ling, VA20164-5405USA. ISASI, 107E.HollyAve.,Suite 11,Ster- and bemadepayabletoISASI. Mailto Checks shouldaccompanythe request ings ited numberofpapercopies quire theCD-ROMfora$75fee.Alim- handling. Non-ISASImembersmayac- nal feeof$15,whichcoverspostageand Proceedings and corporatemembersmaypurchasethe website atwww.isasi.org.Further,active from theappropriatesectionofISASI Proceedings quire, onano-feebasis,copyofthese standing andcorporatemembersmayac- S Deficiencies andSolutions Solutions—100 YearsofIdentifyingSafety “ ments inconcertwiththe and industryaccident-preventiondevelop- investigative techniquesandexperiences, emphasizes accidentinvestigationfindings, Society ofAirSafetyInvestigators.Content tors whoaremembersoftheInternational lished forprofessionalairsafetyinvestiga- ings Publisher’s EditorialProfile: the editorialoffices. reprint isavailableuponapplicationto hibited withoutpermission.Permissionto reserved. Publicationinanyformispro- ety ofAirSafetyInvestigators,allrights Copyright ©2004 plied bytheauthors. herein intheoriginaleditorialcontentsup- community. Thepapersarepresented on safetyissuesofinteresttotheaviation Washington, D.C., Annual InternationalSeminarheldin Notice: Internet website:http://w 1745. E-mailaddress:[email protected] phone (703)430-9668.Fax:450- 11, Sterling,VA20164-5405 Editorial Offices: position orpolicy. sarily endorsedorrepresentofficialISASI ions expressedbyauthorsarenotneces- Society ofAirSafetyInvestigators.Opin- published annuallybytheInternational ISASI Proceedings From theWrightBrotherstoRight ubscriptions: ISASI (FrankDelGandio,President) is printedintheUnitedStatesandpub- 2002 isavailableatacostof $150. VOL The Esperison Martinez on aCD-ROMforthenomi- by downloadingthematerial Richard B.Stone ISSN 1562-8914 UME 7,NUMBER1 William A.Ford Editorial Advisor Proceedings Editorial Staff Susan Fager Activemembersingood Publisher Design

107 E.HollyAve.,Suite —International Soci- features presentations (ISSN 1562-8914)is of theISASI3 w s w.isasi.org .” eminar theme

ISASI Proceed-

USA. Proceed- Tele- 4th . International Seminar ‘From theWrightBrothersto 100 YearsofIdentifyingSafety August 26–28,2003•Washington,D.C.,USA Proceedings ofthe Air SafetyThroughInvestigation Deficiencies andSolutions’ the RightSolutions— 34th Annual ISASI 2003 Proceedings • 1 ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS 2 Latin American, Russian, United States, European, New Zealand, Canadian, SESA-France Chap., Australian, S N Southeastern, San Francisco, Great Lakes, Rocky Mountain, Florida, Pacific Northwest, Dallas-Ft. Worth, Northeast, Alaska, U New Zealand, European, Canadian, Australian, International, United States, C Treasurer, Secretary, Vice-President, Executive Advisor, President, O Mid-Atlantic, Arizona, C • OCIETY OUNCILLORS HAPTER ATIONAL AND NITED FFICERS ISA SI 2003 CraigBeldsoe BenColeman BillWaldock V.Venkov KeithHagy TomMcCarthy FrankDelGandio DavidW.Graham KenSmart MaxSaint-Germain BarbaraDunn BarbaraM.Dunn KennethS.Lewis LindsayNaylor S P RodneySchaeffer

TATES Inactive Ron Schleede CurtLewis CurtLewis P CajFrostell RonChippindale RonChippindale P

RESIDENTS PeterAxelrod Ron Schleede r oceedings RESIDENTS Marco A.deM.Rocha RichardL.Perry

KevinDarcy RichardStone Curt Lewis VincentFave R R EGIONAL EGIONAL ISASI Information C Global Aerospace, Inc. GE AircraftEngines FTI Consulting,Inc. Flight SafetyFoundation—Taiwan Flightscape, Inc. Finnair Oyj FedEx PilotsAssociation Federal AviationAdministration Exponent, Inc. EVA AirwaysCorporation Era Aviation,Inc. Emirates Airline Embry-Riddle AeronauticalUniversity EMBRAER—Empresa Brasileirade Dutch TransportSafetyBoard Directorate ofFlyingSafety—ADF Directorate ofFlightSafety(CanadianForces) Delta AirLines,Inc. DCI/Branch AIRCO Continental Express Continental Airlines COMAIR, Inc. Civil AviationSafetyAuthority,Australia Cavok, International,Inc. Cathay PacificAirwaysLimited Bombardier AerospaceRegionalAircraft/ Bombardier Aerospace Boeing CommercialAirplanes Board ofAccidentInvestigation—Sweden Bell HelicopterTextron,Inc. Belgian AirForce,StaffBrussels,VSF Avions deTransportRegional(ATR) Aviation SafetyCouncil Avianca &SAMAirlines Australian TransportSafetyBureau Austin Digital,Inc. Atlantic SoutheastAirlines—DeltaConnection Association ofProfessionalFlightAttendants ASPA Mexico American UnderwaterSearch&Survey,Ltd. American EagleAirlines American Airlines Allied PilotsAssociation All NipponAirwaysCompanyLimited Alaska Airlines AirTran Airways Airservices Australia Airclaims Limited Airbus S.A.S. Air NewZealand,Ltd. Air LinePilotsAssociation Air CanadaPilotsAssociation Air Canada Air AccidentsInvestigationBranch—U.K. Air AccidentInvestigationUnit—Ireland Air AccidentInvestigationBureau Aeronautical &MaritimeResearch ACE USAAerospace Accident InvestigationBoard/Norway Accident InvestigationBoard,Finland Aer de Havilland,Inc. of Singapor Laboratory ORPORATE onautica S.A. e M EMBERS WestJet Volvo AeroCorporation University ofSouthernCalifornia University ofNSWAVIATION U.K.-Civil AviationAuthority Transportation SafetyBoardofCanada Transport CanadaAviation The MinistryofLand,Infrastructure, TAM BrazilianAirlines SystemWare, Inc. Southwest AirlinesCompany Southern CaliforniaSafetyInstitute South AfricanAirways SNECMA Moteurs Smith, Anderson,Blount,Dorsett, Singapore Airlines,Ltd. Sikorsky AircraftCorporation SICOFAA/SPS School ofAviationSafetyandManagement, Scandinavian Airlines Saudi ArabianAirlines Sandia NationalLaboratories Royal NewZealandAirForce Rolls- RoyceCorporation Republic ofSingaporeAirForce Qantas AirwaysLimited Pratt &Whitney Phoenix International,Inc. Northwest Airlines NAV Canada National TransportationSafetyBoard National BusinessAviationAssociation National AirTrafficControllersAssn. National AeronauticsandSpace Middle EastAirlines Lufthansa GermanAirlines Learjet, Inc. L-3 CommunicationsAviationRecorders KLM RoyalDutchAirlines JetBlue Airways Japanese AviationInsurancePool Japan AirSystemCo.,Ltd. Interstate AviationCommittee Int’l. Assoc.ofMach.&AerospaceWorkers Independent PilotsAssociation IFALPA Hong KongCivilAviationDepartment Hong KongAirlinePilotsAssociation Honeywell Hall &AssociatesLLC & T Mitchell &Jernigan,L.L.P ROC AirF Administration ransport, AAIC,Japan N or ce Academy . 77 75 SESSION IV 69 64 60 56 54 SESSION III 49 43 38 34 SESSION II 26 17 11 8 SESSION I 5 4 2 By WilliamD.Waldock, Embry-RiddleAeronauticalUniversity Advances inDesign—AView formtheWreckage Protection ThroughCrashworthiness Evaluationand Crashworthiness Investigation: EnhancedOccupant By JohnCarr,President,NationalAir Traffic ControllersAssociation,USA Growth ofATCSystemand ControllersUnion By MichaelR.Poole,Flightscape Accident InvestigationWithoutthe By CandaceK.Kolander,AssociationofFlightAttendants in AccidentInvestigations Historical ReviewofFlightAttendantParticipation By JayGraser,GalaxyScientificCorporation Technology Investigation EnhancementThroughInformation By StéphaneCorcosandGéraldGaubert,BEA,France Otter Accident,March2001 Investigating TechniquesUsedforDHC-6Twin By MarionC.Blakey,Administrator,FAA,USA Learning for‘KickingTin’ Research, UniversityofSouthampton,UK By StuartDyne,ISVRConsulting,InstituteofSoundandVibration Failure Decompressions CVR RecordingsofExplosionsandStructural Council Taiwan.PresentedbyVictorLiang. By Wen-Lin,Guan,VictorLiang,PhilTai,andKayYong,AviationSafety ion TechnologyattheAircraftAccidentInvestigation Application ofthe3-DSoftwareWreckageReconstruc- Taiwan, ROC By DavidLee,StevenSu,andKayYong,AviationSafetyCouncil, Comparisons andLessonsLearned CI611 andGE791WreckageRecoveryOperations— By JohnW.Purvis,SafetyServicesInternational Accident Reconstruction—TheDecisionProcess NASA AmesResearchCenter By BarbaraK.Burian,R.KeyDismukes,andImmanuelBarshi, The EmergencyandAbnormalSituationsProject Space Museum,USA By JeffGuzzetti,NTSB,USA,andBrianNicklas,NationalAir ical PerspectiveofAircraftAccidentInvestigation From theWrightFlyertoSpaceShuttle:AHistor- By Jean-PierreDagon,DirectorofCorporateSafety,AirTranAirways dent/Incident Investigations (RCA) UsingReason The PracticalUseoftheRootCauseAnalysisSystem By EllenG.Engleman,Chairman,NTSB,USA Human SpiritandAccomplishmentAreUnlimited By EsperisonMartinez Caj Frostell:2003LedererAwardWinner By FrankDelGandio Preface ISASI Information ® : ABuildingBlockforAcci- Table ofContents 179 173 168 160 TheAccidentDatabaseoftheCabinSafetyResearch 146 141 135 133 SESSION VI 126 108 105 100 97 SESSION V 93 90 85 82 ISASI PictorialReview By MarkGeorge,NTSB,USA tions RelatingtoInflightFire Emergencies National TransportationSafetyBoardRecommenda- By StevenSaintAmour,PhoenixInternational,Inc. Search &Recovery:TheArtandScience By RayCherry,R.G.W.Cherry&AssociatesLimited,UK Technical Group By ThomasA.Farrier,AirTransportAssociationofAmerica,Inc. Revisiting thePasttoLookFuture Investigating SurvivalFactorsinAircraftAccidents: By ScottA.Warren,NTSB,USA Accident Investigation Use ofComputedTomographyImagingin By TimothyJ.Logan,SouthwestAirlines We Going? Airline SafetyData:WhereAreWeand Prevention andInvestigationCenter(CENIPA),Brazil By Col.MarcusA.AraújodaCosta,ChiefAeronauticalAccident Accident InvestigationinBrazil By RobertMatthews,FAA,USA U.S. Carriers,1987-2002 Ramp AccidentsandIncidentsInvolving By MichaelHuhn,AirLinePilotsAssociation.PresentedbyChrisBaum. Aircraft AccidentReports Recommendation fromSelectedRecentNTSB An AnalysisoftheRelationshipFinding-Cause- By MikeHorne,ADAerospace,Ltd.,Manchester,UK Flightdeck ImageRecordingonCommercialAircraft By PippaMoore,CAA,UK Flightdeck ImageRecordingonCommercialAircraft By JimBurin,FlightSafetyFoundation Killers inAviation The CFITandALARChallenge:Attackingthe By CajFrostell,Chief,AccidentInvestigationandPrevention,ICAO Accident andIncidentInvestigation Audit ProgramtoIncludeAnnex13—Aircraft Expansion oftheICAOUniversalSafetyOversight By AllenJ.Parmet,MidwestOccupationalMedicine,KansasCity,Mo. Aerospace andPreventiveMedicine Aircraft AccidentInvestigation—TheRoleof By MaryCimrmancic,TransportationSafetyInstitute,OklahomaCity,Okla. Victim Identification Forensic AspectsofOccupantProtection: Commonwealth ofVirginia By WilliamT.Gormley,OfficeoftheChiefMedicalExaminer, Pattern Analysis Enhanced OccupantProtectionThroughInjury ISASI 2003 Proceedings • 3 ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS 4 • air powerandspaceexplorations. ability, becauseoftheiraccomplishments inthemilitaryuseof airspace systemandgreatlyimproved aviationsafetyandreli- be themilitaryandspacepioneers whohaveimprovedour fostering andimprovingourindustry.Anothergroupwould goal. Thesefolkshavebeenveryimportantandinfluential in to carrymorepassengers,orsomethingsurpassaprevious lades becausetheirgoalistoattainanewaltitude,speed, vent furtherreoccurrence. improve theman,machine,andenvironmenttopre- a resultofaccidentinvestigationandourconstantquest to are thestarsandlegendsofourprofession.Thesecond is followed likeJerryLederer,whoisherewithustoday.They ideas ordreamsliketheWrightbrothersandmanywho the directresultoftwothings:Thefirstisthatpeoplewith cident” or,moreappropriatelyby“accidentinvestigation.” I canmakethestatementthatourindustryhas“grownbyac- pened byaccident”wouldbeappropriate.However,Ibelieve industry. Normallyastatementsuchas“Thishasnothap- flight. days torepairtheFlyerinpreparationforhistoricfirst reaction, “Thereisnoquestionoffinalsuccess.”Ittookthree The aircraftnosedup,stalled,anddoveintothedunes.Wilbur’s it butwassurprisedbythesensitivityofaircraft’selevator. 14, WilburtriedtocoaxtheFlyerintoair.Healmostmade 1903, threedaysbeforethehistoricfirstflight.OnDecember why theaircraftcrashed. which resultedinaninvestigationbytheWrightbrothersasto accident investigation.Thefourthflightcrashedonlanding, ered flightonDecember17,wealsocelebrate100yearsof and traveled852feet.Butaswecelebrate100yearsofpow- accomplished fourflights,andthelastflightwas59seconds 121 feetoverthesanddunes.Beforedaywasthey their dream.Thefirstflightlasted12secondsandtraveled Finally, onDec.17,1903,afternumerousfailures,theyachieved nels wheretheyconstantlyimprovedthewingsandstructure. preponderance oftheirworkwaswithglidersandwindtun- I in Washington,D.C.—Editor) (President DelGandio’swelcomingremarkstoISASI2003attendees ISA The groupthatwearemostconcerned aboutistheunsung The peoplewithideasanddreamsusuallyreceivetheacco- The phenomenalimprovementinsafety,Ibelieve,hasbeen Aviation hasprogressedandexpandedfasterthananyother In reality,thefirstaircraftaccidenthappenedonDec.14, SI 2003 ing beforetheyachievedtheirgoalofpoweredflight.The light. OrvilleandWilburspentmanyyearsexperiment- n afewmonthswewillcelebrate100yearsofpowered P r oceedings 100 YearsofPoweredFlight By FrankDelGandio,President PREF tive skills. safety becauseofyourideas, your dreams,orinvestiga- nized in50yearsashavingdriven thosenextgreatleapsin bility isthatafewfolkssitting rightheretodaywillberecog- landish in50yearsorsooner. Arelatedhopeandrealpossi- improve atsuchapacethattoday’ssafetylevelswillseemout- the centennialcelebration.Ihopethatsafetycontinues to flight, peoplewillspeakoftheenormousgainsmadesince learn fromoneanother. three days,wewilldowhatasinvestigatorsbest.We plishing extraordinarythings.Aswemeethereoverthenext as aircraftaccidentinvestigatorsareordinarypeopleaccom- efficiency ofourindustry. faction fromaccomplishmentsthatenhancethesafetyand notoriety oraccolades.Butwedon’tneedit.Wegetoursatis- to helpsolvethemystery.Traditionally,wedon’tgetmuch everyone intheindustry,andwe—asinvestigators—areeager training, dispatching,etc.Whenadisasterhappens,itaffects we participateintheairspacesystembyflying,maintaining, aviation community.Weareapartofthemaingroupbecause sastrous accidentsandincidents. the workofinvestigatorsfollowingairdisastersandnon-di- of changeshavebeendevelopedandincorporatedbecause dent andinitiatechangetopreventreoccurrence.Thousands we accomplishourtaskofdeterminingwhatcausedtheacci- industry, butthestarsofaccidentinvestigationshinebrightas today. Whenanairdisasteroccurs,itbringsdarknesstoour heroes ofaccidentinvestigationandmanythemarehere ACE The realhopeisthatatthe150thanniversaryoffirst Our professionisaclassicexampleofintrinsicrewards.We We, asaccidentinvestigators,areanintegralpartofthe N

E. MARTINEZ aviation safety,especiallyinthose remoteareasthattrulyneed cillor. Heisabletotravelthe worldandspreadthewordof supporter ofISASIwherehe serves asitsInternationalCoun- tremely tensesituations,”Del Gandiotoldtheaudience. muster tobringaboutsuccessful conclusions,oftenunderex- His assignmentsrequireeveryounceofdiplomacyhecan commitment toaircraftaccidentinvestigationandprevention. yet hedisplaysatalentforfairnessthatcontinuestoreflect his the internationalroleheplays,mustremainindependent, organization andtotheworldofaviationsafety.Because ing oftheICAOaccidentinvestigationmanual. 13. Heiscurrentlydeeplyinvolvedintheoverhaulandrewrit- 99 meeting,bothofwhichresultedinmajorrevisionstoAnnex on AIG92andwasresponsibleforthesuccessofrecent gators, aswellothermajorissuesandpublications.Heworked he isresponsibleforAnnex13,thebibleofworld’sinvesti- the internationaleffortstopromoteaviationsafety.Onjob gation andPreventionSection(AIG),Cajplaysamajorrolein 70s, heinvestigatedsome300aviationaccidents. chief oftheaccidentinvestigationsectionin1972.During ticipated inmoreaccidentinvestigationsandbeganactingas thority (CAA)asanairworthinessinspector.Graduallyhepar- gan hiscivilaviationcareerattheFinnishCivilAviationAu- accident investigationsysteminhisnativeFinland.”Hebe- tion. Further,hecanbejustlycalledthecreatorofaviation ship amongtheworld’sexpertsinaircraftaccidentinvestiga- beacon ofdedication,objectivity,professionalism,andleader- the highestawardyoucangetfromISASI.Cajstandsoutasa most prestigiousawardyoucangetinaccidentinvestigation, dent FrankDelGandiosaid,“TheJerryLedererAwardisthe awards. Inintroducingthewinnertoaudience,ISASIPresi- The presentationisthehighlightofISASI2003seminar Society’s objectivesandtechnicalexcellenceoftherecipient. Investigators (ISASI),theawardrecognizesachievementof Award. the awardtoCajFrostell(MO3596),recipientof2003 office forNASA.At101yearsold,hewasonhandtopresent Flight SafetyFoundation,andorganizedthefirstflightsafety Air SafetyintheCivilAeronauticsBoard,established Lindbergh, LedererwasthefirstdirectorofBureau of aviationsafetysinceitsinfancy.AlifelongfriendCharles created byISASItohonorJerryLederer,aleaderintheworld T He added,“Inhispositionin ICAOheisgoodfriendand “All whoworkwithhimconsiderCajasuperbassettothe In hiscurrentpositionwithICAOaschief,AccidentInvesti- Awarded annuallybytheInternationalSocietyofAirSafety accident investigationandprevention.Theawardwas standing lifetimecontributionsinthefieldofaircraft he JeromeF.LedererAwardisconferredforout- 2003 LedererAwardWinner Caj Frostell: By EsperisonMartinez,Editor have transcendedtime;theyare asvalidandrelevanttoday about theneedforchangeand reform.ButJerry’sarticles mary. Itwasfascinatingreading. Atthisseminarwehavetalked the 1930s.Mytaskwastoprepare a1.5-pagejustificationsum- publications writtenbyJerry, muchoftheseworkswerefrom award, Ihadtheopportunity toreadnumerousarticlesand with ICAO. the bestacceptancespeechthatIhaveheardinmy23years found, significant,andembeddedwithexcellenthumor.Itwas Warner AwardonJerry,whoseacceptancespeechwaspro- Assad Kotaite,wasdelightedtopersonallybestowtheEdward ward WarnerAward.ThepresidentoftheICAOCouncil, Dr. 1999, JerryLedererreceivedICAO’shighestaward,theEd- great honor.Thankyouverymuch,Jerry.” He said,“Iamoverwhelmed.Thisisagreatsurpriseand 2003 JerryLedererAwardwinner.” uniquely qualifiedtoreceivethehonorofbeingnamed person ofsuchdedication,energy,andtalent.CajFrostellis dividual initsranks.Weareluckytohavethesupportofa jects. ISASIisindeedblessedtohavesuchanoutstandingin- background informationandmaterialsonvariousICAOsub- for ISASImembersgivingpapers,readilyprovidingneeded Dar-es-Salaam, andSanJose).Further,heisawelcomesource nars, participatinginatleastfive(Prague,Beirut,NewDelhi, it. HehasbeenamajorcontributortoISASI’sReachoutsemi- Looking onisISASIPresidentFrankDelGandio. the covetedJeromeLedererAwardtoCajFrostell,right. Jerry Lederer,centemakesafewcommentsafterpresenting “Behind thescene,inICAOselectionprocessfor In aswitchofroles,hepraisedtheaward’snamesake:“In After theacceptanceceremony,Cajaddressedaudience. ISASI 2003 Proceedings • E. MARTINEZ 5 ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS 6 • Singapore, toconveymythankstheSingaporeAviation ask Mr.ChanWingKeong,thedirectorofAAIB in the UniversityofSouthernCaliforniaandMikeBarr.May I ISASI corporatemembers.Iwishtoacknowledgeandthank with severalofthem.Manythesetraininginstitutesarealso stitutes arealsoclosetomyheart,andIhavebeeninvolved Ron inMozambiquesetthestageforalifelongfriendship. accredited representativeofMozambique.The3monthswith Mozambique wasfatallyinjured.RonandIwereassistingthe dent justinsideSouthAfricainwhichthepresidentof in 1986-1987ontheSamoraMachelaccident,aTU-134acci- edge RonChippindale,whomIworkedwithfor2-3months dent ofISASIsome10-12yearsago.Ialsowishtoacknowl- 23 yearsago.ManyofyourememberOlofasaformerpresi- tunity toflyanairplane. cess atthattime.Simplystated,Iwasfascinatedbytheoppor- safety andaccidentinvestigationwasnotinmythoughtpro- and flighttrainingsome38yearsago.AndIadmitthat ish AirForceformyfirstexposuretoaviation,officerschool, vestigator intheFinnishAirForce.IamgratefultoFinn- Finnair, andLieutenantColonelJaakkoSaatsi,thechiefin- to acknowledgeCapt.PekkaKärmeniemi,safetymanagerwith had twootherparticipantsfromFinlandatthisseminar.Iwish country Finland,asmallwithfivemillionpeople.We Reachout program. financial supportisessentialforustobeablecarryoutthe ‘Reachout.’ Ialsowishtothankthecorporatesponsorswhose seminar. JimandLadicouldwellbecalledthefathersof the CzechRepublicashostoffirstISASIReachout tee, forhisexcellentwork,andIwishtothankLadiMikafrom to thankJimStewart,thechairmanofReachoutCommit- tion oftheinternationalforumthatISASIrepresents. tion ofnumerousaccidentinvestigationagenciesasanindica- ated verymuchhisopeningofthisseminarandtheintroduc- Gandio fortheexcellentwayheisleadingISASI.Iappreci- safety prophet.” they wereinthe1930s.Theyaretrueproofofanaviation ISA “The aviationsafetyandaccidentinvestigationtrainingin- “I alsowishtothankMr.OlofFritsch,whohiredmeICAO “This isnotonlyanhonorforme,italsomy “The ISASIReachoutseminarsareclosetomyheart.Iwish He thenturnedtothepresent:“IwishthankFrankDel SI 2003 P r oceedings in particular.Thankyou.” much. MayGodblessyouall,andmayJerryLederer, numerous friendsintheaudience.Iwishtothankyouvery Republic. grams intheUSAandnewcoursesPrague,Czech John Purvis,andRonSchleedeforinvolvingmeintheirpro- California SafetyInstitute,MarleneFoulk,GaryMorphew, courses. Andlastbutnotleast,IwishtothanktheSouthern Academy forinvolvingmeintheiraccidentinvestigation “I apologizethattimedoesnotpermitmetomentionall 2002—Ronald L.Schleede 2001—John PurvisandtheTransportationSafety 2000—Nora Marshal 1999—Capt. JamesMcIntyre 1998—A. FrankTaylor 1997—Gus Economy 1996—Burt Chesterfield 1995—Dr. JohnK.Lauber 1994—U.K. AircraftAccidentsInvestigationBranch 1993—Capt. VictorHewes 1992—Paul R.Powers 1991—Eddie J.Trimble 1990—Olof Fritsch 1989—Aage A.Roed 1988—H. VincentLaChapelle 1987—Dr. CarolA.Roberts 1986—Geoffrey C.Wilkinson 1985—Dr. JohnKenyonMason 1984—George B.Parker 1983—C.O. Miller 1982—C.H. PraterHouge 1981—Dr. S.HarryRobertson 1980—John GilbertBoulding 1979—Gerard M.Bruggink 1978—Allen R.McMahan 1977—Samuel M.Phillips Past LedererAwardwinners Boar d ofCanada N SESSION I ISASI 2003 Proceedings • 7 ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS and willcontinue tosaythattheNTSB’scredibility isbasedon tory authority.Ourpowerlies solely inourcredibility.Ihavestated and practices.Asyouallknow, theNTSBdoesnothaveregula- flight andawillingnesstoquestion commonlyacceptedtheories were aresultofmeticulousinvestigation intotheproblemsof to learnfromthanwedotoda y, butalloftheirimprovements tion. Theearlypioneershad many moremishapsandaccidents no differentthanourmodernapproachtoaccidentinvestiga- as accidentinvestigators,butclearlytheirapproachtoaviation is one oftheresultsthis. the accidentinvestigationprocess.TheindependentNTSB is tigation changesweremadetobothimproveaviationsafety and Unfortunately otheraccidentswouldfollow,andwitheach inves- the deathofLt.ThomasSelridgeatFortMeyer,Va.,in1908. five yearsaftertheWrightbrother’shistoricflightandwas due to dent investigations. are allprinciplesthatweadheretotodaywhenconductacci- eses, questioningcommonlyheldbeliefs,andadesiretocutcosts of information,theusescientifictestingtosupporthypoth- enormously importantinaviationtoday.Internationalsharing world ofthe19thcentury. peck—an approachtoproblemsolvingthatwasstandardthe results ofeachtest.Theywouldallownoguesswork,huntand methodically, keepingmeticulousnotesonthevariationsand monly acceptedscientificdata.Theyapproachedeachproblem glider experiments,theybegantoquestionsomeofthecom- the bestscientificprinciples.Asaresultofanalyzingtheirown flight, theyweremeticulousintheirexperimentsandadheredto human spiritandaccomplishmentareunlimited. more than3trillionmilesofpassengerflightintheyear2000, to thedevelopmentofaninternationalairlineindustrythathad flight of12secondsthatwent121feetataheightabout10 we lookbackinamazementatthelast100years,fromawobbly to joinyouthismorning. and VickyAndersontheISASImembershipforinvitingme it isanhonortotalkwithyouthismorning. Carmody, aswellthe429familymembersofNTSBteam, 8 • We maynotlabeltheWrightbrothersandotherearlypioneers The firstofficialinvestigationofanaviationaccidentoccurred The qualitiesthatmadetheWrightbrothersasuccessarestill As theWrightbrothersworkedtowardtheirgoalofhuman Much istobecelebratedwiththecentennialyearofflight.As Thanks toFrankDelGandio,RonSchleede,NoraMarshall, ISA SI 2003 Accomplishment AreUnlimited P By EllenG.Engleman,Chairman,NationalTransportationSafetyBoard,KeynoteSpeaker r oceedings Rosenker, MembersGoglia,Healing,and bers andstaff.OnbehalfofVice-Chairman work withanamazingteamoffellowBoardMem- fessionals andenjoytheuniqueopportunityto low inthefootstepsofdedicatedandgiftedpro- the NationalTransportationSafetyBoard.Ifol- It isaprivilegetoserveasthe10thchairmanof Human Spiritand participates in these investigationsthrough an NTSB-appointed rence isresponsiblefortheinvestigation, theU.S.government U.S. registry,manufacture,or design,whiletheStateofoccur- in aforeignstateinvolvingcivil aircraftofaU.S.operatoror tional CivilAviation.Thismeans thatforanaccidentorincident presented inAnnex13tothe ChicagoConventiononInterna- of Staterequirements,tofulfill theobligationsofUnitedStates protect. may comethepromiseofasaferfuture.Maywelearninorder to them thattheworkofNTSBisfocused—thatouttragedy an accidentoccurredandthatliveswerelost.Butitisintribute to and thoughtswheneverythingyousayisbasedonthefact that understand. me thefirsttime,theyhopedtoneverhavemeetagain. I than onepersontellmethatwhiletheyweredelightedto meet As youknow,theroleofNTSBisunique—Ihavehadmore fidence inthenationalandinternationaltransportationsystem. try backinthesky.Ourultimatemissionistoensurepubliccon- internationally andnationally—worktogethertogetthisindus- ment, maintenance,thepilots,flightattendants,airports— critical thatallpartnersinthisindustry,manufacturers,manage- tries suchasthetravelandhospitalityindustries.Thereforeitis lion, andwehaven’tevenincludedtheimpactofrelatedindus- revenues ofthetop150airlinesgroupsareestimated$300bil- product intheUnitedStatesis$306billion.Internationally, est weaponsagainstterrorism. tinue. Letusrememberthateconomicstrengthisoneofthegreat- and mustbeachievedinorderforpeaceprosperitytocon- be safetyANDsecurity.Thereisabalancethatwillachieved without jeopardizingornegativelyimpactingtheother.Itmust post 9/11world,wemustfindawaytoaccomplishbothtasks or canbeaquestionofchoicebetweensafetyORsecurity.In main thesame—theissueofsafety.Idonotbelievethatthereis flight, oneconstantremainsthesame,however,andmustre- redesign. yield newsafetyknowledgefromfueltankinertingandrudder and awillingnesstoquestionacceptedtheoriespracticeswill a curiositytoexplainwhathappenedwhensomethinggoeswrong, and tootherfederalagenciestheDOT,includingFAA. safety recommendationsthatwesendtoindustry,the50states, assessments andallowsustogoforthissuethesignificant gives theNTSBitsworldwidecredibilityforunbiased,fact-based issuing oursafetyrecommendations.Itisthisstrictdisciplinethat desire inmakingourdeterminationsofprobablecauseaswell our useoffact,science,anddata,NOTsupposition,guess,or The NTSBisresponsible,consistentwiththeU.S.Department It’s sometimeshardtodeterminehowframeone’swords The directimpactoftheairlineindustryongrossdomestic As wereviewthepastandlooktonexthundredyearsof Constant reviewofdatafromaccidentsandnormaloperations, certification issues andcanalsohaveresultsthat return.Lastyear often affecttheinternational community throughstandardsand by theFederalMotorCarrier SafetyAdministration. as illustratedthenewhoursof servicerulesrecentlycompleted are archaic.Ihopethatallmodes willsoonrespondtothisissue and regulationsgoverningthis aspectoftransportationsafety rest, inanymodeoftransportation,isdangerous.Thelaws, rules, Human Fatigue publish theinformationinpilottrainingandoperatingmanuals. speeds inbothicingandnon-icingconditionsthatcarriers tures ofturbine-engineaircraftclarifyminimumsafeoperating merited. TheNTSBrecommendsthattheFAAensuremanufac- airplanes currentlycertificatedforflightinicingconditions, is tion testprogram,includingapplicationofrevisedstandards to Icing gate. land, withthecenterfueltankexplosionthatoccurredat the tragedythatoccurredonMarch3,2001,inBangkok,Thai- certify thesystem.Soonerisbetterthanlater.Wecannotforget to preventfueltankexplosions,theyhavenotsetadeadline working withBoeingtotestafueltankinertingsystemdesigned explosive fuel-airmixturesinfueltanks.TheFAAiscurrently prevent operatorsfromflyingtransport-categoryaircraftwith Center WingFuelTanks trated, theissueisnotyetresolved. 31, 2000,andtheaccidentinMilan,Italy,onOct.8,2001,illus- approach tosafety.ButasthetragedyinTaipei,aiwan,onOct. provide immediateimprovementsandsupportamultilayered use of24-hourrunwayguardlightsatLasVegaswillhopefully tus lightingsystemtobeinstalledatDallasFortWorthandthe ports withthehighestrisk.InUnitedStates,runwaysta- or phased-inapproaches,focusingonthedozensoofair- tech solutionandmustfindimplementlow-techalternatives Runway incursions are areasofinteresttousaswecontinuetheseendeavors: protecting livesmustbeonparallelifnotoverlappingpaths.Here achieving safety,ourgoals,mission,anddedicationto should notbeinterpretedasadversarial.Wemustpartnersin plish thistask.IfiercelybelievethattheNTSB’sindependence cused ineachmode,withthestatesandindustrytoaccom- up therecordofoutstandingrecommendations,andwearefo- pleted. SoakeyaspectofmytenureattheBoardwillbetoclean that havebeenidentifiedstillremainandactionisyettobecom- Open recommendationsmeanthatourjobisnotdone.Therisks we had1,025openrecommendations. while thatisgoodonitsface,whenIcametothisofficeinMarch sued morethan12,000with80+percentacceptancerate;and cause isdetermined,weissueourrecommendations.Whaveis- ing theprobablecauseirrespectiveoffault.Oncethat must remainsoinordertoaccomplishourmissionofdetermin- the U.S.-manufacturedfleet. participate inforeigninvestigationsandmaintainthehealthof the safeoperationofaircraft.Thusourroleistoappropriately transmit informationtomaintaincontinuedairworthinessand named bytheNTSB.TheUnitedStatesisalsoresponsibleto and accreditedrepresentativeateamoftechnicaladvisors As youknow,recommendations thatwehavemadetotheFAA Open recommendationsmeanthatthesafetyloopisnotclosed. As youknow,theNTSBisafiercelyindependentagencythat . Icingisacontinuedseriousproblem.Athoroughcertifica- . Operatinganyvehicleorvesselwithoutadequate . Wecan’taffordtowaitfortheperfecthigh- . TheFAAmustcompletearulemakingto for research,development andimplementation ofnewtechnolo- roundtable discussionsandforums, formulatesafetypartnerships training programs,offeropportunities forsafetysymposia, safety partnership.Itwillhouse theNTSBinvestigationandsafety ton UniversityinAshburn,Va.,andoffersnewopportunities for This leasedfacilityislocatedon thegroundsofGeorgeWashing- new recommendations,sothebeatgoeson. accidents, andEmeryWorldwideFlight17.Wehavealsoissued 47 the most-wantedlist,15passengervans,driverdistraction,two rail review andheldeightmeetingspublichearingsthatincluded hearings. Wehavesavedmorethan$250,000viaprocurement families, andourlawjudgeshaveclosed131casesheld 40 We havefieldedmorethan1,350callsfromthemediaorvictims’ of theChinaAirBoeing747thatcrashedinstraitsFormosa. 14 fatalities,andtheNTSBcontinuestosupportinvestigation Boeing 737with116fatalities;Kenya,aFairchildMetroliner with that crashedaftertakeoffwith102fatalities;SudanAirways,a 112 accidentinvestigations,includingAirAlgerie,aBoeing737 Now that’saperfectworld,butitisgoalaswell. eral aviationandotherswouldbefinishedinoneyearorless. major accidentreportwouldtakelongerthan2years,andgen- our efficiencywithoutaffectingthequality.Inaperfectworld,no focused oninternalreviewofprocessestoseeifwecanincrease hurried forfalseorartificialdeadlines.Thatbeingsaid,Iam reports. Yes,theymustbethoroughlydeveloped,andcannot to determineifthereisawayincreasethetimelinessof supposition, ordesire.Ourinternalproceduresarebeingreviewed based onscience,data,andfactsunswayedbyguesswork, deliver anaccidentinvestigationreportthatissoundlydeveloped product delivery.TheSafetyBoardmustimproveourabilityto on increasedperformance,fiscalmanagement,andqualityof advice. As“CEO”oftheBoard,Iamleadingstaffinfocusing We cannotmakerecommendationsifwedonotfollowourown regulator, weareabullypulpitbutIamholdingdailyservices. will beaddressed.Ofcourse,asyouknow,theNTSBisnota retary MinetathattheopenNTSBrecommendationsinallmodes der hermanagementandtheleadershipofTransportationSec- the experienceofsharedmoccasins,andItrulybelievethatun- sued andreceivedrecommendationsfromtheNTSB.Wehave dations, andIwantanupwardslopeonthatgraph. every oneofthem.SinceMarch24wehaveclosed68recommen- open NTSBrecommendationsandwillcontinuetodogeach other DOTmodesinthisSWATteamapproachtoaddressall through theremainingopenrecommendationswitheachof ments intheU.S.system. tively totherecommendationandisworkingmakeimprove- and IamgladtonotethattheFAAhasrecentlyrespondedposi- tion totheFAAaddresspotentialsafetyissuesinU.S.systems, aircraft structures.ThisledtotheBoard’ssafetyrecommenda- factors, airtrafficcontrol,collisionavoidancesystems,and sisted theGermanauthoritieswithexaminationofoperational flight andaTupolevpassengerairliner.Ourinvestigationsas- vestigations ofafatalmidaircollisionbetweenBoeing757cargo an NTSBteamassistedourcolleaguesinGermanywiththein- A newbeginningwillbethe opening oftheNTSBAcademy. And we’reseeingresults.SinceMarch,theNTSBhasconducted Performance andfundingissuesarealsointernaltotheNTSB. Uniquely bothFAAAdministratorBlakeyandIhaveis- We believethatsafetyisjoboneandwillcontinuetowork ISASI 2003 Proceedings • 9 ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS and interests,aplacewheresharedknowledgeopendebate emy willbetheforumforinternationaldiscussiononsharedissues in safety,andtodayisnodifferent.WehopethattheNTSBAcad- topics. gies, andcreateasanctuaryfordiscussionofkeysafetyissues 10 The NationalTransportationSafetyBoardreliesonitspartners • ISA SI 2003 P r oceedings sion canandwillmakeadifferenceinachievingsaferskies. of otherindustryandtransportationleaders,thistimelydiscus- solicit yourcommentsandsupport.Withhelpthe veloping keyissuesthatwillbeappropriatetothisvenueandI government, academiaandinpersonnel.Weareworkingonde- will helpgrowtheoverallbodyofsafetyknowledgeinindustry,

N etrate thesystem’smanydefensive layers tions combinewithlocalcircumstances andactivefailurestopen- ment, whicharepresentforsometimes yearsbeforethesecondi- poor design,ashortfallintraining, inadequacyoftoolsandequip- mitted atthesharpendof system.Alatentfailurestemsfrom failures. Theactivefailureconsists oferrorsorviolationscom- conditions fensive layercominginandoutoftheframeaccordingto local metaphor isbestrepresentedbyamovingpicture,witheach de- weaknesses andgapsakintoaSwisscheese.TheCheese Cheese Theory.AccordingtoProf.JamesReason,eachlayerhas down oftheoneinfront,arecommonlyillustratedby Swiss T George WashingtonUniversity:Safetycourseinsafetydataanalysis. for HighConsequencesIndustriescourse.TSI:SafetyOfficerCourse. College ofGeneva(Switzerland).SandiaNationalLabs:SystemSafety 200: FlightSafetyBoeing,LongBeach.NAIA,ASMEL,CFI&I SAS FlyingAcademy.SA-2000:FlightSafetyInternationalB-717- ment option,BoiseStateUniversity(cumlaude).Typerating:SF-340 2000. Checkairman,SF-340.Education:B.B.A.,aviationmanage- launch customerforsuccessfulSF-340and(notsosuccessful)Saab pilot trainingforoneofthelargestregionalairlinesinEurope,and instructor. Priorto1995:CrossairLTD,Basel,Switzerland;headof addition TCE(TrainingCenterEvaluator)SF-340andSA-2000 Standards—helped obtainfirstPart142certificateinthenation.In FlightSafety Texas,SanAntonioLearningCenter:DirectorofPilot charge oftheOpsSpecs(PartA,B,andC).May1996to1998: carriers; implementedATOSatNWA.Forashortperiodoftimein Oversight andinchargeofcodeshareaudit,safetyauditsregional Airlines, Inc.,Manager,FlightOperationalQualityAssurance. ing) WorkingGroupsince1999.May1998toJune2000:Northwest tion. ParticipationinACOSM(AirCarrierOversightSystemModel- of AIAA,corporatemembersISASIandtheFlightSafetyFounda- As such,therare conjunctionofasetholes insuccessive These holesarecreatedbyacombinationofactiveandlatent The PracticalUseoftheRootCause hind theother,eachguardingagainstpossiblebreak- vention doctrine.Successivelayersofprotection,onebe- he defensein-depthstrategyiscommontoallsafetypre- 2 . Reason Accident/Incident Investigations SA-2000. NativeofSwitzerland(Geneva).Member B-717-200, ratedonDC-9(B-717),SF-340,and Oversight oftheinternalevaluationprocess.Captain FOQA), occupationalsafety,andemergencyresponse. oversight ofoperational(includingASAPand Director CorporateSafety—AirTranAirways— Jean-Pierre Dagon Analysis System(RCA)Using By Jean-PierreDagon,DirectorofCorporateSafety,AirTranAirways(CP0204) , fromJune2000topresent,is 1 ®: ABuildingBlockfor 3 . 2. Changeisproduced bychange. explains onecause,orchange foranylevel). 1. Therecanbeonlyonechange inaset(agroupofanswersthat step (theirconsequent)inthe model. set isagroupoffactorsthatcausally accountforthenexthigher of thatset,andnothingmore. contain onlyfactorsthatarenecessarytoexplaintheconsequence step intheeventoccurred,asetoffactorsisbuilt.Each must As youlistthecomponentfactorsthatexplainwhyaparticular one steptooccurwithinchaintheReason®model. pull-up command,andsoforth. high rateofdescent. signals werepresent.Theairplanewasallowedtocontinue with For example,thepilotdidnotstopwhenconfusingmarshalling akin toallowingthechainofeventscontinueunchallenged. prevent thenextstepinproblem,butdidnot.Theinactionis An inaction fuel truckwasparkedonthesafetyzone,pilot’sscanpoor. over someperiodoftime,i.e.,itwasdark,therampwet, A condition pened thatcausedtheendresult,acollision. lided withaparkedfueltruck. For example,aircraftshipnumber123,Flt.456’sleftwingcol- problem. Theinitialchangecomesfromtheproblemstatement. A changeorchanges built causalpatterns: In rootcauseanalysis,onerecognizesthreebasicelementsthat Basic elementsofrootcauseanalysis,usingReason® dures, processes,manuals,oversight,andtraining. dent/incident investigations.RCAcanleadtochangesinproce- and reliabilityprograms,butisnotalwaysemphasizedinacci- tent conditionsledtotheaccident?” reoccurrence. Itmayofferanavenuetothequestion:“Whatla- and bringsforthsomemeasurableindicatorsofthelikelihood up totherecipientofreportdeterminerootcauses. times tocausalfactors(i.e.,bringingforththefacts)butleaveit the accidenttrajectory. people andassets,accordingtoJamesReason®,ashedefines defenses allowshazardstocomeintodamagingcontactwith Certain ruleshavetobemetin orderforthesystemtowork.A A setoffactsidentifiesallthefactorsthatareessential for Root causeanalysis(RCA)iscommonlyusedinengineering This approachoffersanopportunitytoexaminerootcauses To date,however,accident/incidentinvestigationspointmany : Astateofbeingthatexistedwithintheenvironment : Anythingthatcouldhaveorshouldoccurredto : Anactionthattriggeredanotherstepina The flightcrewdidnotreacttoaGPWS

This isachange—somethinghap- ISASI 2003 Proceedings • 11 a ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS pyramid: ing totheaccident.Theprocessisbasicallystructuredarounda to onefactorwhichaccountsforseveraloftheseeventsetslead- find repeatingpatternsthatcanbeloopedandthusconnected into effecttobringaboutthechanges.Inthisprocess,onewill initial event.Thisallowstheinvestigatortoretracestepsthatcame change isdevelopedwithasetsoffactorsthatcontributedtothe of theincidentbrokendowninsimplebuildingblocks).That The buildingstepsstartwithachange(whichmaybesummary Building stepsforrootcauseanalysis tions fortheset. tion youaredealingwithbeforetheadviceareaactivatesques- This softwarewillaskyoutodesignatewithwhichtypeofcondi- by change,andthosethatarebroughtaboutalackofchange. a conditionineveryset. 4. Conditionscanoccurinanyset,butitisnotnecessarytohave plan didnotwork.) get done,ordidn’thappen,eitherduetoalackofplan,the cannot haveachangeanswerinaninactionset.(Somethingdidn’t 3. Inactionsarealwaysbroughtaboutbyinactions;therefore,you 12 to reoccur.Engineeringsafety bringsforthacomprehensive and changes thatmaynotbenecessary giventhepropensityforevent This approachoffersanalternative toeventuallycostlyengineering Engineering fixesversusorganizational fixes one hasatypicalactivefailure model(avirtualimpossibility). tial atthetopofpyramid. stream analysis,whereasactivefailureswouldmorebeconsequen- It islikelythatlatenteffectswouldbebestdescribedby down- folding, yettheseopportunitieswereeitherignoredorunknown. were manyopportunitiestobreakthechainofeventsfrom un- tom ofthepyramid),wouldindicateabiggerproblem,for there A longerchainofevents,asanalyzeddownstream(oratthe bot- Downstream analysis at thetopofmodel. values arebecomingsmaller;itlendsapredominanceofweight event (thelastchainofthechain).Asyoumovedown, Upstream (topofpyramid)accountsforcriticalstepspriortothe Upstream riskanalysis Note: Thesoftwarestringentlyenforcesthesefourrules. There aretwotypesofconditions:thosethatbroughtabout If oneassumesasinglelevelof eventscausedtheaccident,than • ISA SI 2003 P r oceedings of breakdownin theorganizationalprinciples of control.Afterlook- • • to complywithpolicy. • • sider ineachofthefollowingstatements: At thispoint,thesoftwareofferssupervisionprinciplesto con- B. Supervision-levelactionrequired the statementsarenotapplicable,nextstepisanalyzed. point ofbreakdownintheorganizationalprinciplescontrol. If • tion wasfound. • TORING forcompliance. • pliance withthispolicy. • this specificsituation. • • are management-levelstatements: policy isinplace,enforced,andcontrolsareestablished.Here Management principlesmustbefirstconsideredtoensurethata A. Management-levelactionrequired A rootcausecanbecategorizedinthefollowinghierarchy: Root causesdivision preferred. tion toaidthemindecidingwhetherengineeredcontrolsare event torecurprovidesdecisionmakerswithimportantinforma- coupled withanunderstandingofthepropensityforspecific for preventionofacertainevent.Thiseffectivenesscomparison son® system,thesecontrolscanbecomparedforeffectiveness engineering solutionandorganizationalalternatives.IntheRea- disproportional fixtoasingleandremotepossibilityoffailure. [3] Itbringsforthabulldozerapproachtolevelananthill.A against theMt.St.Odile. FPA mode]desired,leadingintoacontrolledflightterrain the PIC,inlieuoftrack/flightplananglemodeor3.3°[TRK/ combined mode]mayhavebeenleftorerroneouslyselectedby were averticalspeedof3.3(asin3,300fpmdown)[VS/HDG catastrophe. Example:theAirInterA320crashinStrasbourg, automation, orafailureinprogrammingandcouldleadto edge bythepilotwhoisnotcognizantatfirstofafailurein rameters. Iftheautomationfails,itreliesonintuitiveknowl- which isintendedtoalleviatetheworkloadandmonitorpa- ample inpointistheautomationintroducedmodernjets, [2] Itcanintroducenewthreatsbecauseofthefixinitself.Ex- siders theflyingpublic’slikelydistastefor“backwards.” sition foraft-facingpassengerseatscouldbeinterestingifonecon- aft-facing seats.Althoughusedextensivelyinthemilitary,apropo- Example: Consideringaninitiativetoequippassengeraircraftwith [1] Itcanbeimpracticalorhardtomarketfortheindustryatlarge. permanent fix;however,thiscanhavealternativedrawbacks. The pointatwhichastatement canbeaffirmedastrueisthepoint The pointatwhichthestatementcanbeaffirmedastrueis the Root causeanalysisisaprocessdesignedtodiscoverbothan Supervision didnotENFORCE thepolicyinpast. Supervision didnotFOLLOW thepolicyinpast. Supervision didnotPRO Supervision didnotCOMMUNICA Management didnotEST Management didnotENFORCEthepolicywhenaninfrac- Management didnotCOMMUNICA Management didnotestablishameanstoMONITORcom- Management didnotDESIGNA Management didnotCOMMUNICA 6 5 VIDE thethingsnecessaryin or ABLISH apolicytocontr TE thatthispolicyapplyto TE whatwaswanted. TE thisr TE howitwasMONI- equir ement. ol this. 4 der • he mustbeREMOVEDfromthatparticularenvironment. • • policy canbechanged. • C. Individual-levelactionrequired visory level,theindividualperformancemaybeexamined ing atafailurethemanagementlevel,filteringdowntosuper- wet, andsince thesafetyzonelinewasnot visiblefromthe was notvisiblefromthemarshaller’s position.Astherampwas to repaintthelinesperiodically duetowear,thesafetyzoneline safety zoneline,afueltruckwas parkedinthesafetyzone. parking area,andbecausethe fuelerdidnotparkoutsideofthe and becausethefuelerdidnot parkthevehicleinadesignated fueler parkedthetrucktodeliverabillfuelvendor’s office, fueler didnotparkoutsideofthesafetyzoneline.Because the bition againstparkinginthesafetyzonesforofficebusiness, the the vehicle,andbecausefuelcompanydidnothaveaprohi- office business.Sincethefuelerwasnotattentivewhenheparked not haveaprohibitionagainstparkinginthesafetyzones for vehicles indesignatedparkingspotsonly,thefuelcompany did not attentivewhenheparkedthevehicle. the importanceofnotparkinginsafetyzones.So,fueler was importance ofsafetyzonelines,thefuelcompanydidnotstress not establishapolicytoadvisefuelcompanypersonnelonthe nated parkingarea. business process,thefuelerdidnotparkvehicleinadesig- individual(s) didnotcomplyontheirownwiththeestablished of parkingfueltrucksindesignatedareasonlyandthe Because thefuelvendor’ssupervisiondidnotenforcepolicy Narration obtainedbytheReason®software underside oftheleftwing. inside thecontainmentzone,damagingleadingedgeand ward intothegate,itstruckanunattendedfuelhydranttruckleft when notifiedofthewaitingaircraft.Asaircraftmovedfor- not inpositionforanarrivalatC-3,butrantotheirpositions assigned toGateC-3.Therampcrewwasatanothergateand delphia on14Junewith59customersandacrewoffivewas Aircraft #123arrivedatDestinationasFlight456fromPhila- Problem statement take arealexampleofsimplegrounddamage. To illustratethepracticalapplicationofrootcauseanalysis,let’s Application ofRCAusingReason® sequence environmentsseldomcreateintentionalaccidents. part haveaninnatedesireforself-preservation,andinahighcon- ing againsttheacceptedpropositionthatindividualsformost expense offindingasystemicproblemtoanincident/accidentgo- portion ofaRCwheretheindividual(s)shareresponsibility.” sion. UsingtheRCwizardwillhelptoavoidmissingsystemic stating: “SelectinganindividualRCisaseriousandraredeci- Moreover, becausethemanagementdid notestablishapolicy Also, becausethefuelvendordidnotestablishapolicytopark Additionally, becausethecustomerserviceorganizationdid Often anorganizationwillresorttodisciplinaryactionatthe The softwar The individual’sincor The individual’sincor The individual’sincor e isdiligentingivingthisasalast-r r r r ect behaviorcanbeMODIFIED. ect behaviorcannotbechanged,and ect actionisnowacceptableandthe esort option . Figure 1:Leadingedgeslats4and5damaged. Figure 2:Detailedviewofthedamagewithfueltruck. needed thewingwalkers inpositiontoguide theaircraftin, in communicatingwithwingwalkers. Then,becausethemarshaller lished businessprocess,themarshaller didnotfollowprocedure and theindividual(s)didnotcomply ontheirownwiththeestab- sion didnotenforcetheillustration ofhandsignals(SP6720.3 shalling agentwasnotfollowing hertraining,andrampsupervi- wands tosignalherwingwalkers. tablished signalprocedures,themarshallingagentwasusing her ing hertraining,andbecausethemarshallerdidnotfollow es- was understress,andsincethemarshallingagentnotfollow- did notfollowestablishedsignalprocedures.Asthemarshaller on theirownwiththeestablishedbusinessprocess,marshaller tor themarshallingpolicyandindividual(s)didnotcomply her training. on-the-job experience,themarshallingagentwasnotfollowing pilot’s scanwaspoor. the pilothadtoturnovera135-degreeanglepark.So zone linewasuncleartothemarshaller. side ofthesafetyzoneline,fueltruck’spositionto marshaller’s position,andbecausethefuelerdidnotparkout- Meantime, asthemarshaller wasunderstress,andthemar- Then, becausethecustomerserviceorganizationdidnotmoni- Furthermore, becausethemarshallerdidnothaveadequate In addition,becausethe3-Cgaterequiredahigh-angleturn, ISASI 2003 Proceedings • 13 7 ) ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS option, removing 9percentofthismodel. not complyontheirownwiththe establishedbusinessprocess. of parkingfueltrucksindesignated areasonly,andtheindividual(s)did tions andactions of 100percent.Thiseventcontains atypicalmixofbothcondi- identified rootcausestoeachother,givenacalculatedreliability Analysis ofthisinvestigationshowsthatitisvalidtocompare the Interpretation ofthecase Flt. 456’sleftwingcollidedwithaparkedfueltruck. stop theaircraft,whenconfusingsignalswerereceived,aircraft 123, taxied intogateC-3,andbecausethepilotincommanddid not marshaller, andsincethepilot’sscanwaspoor,becauseFlt.456 fuel truck’spositiontothesafetyzonelinewasunclear to the and sinceafueltruckwasparkedinthesafetyzone, the received. Asthefueltruckheightwasabovewing’sleadingedge, in commanddidnotstoptheaircraftwhenconfusingsignalswere constant, andbecausethePICdidnotfollowsafepractices,pilot follow safepractices.Sincethemarshaller’scrossingsignalwasnot safe practiceswhenunclearsignalsarereceived,thePICdidnot crossed themarshaller’scrossingsignalwasnotconstant. cause themarshallingagentdidnotstayinpositionwithherwands agent wasusingherwandstosignalwingwalkers,andbe- not stayinpositionwithherwandscrossed.Sincethemarshalling take theirpositioninatimelyfashion,themarshallingagentdid position usingherwands,andbecausethewingwalkerscouldnot was tryingtocommunicatetheneedforwingwalkersgetin could nottaketheirpositioninatimelyfashion.Sincethemarshaller busy unloadingacargobinatanadjacentgate,thewingwalkers adequate staffingthatday.Consequently,asthewingwalkerswere own withtheestablishedbusinessprocess,rampdidnothave insure adequacyandtheIndividual(s)didnotcomplyontheir customer serviceorganizationdidnotmonitorthestaffinglevelto walkers togetinpositionusingherwands. marshaller wastryingtocommunicatetheneedforwing- 14 cause. Theprocessofeliminationproducestheinterpretationcase. Tree diagramoftheevent.Eachbranchrepresentsanopportunitytoeliminatearoot In termsofpreventingthisproblem, thisistheseventhbest The fuelvendor’ssupervisionhas theopportunitytoenforcepolicy Also, becausethePICwouldnotcomplywithpolicyrequiring Additionally, asseveralrampworkersweresickthatday,andthe • ISA SI 2003 P r oceedings . surprising thatrootcausesoften happenasaresultofaninaction. tions aregre the aforementionedexample. Changesare The treemodel Brief explanationofthetree model cent ofthisproblem. practices whenunclearsignalsarereceived. tion, removing15percentofthismodel. their ownwiththeestablishedbusinessprocess. staffing leveltoensureadequacy,andtheindividual(s)didnotcomply on percent oftheprocessthatproducedthisproblem. own withtheestablishedbusinessprocess. hand signals(sp6720.3),andtheindividual(s)didnotcomplyontheir percent ofthisproblem. with theestablishedbusinessprocess. marshalling policy,andtheindividual(s)didnotcomplyontheirown this problem. lines periodicallyduetowear. deal with22percentofthecausesthatproducedthisproblem. hicles indesignatedparkingspotsonly. this problem. lines. policy toadvisefuelcompanypersonnelontheimportanceofsafetyzone A levelisbestdescribed asacollectionofevents occurringhorizon- This isthesixthbestpreventionoption.Iteliminates12per- The PIChastheopportunitytocomplywithpolicyrequiring safe In termsofpreventingthisproblem,isthethirdbest op- The customerserviceorganizationhastheopportunitytomonitor the This preventionopportunityisthefourthbest,eliminating14 Ramp supervisionhastheopportunitytoenforceillustrationof This optionisthefifthbestavailableoption.Itwillremove13 The customerserviceorganizationhastheopportunitytomonitor This action,theeighthbestoption,willremove7percentof Management hastheopportunitytoestablishapolicyrepaint Preventing thisrootcauseisthesecondbestoptionandwill The fuelvendorhastheopportunitytoestablishapolicyparkve- This isthebestpreventionoption.Iteliminates22percentof The customerserviceorganizationhastheopportunitytoestablisha y circles.Inactionsare above illustrates acompleterootcause analysison grey rounded squares.Itisnot dark squares. Condi- REASON SUMMARYSHEET CHART this problem. to advisefuelcompanypersonnelontheimportanceofsafetyzonelines. the interpretation. same rootcauseeliminates22percentofthemodel,henceforth (first levelofthemodel)followinginitialchange. at therightidenticalbranchesofsecondandthirdconditions actions (mosteffectiveaction)toaffecttheoutcome,onecanlook portion ofthepyramid.TounderstandhowReason®prioritizes tally, henceforthasetnoticethatthereisonechangeperlevel. LISTING OFSPECIFICROOTCAUSES This isthebestpreventionoption.Iteliminates22percent of The customerserviceorganizationhastheopportunitytoestablishapolicy As weeliminateonerootcause,theyareduplicated,i.e., the As thetreemodelbuildsup,considerbottomasflat recurrence, and thereareseveralreasonswhy itshouldbecon- cates thedegreeofquickness and frequencyrelatedtoanevent’s recur quicklyorfrequently.IntheReasonsoftware,TTP indi- is slightlyelevatedyetindicates thattheeventisnotproneto around 3.0arenormal.Thisparticular casehada3.7TTP,which sure ofthepotentialityrecurrence. ticipate arecurrenceofthesameevent.Inmanywaysitis mea- cates howquicklyand/orfrequentlytheorganizationcould an- the discoveredcorrectiveactionsarenotputinplace,TTP indi- “causal stress”presentwithinaspecificeventmodel.Given that lated withintheReason®softwarethatindicatesamount of The tendencytowardprocess(TTP) Tendency towardprocessinterpretation existing conditions. ity TTPisinterpretedtomeanthesame,butwediscount the (discounting thelevelatwhichitoccurs).Thegeneratingcausal- (including conditions)dividedbythepropernumbersoffactors process) isinterpretedtomeantherelativenumberofallfactors top ofthepyramid).ThecausalstressTTP(tendencytoward TTP ischartedinanumeric rangeof0-10.TTPscores to factorsoccurringearlyinthemodel(i.e., Relative givesmoreimportanceandweight Proper meansaequalnumberperlevel. the outcome,heavierweight). these factorsoccur(theclosertheeventto depending onwhichlevelofthemodel isting conditions). changes andinactions(wesubtracttheex- generating causality:Thevalueofall changes inactionsandconditions.Proper proper causalstress:Thevalueofeach to ourinputs).Therawnumbersinclude data, itis100percentreliable(according Since themodelcontainsnoinsufficient Interpretation ofthesummarysheet contributors tothisevent. active failure(individualfailingtostop)as of alatentfailure(systemicissue)andan tunity todamageanairplane. ditions arestillpresentforanotheroppor- could exist,forthisaccidenttoreoccur.Con- do notremovetheconditionsthatexist,or It eliminates12percentofthisproblem. signals arereceived. the policyrequiringsafepracticeswhenunclear Hence thisinterpretation: nals whereconfusing(lastlineofdefense). pilot incommanddidnotstopwhensig- at thefarright-handcorner,firstlevel:The diate contributortotheaccident,looking active failureatthefirstlevelwasanimme- Relative meanstheimportanceassigned In thisexample,weseeabriefoverview By removinghimfromthepicture,we This isthesixthbestpreventionoption. The PIChastheopportunitytocomplywith However italsoworthnoticingthatan 8 number isametriccalcu- ISASI 2003 Proceedings • 15 ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS tion. YetiftheTTPislowforanincident, and iftheyaredesignedcorrectly,provideinstantprotec- Engineered solutionsaresometimesquick—youputtheminplace are themostcostlyoptionsavailablefordealingwithanevent. tional systemseemtobeequallyeffective. engineered controls,evenifsolutionsdealingwiththeorganiza- corrective actionwithamindsetthatoftendefaultstoputtingin ous consequences,itproceedsintothedecision-makingphaseof templated whenprioritizingeventsforcorrectiveaction. 16 covered engineeredsolutions. ganizational fixeswouldbejustasprudentlychosenanydis- slightly abovenormal.Thisscorewouldtendtoindicatethator- as preventativeandmorecosteffective. solutions and/orfixesintheorganizationalsystemthatoftenare tor thatassistsadecision-makerindecidingbetweenengineered TTP numbers.Thusthemetricitselfcanserveasanindica- that arenotasrepetitive.Thesetypesofeventstendtohavelow exceptional, infrequenteventsassociatedwithbusinessprocesses lution. as consistentacontrolfortheseproblemsanengineeredso- with voluminouslyrepetitiveproblemissuesisnotgoingtobe correction. Relyingonorganizationalsystemsandpeopletodeal those businessprocessessothattheydon’trelyonpeoplefor advantages ofengineeredsolutionsinthatthey“dummyproof” Repetitious eventssuchasthesematchwellwiththeinherent Such problemeventstendtohavehighTTPscoresinReason®. be expectedthatthoseproblemswouldhappenagainandagain. lems existinthe“boardingpassengerbusinessprocess”itcan cess. Duetotherepetitivenatureofthisprocess,ifanyprob- ing passengersonaplane,whichisanextremelyrepetitivepro- with them.Anexampleofthisisthebusinessprocessboard- the repetitivenatureofspecificbusinessprocessassociated effective thantheengineeredsolution. organizational systemcanbebothmoreeffectiveandcost ent correctionisnotasgreat.Veryoftenadiscoveredfixinthe Engineered solutionsareindeedofteneffective,yetthey Often, afteranorganizationexperiencesincidentwithseri- In the“aircrafttruck”casestudy,TTPis3.7,whichjust But oftentheeventswedealwith(seriousorotherwise)are Some unwantedeventstendtohappenoveranddue • ISA SI 2003 P r oceedings need foranexpedi- 2 1 References/Footnotes tems Inc.,forbringingtheirexpertiseinthispaper. Laboratories, forintroducingmetoReasonandDecisionSys- member oftechnicalstaffSystemSuretyStudies,SandiaNational I wouldliketoacknowledgePaulW.Werner,Ph.D.,principal Acknowledgements of removingthecause sis investigations.Finallyitpreemptstheold-fashion knowledge. Thistoolalsoaffordsaframeworkforrootcauseanaly- tices” intheindustrytermsofprocedures,processes,andgained reoccurrence ofanevent,andencouragesharing“bestprac- approach toriskanalysis,offeringaninsightinthelikelihoodof mendations followinginvestigations.Inparticularincangivean Hopefully thisapproachwillprovideadditionalweightinrecom- “why” ofanaccident,complementingthe“how”and“when ally) leadingtoanaccident,itmayhelpanswerthesystemic tigation. Itmayassistatlookingasystemicfailure(organization- Reason’s® softwareoffersanadditionalfacettotheaccidentinves- The rootcauseapproachtoincident/accidentinvestigationusing Conclusions 9 8 7 6 5 4 U.K.AAIBrecommendationsAircraftAccidentReportNo:4/90(EW/ 3 REASON®isatrademarkofDecisionSystems,Inc.,companylocatedin Removingtheonewhodidit. CourtesyofScottAlanJones,fromDecisionSystems,Inc.,802N.High SP6710.02,referstoAirTran’sacceptedStationOperationsManual,Stan- TheManagementPrinciples,inquirylogicrulesandcausalpatternsde- BEA,BureauD’EnquêtesandD’AnalysespourlaSécuritédeL’Aviation Longview, Tex.,nottobeconfusedwithDr.JamesReason. Street, Longview,TX75601,website:www.rootcause.com. dard Practice,concerningsignals(Ground-Air). right 1986-2003,andareincludedhereinwithapprovalofDII. tailed inthispaperareprotectedbyDECISIONSystems,Inc.,©Copy- ter. Hypothesisoftheaccident,descriptionat21.58,para.21.4,-1. Mt. St.Odile(LowerRhine),AirbusA-320,F-GGED,operatedbyAirIn- Civile. ReportFED9202120,AccidentoccurredJanuary20,1992bythe senger seats.(Made30March1990). thorities intopassengerseatdesign,withemphasison.2.Aft-facingpas- tiate andexpediteastructuredprogramofresearchwithEuropeanau- Kegworth, Leicestershire,on8January1989.Recommendation4.25:ini- C1095). ReportontheaccidenttoBoeing737-400-G-OBMEnear Publishing Company,1997reprinted1999,pp.10-11. Publishing Company,1997reprinted1999,p.9. Managing theRiskofOrganizationalAccidents Managing theRiskofOrganizationalAccidents 9 , andtheproblemceasestoexist. , Dr.JamesReason,Ashgate , Dr.JamesReason,Ashgate ed N approach . ” Icarus hadflown tooclosetothesun,despite hisfather’srecom- The fatherandsonwere,of course, DaedalusandIcarus.Since were constructedfromfeathers andheldontotheirbodiesbywax. nation ofSicily,inordertoescape prison.Bothutilizedwingsthat ing afatherandsonflyinginformation fromGreece,withadesti- century B.C.,theancientGreeks reportedonanaccidentinvolv- with flight,andalarmedbyits potentialfortragedy.Inthesecond Long beforetheWrightbrothers,mankindwasalwaysfascinated Ancient history have redefinedaviationsafety. continues torevealitselfthroughseverallandmarkaccidents that old adage“themorethingschange,thetheystay same” Hawk, andwhereweneedtogofromhere.Aslookback, the spective onhowfarwehavecomesincethatwindydayat Kitty evolution ofaircraftaccidentinvestigationinordertogain aper- A Guzzetti, whoalsowroteforTheAvion. at Embry-RiddleAeronauticalUniversitywhileattendingcollegewith served asastaffreporterandphotographerforTheAvionnewspaper and heisacontributingeditortoAir&Spacemagazine.Nicklasalso continues towriteon,andphotograph,currentsubjectsinaviation, Brian Nicklas commercial multiengineinstrumentrating. University withadegreeinaeronauticalengineeringandholds ISASI member.HegraduatedfromEmbry-RiddleAeronautical investigated hundredsofgeneralaviationaccidentsandisalong-time Command, andtheFederalAviationAdministration.Guzzettihas positions withtheCessnaAircraftCompany,NavalAirSystems 11 years.PriortocomingtheNTSB,heservedinairsafety From theWrightFlyertoSpace ting thatwe,asairsafetyinvestigators,lookbackonthe of theWrightbrothers’historicfirstflight,itisonlyfit- s theworldpreparestocelebrate100thanniversary Shuttle: AHistoricalPerspectiveof ([email protected]) istheDeputyReference working onahistoricaloverviewofthatfacility.He detailed totheNASALangleyResearchCenter During his15yearsatNASM,hespent1year special interestsareflighttestingandresearch. able aboutmanysubjectsinaerospacehistory,buthis and SpaceMuseum(NASM).Heisveryknowledge- Chief fortheArchivesDivisionofNationalAir investigator-in-charge withtheNTSBoverpast investigator, aerospaceengineeandmajorairline current position,heservedasaregionalairsafety with theNTSBformorethan11years.Priortohis NTSB’s OfficeofAviationSafety.Guzzettihasbeen Deputy DirectorforRegionalOperationsatthe Jeff Guzzetti Aircraft AccidentInvestigation (AO3317) ([email protected])isthe By JeffGuzzettiandBrianNicklas Army specifications thatwereissuedonDec. 23,1907.Thefol- was destroyedandneverflew again. hold oftheaircraftandtumbled itontotheground.Thecraft the startingpointforanother flight,strongwindgustsgrabbed mous flight,astheairplanewas beingmanually“taxied”backto and successfullyflewitonDec. 17,1903.However,afterthefa- into airfoils,theWrightbrothersconstructedapoweredairplane Influenced bythesuccessesofLilienthalglidersandhisresearch The Wrightbrothersandthefirstfatalaccident an oversighthepaidforwithhislife. he hadneglectedtoinstallthereboundinghooponlaterglider, cited thereboundinghoopdeviceasanimprovement.Ironically, survival, Lilienthalinvestigatedandreportedhisownaccident, and vice thatshatteredandabsorbedtheimpactforces.Because ofhis Lilienthal survivedthecrash,undoubtedlyduetobow-like de- on theiraircraft,thusproducingthe“canard”configuration.) were influencedtoplacethehorizontalstabilizeraheadofwing device. (TheWrightbrothersbecameawareoftheaccidentand flipped over,anddoveintotheground,splinteringbow-like accident. Thescienceofsurvivalfactorswasborn. glider sothatitwouldabsorbimpactforcesintheeventofan stalled acurved,bow-like“reboundinghoop”tothefrontofhis to preventseriousinjuryduringhisflighttesting,Lilienthalin- last words,infact)ifprogressaviationwastobeachieved.Still, ous business,andthat“sacrificesmustbemade”(reportedlyhis Lilienthal wasfirsttorecognizethatflighttestingahazard- device thathehaddesignedtomakethecraftcrashworthy. one ofhisglidersandwasinjuredduetofailureinstalla German mechanicalengineernamedOttoLilienthalcrashedin to rightyourmachinebeforehittingtheground.” not flytooneartheground,forifyoufallwillhavetime Leonardo daVinci,whostated:“Intestingflyingmachines,do ting toputatailonthebackpart.” dent, anddetermined“thatthecauseofhisfailurewasforget- ing hislegs.BrotherEilmersurvivedtoinvestigateownacci- greater than600feetbeforefallinghardontotheearthandbreak- According toahistorian,BrotherEilmerglidedfordistance attaching wingstohishandsandfeetdivingoffthetower. attempted toflyoffofthewesttowerhisabbeyinEuropeby inducing structuralfailurewithsubsequentlossofcontrol.” able causestatingthatIcarus“exceededhiscraft’sthermallimits, mendation againstit,wecanimaginethedeterminationofprob- The Wrightsthenbegantomodify theiraircraftdesigntomeet Later, asLilienthalwastestingoneofhisgliders,thegliderstalled, About 800yearsaftertheEilmermishap,inyear1895,a Accident preventionalsoappearsinthewritingsofone About 1,200yearslater,aBenedictinemonknamedEilmer ISASI 2003 Proceedings • 17 ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS the machine, the longerpropellercaught aguywireonthe ratory ofmakinghisofficialspeed trial.”Duetothevibrationof for thepurposeof“tuningup thespeedofhismachineprepa- the 8-foot8-inchpropellers with onesthatwereninefeetlong report explainedthatpriorto thetrials,OrvilleWrightreplaced which resultedinthemachine fallingtotheground….”The propeller bladeandaconsequent unavoidablelossofcontrol, the accidentwasreportedas“…theaccidentalbreaking ofa ough investigationandwroteasubstantivereport.Thecause of perts whorodetotheaccidentsiteonhorses,conducteda thor- the SignalCorps,composedofArmyofficersandcivilian ex- 6 weeks;hebrokehisleftlegandseveralribsinthecrash. somehow survivedwithseriousinjuriesandwashospitalized for military honorsinnearbyArlingtonCemetery.OrvilleWright aircraft accidentfatality.Selfridge,age26,wasburiedwithfull accident anddiedabout3hourslater;thiswasthefirstofficial about 75feet.Lt.Selfridgesufferedasevereheadinjuryinthe and theairplaneplummetedtogroundfromanaltitudeof times. Suddenly,apieceofonethepropellerbladesbrokeoff, Selfridge onboard,tookoffandcircledthefieldfour-and-a-half pate inthetrials. aviation. HehadbeendesignatedbytheSignalCorpstopartici- dirigible airshipandhadakeeninterestinknowledgeof near LeMans.)SelfridgewasaqualifiedpilotoftheArmy’sfirst present, ashewasinFrancedemonstratingaWrightmachine First FieldArtillerywasonboardtoobserve.(Wilburnot in ordertoflywhileFirstLieutenantThomasE.Selfridgeofthe 1908, thetrialsbegan.OrvilleWrightwasonboardairplane the U.S.ArmyinFt.Myer,Va.,forofficialtrials.OnSept.17, lowing year,theydeliveredtheirproducttotheSignalCorpsof 18 Immediately aftertheaccident,AeronauticalBoard of At 5:14p.m.,theWrightFlyer,withOrvilleandLt. • ISA SI 2003 P r oceedings First fatalaccident. safer thanone.Thebanwaslifted5monthslater. all militaryflightsofmonoplanes,statingthattwowingswere weeks oftheirdeaths,theBritishSecretaryStateforWarbanned Thames RivernearGodstow,Oxfordshire,England.Within3 their Bristol-CoandaMonoplanecrashedontakeoffnexttothe Royal FlyingCorpsofficerstodieinanaircraftaccidentwhen ward HotchkissandLt.C.A.BettingtonbecamethefirstBritish the earlydaysofpoweredflight.OnSept.12,1912,2ndLt.Ed- The deathofLt.Selfridgewasunfortunatelythefirstmanyin World arIandtheRoaring20s rudder toitsside,andtheairplanelostcontrol. aircraft andbrokethepropeller.Theguywirepulledrear stream oflife; there wassimplyverylittlepublic interestinavia- likely becauseaviationwasstill consideredtobeoutsidethemain- gated bytheaircraftdesigners orsurvivingpilots.Thiswasmost of accidentswerenotofficially investigated,orwereonlyinvesti- marking systems,andtheneed forinstrumentflying. sacrifices ledtotheimpetus forcommercialairroutes,aerial and asafelanding.Theaircraftlosseswerestaggering,but the planes duetopoorvisibilitythatpreventedaccuratenavigation Lindbergh, frequentlyhadtobailoutoftheirsingle-engine air- out ofthebusinessairmail.Manypilots,includingCharles investigated bytheaircraftdesignersandsurvivingpilots. craft accidentscontinuedtooccur,andthemajorityofthem were documentation ofallaccidents.IntheUnitedStates,civilian air- man militaryappearstohavebeenthefirstrequirephotographic tion, andalsosignificantnumbersofaircraftaccidents.The Ger- World arIandtheRoaring20s. During thistime,manyofthe airmailcrashesandothertypes Following theendofWorldWarI,commercialaviationevolved World WarIbroughtwithitsignificantdevelopmentsinavia- held thefabric ontothewing.Theairplanes weretemporarily could notbeeasilyinspected), anddeteriorationofthegluethat instability, stressfracturesofthe interiorofthewoodenwing(which covered thattheF-10Ahadpreviousproblems offlightcontrol in thedivetoground.Late r, aftermoreresearch,itwasdis- led toaninflightseparationof awingduringtheexcessiveforces the wings,causingairplane tobecomeuncontrollable,which indicating thatthecauseofaccidentwasacoating ice on the erroneousstatement,anotherpublicstatementwas issued to thecausewasinerror.So,onApril7,4daysafterrelease of recovered, providingproofthattheagency’sfirstspeculation as “no blamecanbeattachedtothepilots.” loose, andstruckthepropellerblade.Thestatementended with, piece oficehadformedononethepropellerhubs,broken broke offandseveredthewing.Italsopubliclyspeculated thata statement thatsaidtheagency“assumed”piecesofpropeller blade the CommerceDepartmentinWashington,D.C.,issuedapublic and threeengines.Threedayslater,theAeronauticsBranchof ment ofCommercearrivedthenextdayonlytofindtwowings Anthony Fokker,andahandfulofinvestigatorsfromtheDepart- cluding thepropellersandfuselage.Theairplane’sdesigner, hunters cartedawayanythingthatcouldbecarriedbyhand,in- The nationwasstunned. endary footballcoachKnuteRockneofNotreDameUniversity. Wichita fromKansasCity.Oneofthepassengerskilledwasleg- continental andWesternAirways(TWA)Flight99Eboundfor from themainimpactsite.TheflightwasbeingoperatedasTrans- There wasnofire.Awingsectionfoundaquartermileaway gers andtwopilotsplummetedtothegroundwasdestroyed. three-engine FokkerF-10Amonoplanecarryingsevenpassen- vestigated. InapasturenearthesmallKansastownofBazaar, changed commercialaviationandthewaythataccidentsarein- March 31,1931,thatawatershedaircraftaccidentoccurred ers reached461,743,upfromamere5,782in1926.Butitwason aviation. In1934,thenumberofpassengersflownbyU.S.carri- The decadeofthe1930sbroughtgreatadvancesincommercial Watershed accidentsofthe1930s thus allowingpilotstotakeoffandlandinfogpoorvisibility. famous AirForcepilotJimmyDoolittleinventedinstrumentflight, gation bodyhadbeenestablished.Additionally,in1929,not-yet- latory powersoveraviation,althoughnoformalaccidentinvesti- Air CommerceAct,whichgavetheDepartmentregu- that areworthnoting.In1926,theU.S.Congressenacted the variousaspectsofanaccident. Accident AnalysisForm”thatservedasachecklistforanalyzing tions ofaccidents,andevenincludedaone-page“NACAAircraft of Analysis.”Thereportdefinedtheterminologyandclassifica- Technical ReportNo.357entitled“AircraftAccidents:Method the CommitteeonAircraftAccidents.Theissued odologies begantocoalesce.OnOct.3,1928,theNACAcreated Aeronautics (NACA)in1915,formalaccidentinvestigationmeth- gate accidents. aviation, otherthanthedesignersandpilots,availabletoinvesti- tion atthattime.Additionally,therewerenoreal“experts”in With thehelpoflocalauthorities,severalpropellerbladeswere No attemptwasmadetosecuretheaccidentsite,andsouvenir There weretwomoreeventsthatoccurredduringthe1920s With thebirthofU.S.NationalAdvisoryCommitteefor Hindenburg checklists. The findingsfromthisaccidentledtotheconceptofflightcrew due tothecrew’sfailureremovegustlockspriortakeoff. The U.S.Armyinvestigationrevealedthattheaccidentoccurred takeoff, destroyingtheairplanebyfireandkillingallaboard. bomber prototype.OnOct.30,1935,theprototypecrashedon Field inDayton,Ohio.ThemilitarywastestingaBoeingB-17 warts astheDouglasDC-3. develop all-metalcommercialairplanes,givingrisetosuchstal- investigation alsoprovidedimpetusfortheaircraftindustryto ity togroundafleetofairplanesinthewakeanaccident.The NACA “MethodofAnalysis”foraircraftaccidents;andtheabil- improved documentationofbodyrecovery;acceptedusethe public speculationofcauseimmediatelyfollowingtheaccident; accident sceneanditswreckage;prohibitionsregardingofficial methodology, suchasenhancedauthoritytotakechargeofan the ganglandexecutionofJakeLingleinaChicagosubway. ploded abouttheairplaneinordertokillapriestwhowitnessed tion abouttheaccident,suchastheorythatatimebombex- investigation andlackofsolidevidence the causeofKnuteRocknecrashwithconfidence.Thepoor grounded asaresultofthisresearch. Pacific Oceanduringtheirattempttoflyaroundtheworld. Earhart andhernavigator,FredNoonan,somewhereoverthe what similarprobablecause. tic Oceanwouldoccurnearly60yearslater,resultinginasome- discharge ignitedsuchmixtureappearsmostprobable.” distance forwardoftheupperverticalfin.Thetheorythatabrush an openflamewasonthetopofshipandarelativelyshort part oftheshipinconsiderablequantity;firstappearance bustible mixtureofhydrogenandairtoformintheupperstern dence, aleakatorinthevicinityofcells4and5causedcom- tion ofamixturefreehydrogenandair.Basedupontheevi- report concludedthatthecauseofaccidentwas“…theigni- vestigation andissuedafinalreport3monthslater.The18-page time, theBureauofAirCommerceconductedanexhaustivein- more wereseriouslyinjured.Again,thenationwasstunned.This crewmembers and13passengersdiedintheaccident,many 1930s B-17prototypecrash. About 2yearslater,onMay6,1937,theGermanairship Another watershedaccidentoccurred4yearslateratWright The accidentledtoseveraladvancesininvestigation To thisday,thereseemstobeinadequateevidencedetermine The decadeofthe30sendedwithmysteriouslossAmelia Another accidentinvolvingaTWABoeing747overtheAtlan- exploded intoflamesandwasdestroyed.Twenty-two

spurred mediaspecula- ISASI 2003 Proceedings • 19 ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS began coming intotheRAEatFarnboroughfor awreckagere- died. Anextensivewreckage-recovery effortbegan,andpieces feet fromRometoCairo.Acrew ofsevenandall14passengers another Cometmysteriouslybroke apartwhilecruisingat35,000 lowed toflyagain. new evidencefromthesunken wreckage,theCometswereal- 2 months,withnofindingsyetfromthefuselagetesting,and no and begantestingitsfuselageformetalfatigue.Aftermore than gine failure,andonboardfire.TheRAEprocuredanotherComet evidence, therecamespeculationofabomb,uncontained en- of theBritishAirAccidentsInvestigationBranch).Withalack of Establishment (RAE)inFarnborough,England(thecurrentsite be found.AninvestigationwasconductedbytheRoyalAircraft Comets wereimmediatelygroundeduntilsomeanswers could was strewnalongthebottomofMediterraneanSea.All ofthe killing allsixcrewmembersand29passengers.Thewreckage cruise flightabout24,000feetonafromRometoLondon, of health.Then,onJan.10,1954,anotherCometbrokeapartin encounter withsevereweather,sotheCometreceivedacleanbill The officialcausereportedbytheIndiangovernmentwasan parture fromCalcutta,India,whileflyinginthevicinityofastorm. ler-driven transports. aircraft manufacturerswerecontinuingtobuildslowerpropel- to cornerthemarketoncommercialaircraftsales,sinceallother pressurized cabinforflightathighaltitudes.TheCometbegan The Cometwaspoweredbyfourjetenginesandincorporateda tudes approaching50,000feetatspeedsof500milesperhour. transport aircraftthatcouldcarry36passengersandflyatalti- design withtheintroductionofdeHavillandComet,asleek aircraft, whichusheredinaneweraaviation. Spitfire fighter.TheGermanswerethefirsttobuildoperationaljet rescinded ports andbombers,theneedforspeed.TheBritish,having bringing withitrequirementsforlargefour-enginemilitarytrans- CAB toinvestigateaircraftaccidents.Then,WorldWarIIbegan, Aeronautics Actof1938.ABureauSafetywascreatedunderthe nautics Board(CAB)wascreatedunderthemandateofCivil from theCivilAeronauticsBranchin1933.In1940,Aero- In theUnitedStates,BureauofAirCommercewasformed World arIIandthe1950s 20 Two weeksafterthegrounding waslifted,inearlyApril1954, On May2,1953,aCometbrokeapartimmediatelyafterde- In theearly1950s,Britishledworldincommercialjet • ISA SI 2003

its banonmonoplanes,developedthehighlysuccessful P r oceedings 1930s 1950s deHavilland Hindenberg Comet wreckage. crash. military andairlinepilotsmechanics.Theirvastoperational investigators weretypicallyallmenwhomostlyhigh-time ing aviationcombatpilotsfromWorldWarIIandKorea. These investigate U.S.civilaccidentswithacadreofmostlytough-talk- In theearly1960s,CAB’sBureauofAirSafetycontinuedto The 1960s:adecadeofchanges investigation…the flightrecorder. ably themostimportantadvanceinaviationaccident a classiccaseofweather-relatedaccident. off intoastormywinternightandcrashednearClearLake,Iowa, 3, 1959,knownas“thedaythemusicdied.”Theairplanetook and RitchieValensperishedinaBeechBonanzaaccidentonFeb. investigation. within theCABwouldremaininchargeofcivilaircraftaccident facilities andairtrafficcontrol;however,theBureauofAirSafety of thenavigableairspaceanddevelopoperateairnavigation duced tocreatetheFederalAviationAgency(FAA)controluse bear toincreaseaviationsafety.OnMay21,1958,abillwasintro- ensure safeseparationofcommercialairlines. air trafficcontrolsystematthetimewasnotadequateenoughto came together.Additionalevidenceandresearchrevealedthatthe off. Thewreckagesignaturespaintedapictureofhowtheairplanes slopes ofTempleButteintheGrandCanyon.Thehardworkpaid of wreckagefrombothairlinersthatwerestrewnalongthesteep many gruelingweeksrecovering,identifying,andanalyzingpieces their lives.InvestigatorswiththeCAB’sBureauofAirSafetyspent 21,000 feetabovetheGrandCanyon.Atotalof128peoplelost collided inflightwithaTWALockheedSuperConstellationabout years lateronJune30,1956.Onthatday,aUnitedAirlinesDC-7 with othermanufacturersofnewjettransports. grounded, andthenewCometsbegancompetingunsuccessfully ture wasrecognized.Unfortunately,theolderCometsremained cident wassolved,andtheneedforfatigue-tolerantaircraftstruc- accident Cometandrevealedacrackinthesameplace.Theac- open thefuselage.Asimilarpiecewasthenrecoveredfrom edge ofanantennaholebegantoformandeventuallycracked Comet suddenlyrevealedaflaw.Astresscrackfromthesharp been found.Meanwhile,theongoingpressuretestingofother blown apart,butnoinitiatingsiteforthestructuralfailurehad construction. Thereconstructionrevealedthatthefuselagehad The endofthe1950sbroughtaboutbeginningargu- Rock andRollHallofFamersBuddyHolly,theBigBopper, As aresultofthishorrificaccident,pressurewasbroughtto Another watershedaccidentoccurredintheUnitedStatesafew military base.When theywereoncall,had nopagersorcell doubled uponhotelrooms orbunkedinthelocalYMCA ceived verylittleperdiem on accidentlaunches.Theyoften that werebarelyabletosupport theirfamilies.Investigatorsre- and administrativesupport. on theU.S.DepartmentofT ransportation (DOT)forfunding April 1,1967.Althoughitwasanindependentagency, relied Transportation SafetyBoard.TheNTSBopeneditsdoors on Washington, D.C.,asemployeesofthenewlycreatedNational tors. Aftercompletingtheon-sceneportion,theycameback to peers launchedontheaccident,theydidsoasCABinvestiga- esting noteaboutthisaccidentwasthatwhenRodriguezand his conducted insimulatorsinsteadofanactualairplane.One inter- the FAArequiredthatairlinetrainingforsuchemergencies be plane, alongwith13peopleontheground.Afteraccident, lated. FivecrewmembersandoneFAAinspectordiedonthe air- during whichadual-enginefailureonthesamesidewassimu- into ahotelnearNewOrleans,La.,whileontrainingflight shed accident.OnMarch30,1967,aDeltaAirLinesDC-8crashed ing descent. crew’s failuretocomplywithvectorsfromairtrafficcontroldur- cause oftheaccidentwasdeterminedtobeinexperienced foot-tall SalmonMountainnearMountHoodinOregon.The West CoastAirlinesflightsmackedtheeasternslopeof4,090- with a“GoTeam”)thefirstfatalaccidentinvolvingDC-9.The dent ofISASI. who camealonglaterwasRonSchleede,thecurrentvice-presi- and nowtheNTSB,formorethan40years.Anotherinvestigator Rodriguez, whohasbeeninvestigatingaccidentsfortheCAB, had. OneoftheseinvestigatorswasamannamedRichard experience outweighedanyformaleducationthattheymayhave 1970s Tenerifegroundcollision. During thesetimes,airsafety investigatorswerepaidsalaries Rodriguez wasalsoinvolvedinwhatwouldbeanotherwater- On Oct.1,1966,Rodriguezwasassignedtoinvestigate(along air carrieraircraft beequippedwithground proximitywarning applied. Asaresultoftheinvestigation, theFAArequiredthatall altitude priortotheapproach segmentinwhichthataltitude and thatthecrewimproperly decidedtodescendaminimum equacies andalackofclarity inairtrafficcontrol’sprocedures, proach. Theinvestigationlater determinedthattherewere Washington’s DullesInternational Airportonaninstrumentap- Dec. 1,1974,TWAFlight514wasdescendingtoland at Berryville, Va.,initiatinganothersignificantsafetychange.On directly toCongress. NTSB nolongerhadany“sister”agenciesandnowreported between theSafetyBoardandDOT,includingFAA.The Act of1974,whichseveredallorganizationalandfinancial ties controversies, CongresspassedtheIndependentSafetyBoard partial initsconclusionsandrecommendations.Fueledby these reliance ontheDOT,NTSBmightnotbeobjectiveandim- Nixon Administration,anditwassuggestedthatbecauseofits Paris, killingall346peopleonboard. the airplaneslammedintoErmenonvilleForestnortheastof This time,however,alloftheflightcontrolsweresevered,and enced theexactsamescenariowhileclimbingoutfromParis. years later,onMarch3,1974,aTurkishAirlinesDC-10experi- FAA didnotrequireDouglastocorrectthedoordesign.Two the DC-10cargodoortoitssisteragency,FAA.However, made severalrecommendationsabouttheinadequatedesignof one washurt,theNTSBinitiatedathoroughinvestigationand ment, whichcausedintensemediaattention).Eventhoughno Detroit (althoughacasketwasblownoutofthecargocompart- was fortunatelyabletonursethecrippledairplanesafelyback floor, causingseveralflightcontrolcablestobesevered.Thecrew and thethinair.Theexplosivedecompressionbuckledcabin out duetoanexcessivepressuredifferentialbetweenthecabin troit, apartiallysecuredaftbaggagecargodoorsuddenlyblew During climboutoverWindsor,Ontario,fromitstakeoffinDe- no onerealizedthelastingeffectthatthisaccidentwouldhave. American AirlinesDC-10accidentonJune11,1972.Atthetime, The beginningofthe1970sbeganwithinvestigationan The 1970s:thedeadliestdecade,andtrueindependence History continuedtorepeatitselfwithweather-relatedaccidents. both wereclassicweather-relatedaccidentsinsmallairplanes. a PiperComanchewithverylow-timeprivatepilotonboard); singer OtisReddingandcountryPatsyCline(whodiedin in preventingmoreaccidentsofthesamenature. tors quicklysolvemajorairlineaccidentsand,therefore,assisted eters. Crudeornot,theflightrecordershelpedNTSBinvestiga- solid-state electronicsthatcanstoredataforhundredsofparam- five parameterswrittenontoaribbonoffoil)comparedtotoday’s flight datarecorders.Theserecordersweresomewhatcrude(only nearly aseasyitistoday. them anyway).Travelingacrosscountryinthosedayswasnot very littlenewsmedia(bydesign,therewasn’tmuchroomfor ings wereheldinsmallconferenceroomswithafewtablesand rotary-dial phonestoringwithnewsofanaccident.Publichear- phones; theyhadtowaitathomeduringtheirdutyfor That sameyear,aBoeing727slammedintohillnear At thetime,politicalscandalswereeruptingthroughout The 1960salsobroughtwithittheuntimelydeathsofrock Also duringthisdecade,theFAArequiredairlinestoinstall ISASI 2003 Proceedings

inad- • 21 ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS Bridge andlanded inthepartiallyfrozen Potomac River.Sev- Airport. About1minutelater, itimpactedtheFourteenthStreet Florida Flight90beganitstakeoff rollfromWashingtonNational critics echoedthroughoutthe country.Onthatafternoon,Air and decreasetheamountoftraining thattheirpilotsreceived. tenance parties,hireless-experienced pilotsforlowersalaries, there werefearsthatsomeairlineswouldcontractouttheir main- “cut corners”inordertobecomemorecompetitive.Forexample, ics. Somecriticsdeclaredthatsomeairlineswouldbetempted to airlines begantooperatewithakeeneyeonprofitsandeconom- government determineroutingandticketprices.Asaresult, the pete forroutesagainsteachother,ratherthanhavethefederal Line DeregulationActof1978.Theallowedairlinesto com- The decadeofthe80sbeganwithrippleeffects Air and thespaceshuttle The 1980s:deregulation,aircraftaging, and baseballgreatRobertoClemente. hero andactorAudieMurphy,oceanographerPhillipeCousteau, singer JimCroce,U2pilotFrancisGaryPowers,WorldWarII Thurman Munson,HouseMajorityleaderHaleBoggs,rock aircraft accidentsalsokilledall-starNewYorkMetscatcher grounded allDC-10sforoveramonthasresult. wing, aretractionoftheslats,andanaerodynamicstall.TheFAA maintenance, whichledtotheleftengineseparatingfromits ure oftheleftengine’ssupportpylonasaresultimproper 1979. All273peoplewerekilled.Theaccidentwasduetoafail- right aftertakeofffromChicago’sO’HareAirportonMay25, history, whenanAmericanAirlinesDC-10(Flight191)crashed being taughttoworkasateaminthecockpit. CRM. Experiencedcaptainsandtheirlessercrewmemberswere the impetusforconceptofcrewresourcemanagement,or was runninglowonfuel.Thisaccident,alongwithTenerife, of assertivenessbytheflightengineer,whoknewairplane one flightattendantwerekilled.Theinvestigationrevealedalack crash sixmilessoutheastofPortland,Ore.Eightpassengersand DC-8, Flight173,allowedtheairplanetorunoutoffueland deadliest aviationdisasterinworldhistory. Pan Am747.Thefinaldeathtollwas583.Itstandstodayasthe lives. Only61passengerssurvivedoutofthe396peopleon ing ahorrificfireball.All248onboardtheKLM747losttheir plane, duringitstakeoffroll,slammedintothePanAm747,caus- was back-taxiingontheactiverunwayfortakeoff.TheKLMair- World Airways747,whoweretalkingtoadifferentcontroller, while, ontheotherendofrunway,crewaPanAm traffic controllersandthefailureofhiscopilottospeakup.Mean- clearance totakeoff,dueinpartlanguagedifficultieswithair enced captainofaKLMBoeing747thoughthehadbeengiven 1977. Aftertaxiingonafog-shroudedtaxiway,theveryexperi- Airport ontheSpanishresortislandofTenerifeMarch27, decade, anotherlandmarkaccidentoccurredattheLosRodeas ing controlledflightintoterrain(CFIT). ments inthepreventionofall-too-frequentaccidentsinvolv- systems (GPWS),arguablyoneofthemostsignificantdevelop- 22 On thefrigidafternoonofJan. 13,1982,thevoicesofthese In additiontothesemajorairlinedisasters,intheseventies The decadeendedwiththedeadliestaircraftaccidentinU.S. One yearlater,onDec.28,1978,thecrewofaUnitedAirlines As iftheseincrediblytragicaccidentsweren’tenoughforone • ISA SI 2003 P r oceedings Ironically, 17years later,anotherspaceshuttle, alongwithanother Commission citedareduction infundingandemphasisonsafety. but didnottakeadequateaction tocorrect.Additionally,theRogers agement failurestoaddressproblems thatithadbeenawareof, derlying factorsthatwerecited, however,addressedNASAman- failure ofano-ringinajoint therightsolidrocketbooster.Un- and cited,amongotheritems, thattheaccidentwascausedbya 5 monthsaftertheaccident,Commissionsubmittedits report the independentRoger’sCommissiononaccident.Less than investigate itselfwithpubliccredibility,PresidentReaganformed and thepublicdidn’tlikeit.RealizingthatNASAcouldn’t really granted. Then,the 5 years,andNASAthepublicseemedtotakespacetravel for McAulliffe. Theshuttleprogramhadbeenrunningsmoothly for astronauts, includingthefirst“teacherinspace”Christa ing inlateJanuary1986.Thenationmournedthedeathsof seven exploded lessthanaminuteafterlaunchoncoldFloridamorn- significant accidentofthe1980s.Thespaceshuttle their airplaneswereinitiatedandarestillinforcetoday. investigations, revolutionarychangesinthewayairlinesdeice Group Chairmanonalloftheseaccidents.)Asaresult (Veteran investigatorRichardRodriguezservedastheOperations Flight 405,aFokkerF-28,atLaGuardiaAirportinNewYork. tinental DC-9inDenver,aRyanAirCleveland,andUSAir dents includedanAirborneExpressDC-9inPhiladelphia,aCon- dents inthe1980sinvolvingice-contaminatedwings.Theseacci- ber ofISASI,wastheinvestigator-in-chargethisaccident.) transport winteroperations.”(RudyKapustin,long-timemem- or ice,and(3)thelimitedexperienceofflightcrewinjet leading edgeiscontaminatedwithevensmallamountsofsnow herent pitch-upcharacteristicsoftheB-737aircraftwhen plane wasexposedtocontinualprecipitation,(2)theknownin- of anairtrafficcontroltakeoffclearanceduringwhichtheair- (1) theprolongedgrounddelaybetweende-icingandreceipt engine instrumentreadings.Contributingtotheaccidentwere ing theearlystagewhenhisattentionwascalledtoanomalous of theaircraft,andcaptain’sfailuretorejecttakeoffdur- off, theirdecisiontotakeoffwithsnow/iceontheairfoilsurfaces failure touseengineanti-iceduringgroundoperationandtake- that “...theprobablecauseofthisaccidentwastheflightcrew’s page reportreleasedonAug.10,1982,theNTSBdetermined enty-four peoplelosttheirlives.Onlyfivesurvived.Ina141- 1980s AlohaAirlines737. After theaccident,NASAinitiallysharedverylittlewithpublic, As collegestudents,theseauthorswereeyewitnessestoanother Air Floridamarkedthebeginningofastringtakeoffacci- Challenger horror shockedthenation. Challenger This secondaccident furtherunderscoredthe needtoreviewair- search thecomplexitiesofaerodynamic penaltiesduetoicing. piloted anairplane,buthad the educationalbackgroundtore- younger scientistsattheNTSB, andalsoatNASA,whohadnever Detroit. RichardRodriguez, astheIIC,neededhelpof the samereasonsasRoselawn ATR72whileonapproachto boprop, operatingasComair Flight3272,crashedforsomeof their crude“foil”andmagnetictapepredecessors recorders wereprovingtobeexponentiallymorevaluable than solved withindaysoftheaccident.Newersolid-stateflight data contrast tothe737accidents,RoselawnATR72crash was something thatthe737rudderaccidentsdidnotpossess.In stark not forthenewexpandedparameterflightrecorderonboard, accident wouldhavebeenextremelydifficulttosolveifit were effects thatresultedfromicecontaminationonthewings. The ron hingemomentreversalduetoaerodynamicandautopilot tions prevailed,andtheairplanesuddenlyexperiencedanaile- Ind., whileinaholdingpatterntolandChicago.Icingcondi- operating asAmericanEagleFlight4184 cial aviationsafety:Icingresearch.OnOct,31,1994,anATR72, haustive researchandtesting. more specialistswithformaleducation,andtimeforex- such asthesewereextremelycomplexandrequiredmore age retrofitofthemostpopularjetlinerinworld.Accidents flight datarecorders,theinvestigationsgaverisetoarudderpack- periods oftimeduetothelackinformationfromolder both involvedcatastrophicmovementsoftherudder.Afterlengthy near Pittsburgh,Pa.BothaccidentsinvolvedtheBoeing737,and Flight 535inColoradoSprings,Colo.,andUSAir427 the mostnotableandcontroversialofthesewasUnitedAirlines The 1990sbeganwitharashofcommercialairlineaccidents,but The 1990s:Icing,rudders,JFK,Jr.,andTWA800 volving anelectricalfire. rock singerRickyNelson,whoperishedinaDC-3accidentin- struck himduringthefilmingof cade includedactorVicMorrow,whodiedwhenahelicopterblade changed thewayairplanesandenginesarecertified. engine failureandlossofcontrolinSiouxCity,Iowa,which ance system(TCAS),andtheUnitedAirlinesDC-10uncontained which providedtheimpetusfortrafficalertandcollisionavoid- collision ofaDC-9withsmallairplaneoverCerritos,Calif., damage andfatigue. quired frequentdetailedinspectionsofairplanesforcorrosion a vastprogram,knowastheAircraftAgingProgram,thatre- failure ofafuselageskinlapjoint.Theinvestigationgaveriseto convened. Thecauseinvolvedfatiguedamagethatledtothe killed. Theairplanelandedsafelyandanintenseinvestigation decompression. Miraculously,shewastheonlypersonwho at 19,000feet.Aflightattendantwasejectedduringtherapid lines, suddenlyhadthetopofitsfuselageripawaywhilecruising when an“islandhopping”Boeing737,operatedbyAlohaAir- concerns. Aviation,includingspacetravel,isunforgiving. crew ofsevenastronauts,wouldperishduetosomethesame A fewyearslater,onJan.9,1997,anEmbraerEMB-120 tur- The 90salsousheredinanotherimportantaspectcommer- Celebrities whoperishedinaircraftaccidentsduringthisde- Other notableaccidentsintheeightiesincludedmid-air Yet anotherwatershedaccidentoccurredonApril28,1988, The TwilightZone , crashedoverRoselawn, movie,and the continuing trendtowardconductingextensive scientificre- cade ofaccidentinvestigation.) Theinvestigationagainshowcased investigation wasledbyRichard Rodriguez;itwashisfifthde- Airlines Flight261intothePacific OceaninJanuary2000.(The The dawnofthenewcenturybegan withthefatalplungeofAlaska A newcentury sica Dubroff. the bandofcountrysingerRebaMcEntire,and7-year-old Jes- Sen. JohnTower,Heinz,NASCARdriverDaveyAllison, Denver, golferPayneStewart,bluesguitaristStevieRayVaughan, mercial aircraftaccidentsduringthe90sincludedpopsinger John spiracy theoriesaboutmurderabounded. a VFR-ratedpilotwhobecamespatiallydisoriented.Again, con- the finalprobablecausereflectedyetanotheraccidentinvolving ness aspectsoftheinvestigation,andnonewerefound.Instead, Vineyard, Mass.Oneoftheauthorswasinvolvedinairworthi- wife, andsister-in-law,inaPiperSaratogaoffthecoastofMartha’s of aviation. their accidentcausation,andthehighfinancialpoliticalstakes suits), thecomplexitiesoftechnologicallyadvancedaircraftand more andprooftodefendagainstothertheorieslaw the proliferationoflitigationandnewsmedia(i.e.,needfor phenomenon oflargerandreportscanbeattributedto years afterthe1996accidentandwas425pagesinlength.The ing appendices.TheTWAFlight800finalreportcameout4 report inlessthan7months,andwas98pageslength,includ- crash atChicago’sO’HareAirportin1979gaverisetoafinal of DC-6aircraftacrosstheworld.TheAmericanAirlinesDC-10 six, andinvolvedamechanicalmalfunctionthataffectedthefleet the accident.Theaccidentkilledall46passengersandacrewof Oct. 26,1947,was port ofaUnitedAirlinesDC-6crashinBryceCanyon,Utah,on was thelengthofreports.Forexample,entirefinalCABre- seems, aretimelesstraitsofthehumanspecies. tenacious todayastheyweredecadesago.Distrustandcynicism,it and KnuteRocknecase,provingthatconspiracytheoriesareas niscent ofthetheoriesthataboundedattime believe thattheairplanewasdownedbyamissileorbomb,remi- with thefuelquantityindicationsystem.”Tothisday,somepeople excessive voltagetoenteritthroughelectricalwiringassociated most likelywasashortcircuitoutsideoftheCWTthatallowed certainty, but,ofthesourcesevaluatedbyinvestigation, of ignitionenergyfortheexplosioncouldnotbedeterminedwith ignition oftheflammablefuel/airmixtureintank.Thesource “an explosionofthecenterwingfueltank(CWT),resultingfrom cluded thatprobablecauseoftheTWAFlight800accidentwas nited byanunknownignitionsource.Specifically,theBoardcon- 60 yearsprevious…thatavolatilemixtureoffuelandairwasig- cause oftheaccidentwassimilartothat dent investigationinU.S.history,theNTSBdeterminedthat JFK AirportinJuly1996.Afterarguablythemostarduousacci- that explodedwhileclimbingoutaftertakeofffromNewYork’s be completewithoutmentioningTWAFlight800,aBoeing747 craft icingcertification. Other celebritieswhoperishedingeneralaviationandcom- The decadeendedwiththedeathofJohnF.Kennedy,Jr.,his Another interestingchangeinaircraftaccidentinvestigation Of course,nohistoricalperspectiveofaircraftaccidentswould four pageslong andwasadopted4monthsafter ISASI 2003 Hindenburg Proceedings Hindenburg disaster • 23 ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS to havebeenserving themwell.Ratherthan firstbreakingup orthodox accidentinvestigation methodologies,buttheyappear a shortperiodoftime.TheCAIB membersusedsomewhatun- leased thismonth. gation, andotherdisciplines Theirreportisexpectedtobere- in physics,spacesciences,military andcivilianaccidentinvesti- Retired AdmiralHaroldW.Gehman,Jr.,andincludes14experts Columbia formed, similartotheRogersCommission,andwascalled the regarding theaccident.Anotherindependentcommission was mediately wasforthrightandcandidindivulginginformation ing thepublicresponsefrom Space Center.Sevenmoreastronautslosttheirlives.Remember- disintegrated intheatmosphereduringitsreturntoKennedy disaster, onFebruary1ofthisyear,thespaceshuttle successful missions.Butsuddenly,17yearsafterthe under theradarofpublicscrutinyduetonearlyahundredmore that theU.S.spaceprogramwassafelyandquietlyprogressing After thespaceshuttle The becoming partiallyseparatedfromtheaircraft. meted tothegroundasaresultofitsverticalstabilizer/rudder involved a“composite”airplane(theWrightFlyer),whichplum- that thefirstfatalityinU.S.history,ofLt.Selfridge,also The accidentisstillunderinvestigation.Itinterestingtonote ite materialsinaviationcomponentssuddenlytookthespotlight. airplane andfiveonthegroundwerekilled.Theuseofcompos- tical stabilizerseparatedinflight.Atotalof260personsonthe bus A300,crashedintoaresidentialareaafteritscompositever- New York’sJFKAirport.AmericanAirlinesFlight587,andAir- deadliest inU.S.historyfact,occurredrightaftertakeofffrom accident, amaintenancedeficiencyhadreareditsuglyhead. as wasthecaseinAlohaAirlinesandChicagoO’HareDC-10 airline’s inadequatelubricationduringmaintenance.Onceagain, actuating mechanismasaresultofexcessivewearcausedbythe cause oftheaccidentwasfailurehorizontalstabilizer’s of lubricatinggrease.Intheend,SafetyBoardruledthat Flight 261,onetopicthatwasstudiedextensivelytheeffect search programsaspartofaccidentinvestigations.Inthecase 24 A newcenturyAmericanAirlines587. One oftheauthorsparticipated asanadvisortotheCAIBfor In Novemberof2001,anotherdeadlyaccident,thesecond • ISA Columbia SI 2003 Accident InvestigationBoard(CAIB).Itisbeingledby P r oceedings spaceshuttle Challenger Challenger accident of1986,itseemed accident,NASAim- Challenger Columbia cency andignoranceofsafetycontinuetobeunforgiving. Inde- deadly whenmisunderstoodormisused.Managementcompla- dents. Newtechnologicaladvancesinaviationcontinuetoprove pilots continuetodieinthesametypesofweather-relatedacci- to speculateinthewakeofacrashinvestigation.Generalaviation over thepast100years.Thenewsmediaandpubliccontinue memory chips.Thetypewriterhasgivenwaytothelaptop. gage” cockpitindicatorshavebeenreplacedbynon-volatile placed bydigitalphotography.Needleimpressionsfrom“steam (GPS) receivers.Thenegativesofa35mmfilmhavebeenre- are beingsupersededbyhand-heldglobalpositioningsatellite software forreconstruction.Topographicalmapsandcompasses mentation ofwreckageisbeingovertakenby3-Dscanningand ties havegivenwaytobiohazardsuits.Pencilandpaperdocu- replaced bytheInternetanddigitalstorage.Woolcoatsblack satellite cellphonesandtextpagers.Paperarchiveshavebeen tron microscope.Rotary-dialtelephoneshavebeenreplacedby gation. Themagnifyingglasshasgivenwaytothescanningelec- changed inthemethodologiesandhardwareofaccidentinvesti- After ahundredyearsofaviationaccidents,manythingshave The morethingschange… lenger ing factorswereverysimilartotheRogersCommissionon emphasis onsafetyinNASA’sshuttleprogram.Theseunderly- tionally, CAIBmembershavecitedareductioninfundingand debris, butdidnottakeadequateactiontocorrectthem.Addi- had beenawareofthepotentialhazardsexternaltankfoam tegration duringreentry. heat topenetratethe tank duringlaunch.Thedamagedleadingedgeallowedextreme pact withapieceoffoamthatcameofftheshuttle’sexternalfuel the damageof they declaredthatthemostprobableworkingscenarioinvolved public appearancestoexpoundontheirthoughts.Intheend, narios toproveordisproveeach,allthewhilemakingfrequent Board thenattemptedtofitthedevelopingfactsintosce- tailed briefingsgiventothembyNASAandtheirownstaff.The ining wreckage,interviewingkeypersons,andlisteningtode- several “workingscenarios”byreviewingtelemetrydata,exam- reports foravarietyofdisciplines,theBoardbasicallydeveloped into numerousgroupsandgeneratingfieldnotes/groupfactual Space shuttlereconstruction layout. By thesametoken,thingshaveremainedverymuch The CAIBwasalsostunnedtolearnthatNASAmanagement . Historyrepeateditself. Columbia’s Columbia’s leftwing’sleadingedgeduetoim- structureandinitiateitsdisin- Chal- Crouch, TomD., Columbia Civil AeronauticsBoard.“AccidentInvestigationReport:UnitedAirLines, Bureau ofAirCommerce,“ReportAirship‘Hindenburg’AccidentInves- Adair, Bill. References knowledge andrecommendationstopreventfutureaccidents. it hasforthelast100years,anoblemissiontosavelivesthrough will tomakeaviationsafer.Accidentinvestigationmustremain,as for mechanicswillhelpamelioratethecurrentaccidentrate. analysis offlightparameters,andincreasedtrainingstature cockpit videorecorders,enhancedGPWS,real-timedownloadand population. Thisisachallengetous.Perhapsadvancessuchas one airlinehulllossperweekduetotheproliferationofflying airline accidentrate,then,inafewyears,theworldwillexperience prove. Ithasbeensaidthatifweremainsatisfiedwiththecurrent history andnotbe“doomedtorepeatit.”Wemustconstantlyim- vestigation. safety investigator’sdedicationtopreventaccidentsthroughin- mained thesamedespiteallofchangesinaviationisair lives throughlessonslearned.Mostimportantly,whathasre- pendent, objective,andthoroughinvestigationscontinuetosave Norton: NewYork,1989. Accident InvestigationBoardReleasesWorkingScenario.”May6,2003. No. 1-0097-47.February3,1948. Inc., BryceCanyon,Utah,October,1947”CivilAeronauticsBoard.File 1937, Washington,D.C. tigation.” BureauofAirCommerce,CommerceBulletin,August15, Institution Press:Washington,D.C.,2002. In theend,however,itwillbeairsafetyinvestigator’siron Where dowegofromhere?Ifaresmart,willlearn Accident InvestigationBoardPressReleasePA27-03.“ The MysteryofFlight427:InsideaCrashInvestigation The Bishop’sBoys:ALifeofWilburandOrvilleWright , Smithsonian Columbia , W.W.

N Lahm, FrankP.,1st Jakab, PeterL.andYoung,Rick,(eds.) Hallion, RichardP., Wald, MathewL.andSchwartz,John.“BoardCriticizesShuttleOversight,” Strauch, Barry,andFrankRiche.“TrainingFacilitywillPrepareSafetyEx- “Rockne DiedGangsters’Victims.” “Report ofthePresidentialCommissiononSpaceShuttle Quinn, John.“JFKJr.’sFinalFlightBlastedoutoftheSky,”www. Owen, David. Oliver, David. National TransportationSafetyBoardAircraftAccidentReport.“TransWorld National AdvisoryCommitteeforAeronautics,“AircraftAccidents:Method Nance, JohnJ., Mallan, Lloyd, Hallion, RichardP. Hallion, RichardP. Guzzetti, Jeffrey,InterviewwithRichardRodriguez.NationalTransportation Eckert, WilliamG.,MD.“TheRockneCrash:AmericanCommercialAirCrash Crouch, TomD.,Jakab,PeterL., Orville Wright stitution Press:Washington,D.C.,1978. the tional CivilAviationOrganization.Number4,2003. perts forIncreasinglyComplexInvestigations,”JournaloftheInterna- cident,” June6,1986. freeworldalliance.com 1999. Kingdom. 1998. Ltd.: Shepperton,Surrey,1990. 1974.” ReportNumberAAR-75-16. Airlines, Inc.Boeing727-231,N54328,Berryville,Virginia,December1, dents, NACA,Washington,D.C.,January28,1930. of Analysis”ReportNo.357.PreparedbytheCommitteeonAircraftAcci- New York,1986. necticut. BookNo.517,1962. the WrightAeroplaneatFt.Myer,Virginia,onSeptember17,1908.”1908. Safety BoardSeniorAirInvestigator.Washington,D.C.,June4,2003. Pathology Investigation intheEarlyYears,” Aerial Age New YorkTimes, , Volume3,Number1,March1982. , NationalGeographic:Washington,D.C.,2003. Air AccidentInvestigation British MilitaryAircraftAccidents:TheLast25Years, Great AirDisasters , SmithsonianInstitutionPress:Washington,D.C.,2002. Blind Trust

Legacy ofFlight Test Pilots LT, U.S.ArmySignalCorps.“ReportoftheAccidentto The WrightBrothers:HeirsofPrometheus July 12,2003. . WilliamMorrowandCompany,Inc.:NewYork, , UniversityofWashingtonPress,1981. , FawcettPublications,Inc.:Greenwich,Con- The WrightBrothersandtheInventionof , UniversityofWashingtonPress,1977. Washington Times, American JournalofForensicMedicineand , HaynesPublishing:Somerset,United The PublishedWritingsofWilburand ISASI 2003 January 6,1933. , SmithsonianIn- Proceedings Challenger IanAllen • Ac- 25 ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS T straints andthat arewellmatchedtohuman cognitiveprocesses face ofreal-worldambiguities, workloaddemands,andtimecon- dously challengingtodesign proceduresthatarerobustinthe cedures foremergencyand abnormal situations.Itistremen- guidelines fordesigning,validating, certifying,andtrainingpro- tions, crewshaveinfrequentopportunity topracticethem. executed correctlyandefficientlywhenneededinlineopera- age thesituationasawhole.Althoughtheseproceduresmust be functioning aircraftsystemsratherthanguidingcrewsto man- to complete,andmanyproceduresfocusonresponding mal- pressure, andstress.Someproceduresareconfusingordifficult that arefragileunderthecombinationofhighworkload, time gency andabnormalproceduresdependsoncognitiveprocesses ordination insideandoutsideoftheairplane.Executing emer- workload conditionsthatrequireexceptionallyhighlevels ofco- ambiguous. Almostbydefinitiontheyinvolvehigh-stressand high- plex, andthenatureofunderlyingproblemissometimes in aircarrieroperations.Theyareoftentimecriticalandcom- Emergency andabnormalsituationsrepresentuniquechallenges Introduction andoverview that areevidentinrecentaviationaccidentsprovided. Center. Research andTechnologyDivisionoftheNASAAmes to completeherinstrumentrating“anydaynow.” full-time inAugust2000.Sheisacertificatedprivatepilotandplans Traumatic StressDisorder,toworkattheNASAAmesResearchCenter academia and10yearsofclinicalpractice,withaspecialtyinPost NASA/ASEE FacultyFellowshipsin1999and2000.Sheleft respectively. ShewasawardedStanford/SanJoseStateUniversity/ University ofFloridaandtheMiamiMedicalCenter, completed apredoctoralinternshipandpost-doctoralresidencyatthe psychology fromSouthernIllinoisUniversityinCarbondale.She from OhioStateUniversityandmaster’sdoctoraldegreesin upon pilotdecisionmaking.Shehasabachelor’sdegreeineducation to pilotweathertrainingandknowledgetheinfluenceemotionhas on theflightdeck.Burianisalsoinvolvedinworkandresearchrelated 26 The aviationindustrylackssubstantive humanperformance • By BarbaraK.Burian ISA SI 2003 1 (EAS) ProjectcurrentlyunderwayintheHumanFactors addressed intheEmergencyandAbnormalSituations his paperreviewstheobjectives,goals,andissuesbeing Examplesofvariousissuesbeingcoveredintheproject The EmergencyandAbnormal P r oceedings others respondtoemergencyandabnormalsituations factors thataffectthewaysinwhichflightcrewsand multiyear projectthatexaminesavarietyofissuesand Research Center.TheEASProjectisalarge, through SanJoseStateUniversityattheNASAAmes Abnormal Situations(EAS)ProjectDirector,working Dr. BarbaraBurian Research Center;andImmanuelBarshi, , Ph.D., istheEmergencyand Situations Project

SJSUF /NASAAmesResearchCenter;R.KeyDismukes, • • • • • Issues RelatedtoHumans • • • Issues RelatedtoChecklistsandProcedures • • cies andAbnormalSituations • Broad OverarchingIssues Table 1:EASProjectTaxonomyoftheDomain mal situations. checklists, writtenguidance,andtrainingforuseinnon-nor- agencies, andaircarriersindevelopingequipment,procedures, are alsoconcernedwiththerolesofmanufacturers,regulatory cially cabincrew,controllers,dispatchers,andmechanics. in theaviationsystemwhohelpmanagethesesituations,espe- tions wemustalsoconsidertheirinteractionwithotherplayers the issuesaffectingcrews’abilitytomanagenon-normalsitua- project islargelyorientedaroundtheflightcrew,tounderstand various populationswithintheaviationindustry.Although guidelines forbestpracticestargetedtothespecificneedsof field guidesthatwillsummarizewhatwelearnandprovide the aviationcommunity.Ultimatelywewillproduceaseriesof ies inclosecollaborationandconsultationwithpartnersfrom vulnerabilities inreal-worldemergencyandabnormalsituations. human performancelimitationsandcapabilities,cognitive agement basedonknowledgeoftheoperationalenvironment, and certification,training,crewcoordination,situationman- is todevelopguidanceforprocedureandchecklistdevelopment cussed ingreaterdetailbelow.OuroverridinggoalforthisProject undertaken toaddresstheseandrelatedconcerns,whicharedis- conditions). other traffic;respondingtoemergenciesindeterioratingweather ing withATC,dispatch,maintenance,andcabincrew;avoiding procedures inthecontextofreal-worddemands(e.g.,coordinat- vides limitedopportunitytopracticeemergencyandabnormal and limitations.Pilotinitialrecurrenttrainingcurrentlypro-

Ph.D., We areworkingtowardthisgoalthroughseveralfocusedstud- The EmergencyandAbnormalSituations(EAS)Projectwas R P Human P Checklist Use Cr Checklist T Checklist Structur Development ofChecklistsandP Definitions andP Economic andR Philosophies, P ersonnel Issues oles ofOthers ew Coor

NASA AmesResearchCenter erformance dination andR ype andA olicies, andP egulatory P erspectives e andDesign vailability esponse r ractices ofDealingwithEmer essur r es ocedur

Ph.D., es

NASA Ames 2 gen- We • Issues RelatedtoTraining ten challengehumaninformationprocessingcapabilities. able toreadilyretrievefrommemory,duringsituationsthatof- degree ofknowledgethatcrewsareexpectedtohave,andbe flight crewswhenrespondingtonon-normalsituationsandthe by checklistdesigners’philosophiesregardingthedesiredroleof included inchecklistsalsovaries.Bothoftheseissuesaredriven condition. Theextentofexplanatoryorannotatedinformation isolating theproblemtoallowcontinuedflightinanon-normal to whichtheyfocusontroubleshootingaproblemversussimply ample, checklistsfromdifferentorganizationsvaryinthedegree involved inthesesituationsinfluencetheirperformance.Forex- how theperceptions,attitudes,andpracticesofthosedirectly to emergencyandabnormalsituations.Wearealsointerestedin crews, airtrafficcontrollers,andotherswhomustdirectlyrespond shape thematerialsandguidancetheyprovideflightcabin practices ofmanufacturers,regulatoryagencies,andaircarriers Here, weareconcernedwithhowthephilosophies,policies,and practices ofdealingwithemergenciesandabnormalsituations. overarching issues.Thefirstinvolves Three categoriesinthetaxonomyrelatetoratherbroad, Broad, overarchingissues and illustratethemwithexamplesfromairlineaccidents. paper, organizedaroundourtaxonomy,wesketchouttheissues taxonomy helpsguideourworkinthefocusedstudies.Inthis we refertoasthe“TaxonomyofDomain”(able1).This sorted theseinto15categories,groupedinsixlargerareas,which This Projectaddressesmanydiverseissuesandconcerns.Wehave EAS Project—taxonomyofthedomain • Emergency EquipmentandEvacuations • • Issues RelatedtotheAircraft the averageamount oftimebetweenthedetection ofanonboard occurred over31years(TSB, 2003).TheBoarddeterminedthat tion SafetyBoard(TSB)ofCanada studied15inflightfiresthat coast ofNovaScotia,Canada, onSept.2,1998,theTransporta- Following theinflightfireand subsequentcrashofaMD-11offthe Examples fromrecentaccidents normal situations. velopment anddesignforthebestwaystomanage non- questions havegreatrelevanceforchecklistandprocedure de- checklists andthestepsthatcomprisethem?Theanswersto these what aretheobjectivesandgoalsofemergencyabnormal circumstances? Howdoproceduresandchecklistsdiffer? And, Does suchadistinctionmatter,andifso,when,orunderwhat difference betweenanemergencyandabnormalsituation? is athirdcategoryinourtaxonomy.Forexample,whatthe Several issuesrevolvearound normal situationsrespondtothedemandsofthosesituations. tory pressures We arealsointerestedinthewaysthat Flight andCabinCr Selected Emer Automation Issues Critical Air craft Systems influencehowthevariousplayersinvolvedinnon- gency EquipmentandEvacuationIssues ew Emer definitions andperspectives gency T philosophies, policies,and raining economic andregula- 3 , which safe wejustwanna makesureyouhaveallof theuh…allinfo. DIS: Wellwewannadowhat’s safesoifthat’swhatyoufeelisuh airports. you’re callinguscauseI’mconcerned aboutoverflyingsuitable CA: Ireallydidn’twanttohear abouttheflowbeingreason CA: Welluhyuyouehhuh…boy youputmeinaspothereum…. back inSanFrancisco. we haveamajorflowprogramgoingrightnowuhthat’sfor ATC LA uhwe’llbelookingatprobablyanhourtoand ahalf we willdothatuhwuwe’lltellyouthoughifland in DIS: …IfuhyouwanttolandatLAofcourseforsafetyreasons company dispatcher(DIS)overtheradio: 2002). Thisfirstwasanexchangebetweenthecaptain(CA) anda that crashedoffthecoastofCaliforniaonJan.31,2000(NTSB, from thecockpitvoicerecorder(CVR)transcriptofanMD-83 influence theirperspectivesandactions.Belowarefourexcerpts ties ofindividualsinvolvedinmanaginginflightemergencies can economic issuesmayinfluencechecklistdesign. cost-benefit tradeoffs.Thisclearlyillustrateshowphilosophyand diverted unnecessarily.Thisofcourseisapolicydecisioninvolving a smokeorfirechecklist,evenifitmeansthatmanyflightsare haps “Diverttothenearestairport”shouldbeatorneartopof by acrewmakingdiversionatthefirstsignoftrouble,thenper- versely, however,ifevenonecatastrophicaccidentmightbeaverted ing attheendofachecklistwouldnotseemunreasonable.Con- situations, then,perhaps,placingacheckliststeprelatedtodivert- planned destinations?Ifthishappensinthevastmajorityofthese nate smokeorextinguishafireandcontinuetheirflightsto sible? Howoftenarecrewsabletosuccessfullyisolateandelimi- will thecrewsbecomesoincapacitatedthatadiversionisnotpos- have failed?Iftheisolationoreliminationofsmokeisnotattempted, ous thatadiversionisrequired?Onlyafterattemptsatisolation philosophies illustratedbythedesignofthesechecklists. life anddeath.Thus,itisimportanttoevaluatetheunderlying gency descentandlandingmightmeanthedifferencebetween ever, inotheraccidentsapromptdecisiontoconductanemer- soon asthefireswerediscovered(TSB,2003;NTSB,1997).How- even ifthecrewshaddivertedtomakeemergencylandingsas spread, itisunlikelythatthecrashescouldhavebeenavoided two accidents,becauseofthespeedwithwhichfiresonboard tained noitemregardingdivertingtoanearbyairport.Inthese Unlike thesecondMD-11checklist,however,thischecklistcon- tains numerousstepsdesignedtoisolatethesourceofsmoke. (National TransportationSafetyBoard(NTSB),1997)alsocon- DC-9 thatcrashedintheFloridaEvergladesonMay11,1996, TRICAL SMOKEORFIREchecklistavailabletothecrewofa are noteliminated,landatnearestsuitableairport.”TheELEC- list thecrewwouldhavelikelyaccessedstates:“Ifsmoke/fumes tem (TSB,2003).Thefinalstepatthebottomofsecondcheck- tions andconsequencesrelatedtotheidentifiedinoperativesys- isolate theoriginofsmoke,followedbyalistsystemlimita- crew ofthisMD-11revealsaseriescheckliststepsdesignedto these flightshadlessthanthatamountoftime. of timetocompleteadiversionfromcruisealtitude,andhalf landing, orcrashedwas17minutes.Seventeenminutesisnotalot fire andthepointatwhichaircraftditched,conductedaforced Another accidentillustratesthewaysdifferentjobresponsibili- When arecrewstoconsiderthattheirproblemmaybesoseri- Examination ofthesmoke-relatedchecklistsavailableto ISASI 2003 Proceedings • 27 ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS it schedulingairplanes, takingcareoflanding logistics,trouble- to setasidethemindsetfortheir normalmodeofoperation—be as theynormallydothem.It can beverydifficultforindividuals and maintenancepersonneldo. Bothweretryingtodotheirjobs fixing thingsthatarebrokenon airplanesiswhatoperationsagents the crewtroubleshoottheirproblem. Takingcareoflogisticsand tics wereinorder.Themaintenance technicianwastryingtohelp the operationsagentwasconcernedwithmakingsureall logis- frustrated. Heknewhehadaseriousproblemonhishands but use thealternate.Nothinghappens. me themotor’stryintorunbutbrakewon’tmoveitwhen we it spikesoutwhenweusetheprimary.WegotACloadthattells CA: Yup,itsjustitappearstobejammedtheuhwhole thing MX: Andalternate’sinoptoohuh? CA: Yeawetriedjustabouteveryiteration. nothing. uh otherswitchesthanthesuitcasehandlesaloneor and thesuitcasehandlestoseeifitwasmovininwithanyof MX: umyeaIjustwantedtoknowifyoutriedthepicklesswitches everything…. CA: Yeawetriedeverythingtogether,uh…we’verunjustabout right? MX: Yeadidyoutrythesuitcasehandlesandpickleswitches, script): sulted withmaintenancepriortothebeginningofCVRtran- from maintenance(MX.Itappearsthatthecrewhadalsocon- ter startthatnowcausewe’recomintoyou. CA: OkIunders…Irememberthisiscomplicated.Yeawell.Bet- Customs first. long that’sgonnatakeme…butuhIhavetoclearitallthrough tional arrivalwehavetogetlandingrights.Idon’tknowhow OPS: Okalsouh….justbeadviseduhbecauseyou’reaninterna- was recorded: an operationsagent(OPS)inLosAngeles,thefollowingexchange national Airport(LAX).Soonthereafter,inaconversationwith back andbegintheirdecentforalandingatLosAngelesInter- landing datacouldbeobtainedatwhichtimetheywouldturn lations.) violate legalrequirementsincludedinthefederalaviationregu- able airports”mayhavealsobeenmotivatedbyadesiretonot for anemergencylanding.(Hisconcernabout“overflyingsuit- concern inthefirstexchangeaboutover-flyingairportssuitable was serious,basedupontheotherrecordedcommentandhis ever, thecaptainbythistimewasclearlyawarethattheirsituation aware howserioustheproblemfacedbycaptainwas.How- job. Itisnotclearfromthistranscriptthatthedispatcherwas and adherencetoaschedule—centralaspectsofdispatcher’s airplane’s idn’tgonnagoanywhereforawhile…. gonna fixit,nowthey’reworriedabouttheflow.I’msorrythis know, itjustblowsmeawaytheythinkwe’regonnaland,they’re CA: …just…drivesmenuts…notthatIwannagoonaboutityou someone ontheflightdeck(mostlikelyaattendant): wind againthereinSanFrancisco? CA: Yeawekindaassumedthathad…what’stheuh 28 In theconversationwithoperationsagent,captain was Following thisexchange,thecaptaintalkedwithanindividual The captaindecidedtocontinuetowardSanFranciscountil The dispatcherwasconcernedaboutthemovementofaircraft Soon afterthisexchange,thecaptainwasrecordedsayingto • ISA SI 2003 P r oceedings a checkliststep directingthecrewstoreset the circuitbreakers. checklist, underaheadingentitled “Approachandlanding,”was breakers toplacethesystems intheflightmode. same checklist,thecrewpulled thegroundcontrolrelaycircuit ing systemswerestillinthe ground mode.Asdirectedbythe later theyrealizedthatthecabin pressurizationandtakeoffwarn- this allowedthemtoresolvetheirgearproblem,afewminutes included intheirquickreferencehandbook(QRH).Although used theUNABLETORAISEGEARLEVERprocedurethat was landing gearontheirdepartureclimbout(NTSB,1996). They On Jan.7,1996,thecrewofaDC-9haddifficultyraising the Examples fromrecentaccidents category. checklists aredisplayedfirstalsoconsideredunder this prioritization schemesthatdeterminewhichintegratedelectronic cally accessthechecklists.Issuesrelatedtocomputerized non-integrated electronic,etc.)andthewaysthatcrewsmustphysi- tion ofchecklists(i.e.,paper,mechanical,integratedelectronic, cerned withissuesrelatedtothemodalityusedforpresenta- style guide,formatting,andlayoutconsiderations. the flightcrewsmayreference.Ofcourse,wearealsointerestedin those usedbycabincrews,withtheMELs,andothermaterial tency ofthechecklistsandproceduresusedbyflightcrewswith steps areintegratedwithnon-normalchecklistsandtheconsis- ture, etc.).Weareinterestedinwhetherornotnormalchecklist checklist inthefirstplace(e.g.,organization,indexing,nomencla- specific situation)andwhatfeatureshelpcrewslocatetheproper navigate withinachecklist(tofindtheactionsappropriatefortheir or recallitems.Weareconcernedwithhowwellcrewsable to completeachecklist,andtheinclusion(orexclusion)ofmemory and where,missingorincorrectitems,thelengthoftimerequired useful tothecrews,includingwhichitemsareplacedonchecklists important issuesthatinfluencethedegreetowhichchecklistsare different aircraftfleets? ment andtowhatdegree,ifany,cantheybestandardizedacross and proceduresreflecttherealitiesofoperationalenviron- ing checklistdevelopmentanddesign?Howwelldochecklists guidance existsfordevelopersandregulatoryagenciesregard- regulatory approvalobtainedforchecklistsandprocedures?What and howarechangestochecklistsmaderecorded?Howis types ofsituationswarrantthedevelopmentachecklist?When what checklistsandproceduresaredevelopedbywhom.What of ChecklistsandProcedures used inemergencyandabnormalsituations.Inthe taxonomy ofthedomainpertaintochecklistsandprocedures The issuesgroupedinthenextthreecategoriesEASproject Issues relatedtochecklistsandprocedures 1998). when awareofcircumstancesthatshouldcompelrevision.(Klein, that individualsareslowtoreviseanestablishedmindseteven plane safelyontheground.Cognitiveresearchhasdemonstrated and toputallotherconsiderationsasideinordergettheair- B— torecognizeandcommunicatetheseverityofasituation shooting systemsproblems,orflyinganaircraftfrompointAto In alaterportionoftheUNABLE TORAISEGEARLEVER In the The Checklist StructureandDesign Checklist TypeandAvailability

category, weareconcernedwith categoryincludesmany category,wearecon- Development communication, coordination,andcrewresource management the nature ofthethreatanddegree ofriskareconsideredunder making, prioritizationoftasks, andaccurateassessmentofthe taxonomy categories. specifically relatedtohumans inthesesituationscomprisefive emergency andabnormalsituations.IntheEASProject, issues human elementwillcontinuetobecrucialintheresponse to out pilots,cabincrew,dispatchers,andairtrafficcontrollers, the Until thedaythatcommercialtransport-categoryaircraftfly with- Issues relatedtohumans the aircraft. flight manualupdatesfromthemanufacturerafterpurchasing of theaccident,however,butoperatordidnotcontractfor Such achecklistwasavailablefromthemanufacturerat time recting afuelimbalancethatoccursduringthetransferoffuel. but lostcontrolandalleightindividualsonboardperished. dual-engine flameout,thecrewattemptedanemergencylanding gines wereusingfuelfromtherightwing.Fearinganimminent the fuelinleftwinghadbecome“trapped”andthatbothen- the righttoleftwingtank.Thecrewincorrectlybelievedthat and resultedinaseverefuelimbalanceaswaspumpedfrom ferred totherightwingduringnormalfueltransferprocedures on thefuelcontrolpanel.Thispreventedfrombeingtrans- uncommanded afterenginestartbecauseoftwobondedcontacts the crew,rightstandbyfuelpumpcontinuedtooperate 1995 (FuelImbalanceCited,2000).Duringtheflight,unknownto Learjet 35A)thatoccurredinAlexanderCity,Ala.,onApril17, be foundintheaccidentofaC21A(aU.S.militaryversion limited andworkloadstressmaybehigh. QRH) sothatsuchreasoningisnotrequiredwhentimemaybe list stepsintheprimaryresourcereferenced(inthiscase, have allinformationnecessaryforthepropercompletionofcheck- only beresetonceontheground.Webelievethatcrewsshould have beenabletoreasonoutthatthesecircuitbreakersshould situation. Somepeoplemayarguethattheaccidentcrewshould fer onlytotheQRHwhenhandlinganemergencyorabnormal pilots employedbythesameaircarrier,theyweretrainedtore- according tointerviewswiththeaccidentflightcrewandother during taxiandverifythatcircuitsareinthegroundmode.”Yet ance wasgiven:“ResetGroundControlRelaycircuitbreakers that appearintheaircraftoperatingmanual(AOM),suchguid- completed. However,intheexpandedorannotatedchecklists to thecrewaboutspecificallywhenorhowthisstepwasbe ing” sectionoftheQRHchecklist,nofurtherguidancewasgiven ground controlrelaycircuitbreakersinan“Approachandland- on adifferentrunway. was abletocompleteasuccessfulgo-aroundprocedureandlanded from theaircraft.Theaircraftbouncedintoairandcrew lost lift,hitthegroundveryhard,andnosewheelseparated ground modecausingthespoilerstodeploy.Theaircraft ing this,however,relatedsystemsimmediatelywentintothe proximately 100feet(30.5meters)abovetheground.Upondo- The crewdidthiswhileonfinalapproachintoNashville,ap- Issues suchasthedistribution ofworkloadandtasks,decision The crew’sflightmanualdidnotcontainachecklistforcor- Another exampleofachecklist-orprocedure-relatedissuecan Other thanincludingtheiteminstructingcrewtoreset Crew CoordinationandResponse category.Issuesrelatedto incident. However, thecaptaindidnotcall for andthecrewdid formed anemergencydescent andtheylandedwithoutfurther was stillonhisinitialoperating experienceintheB-727,per- lost consciousness.Thefirstoffice r, whowasthepilotflyingand attendant, whohadbeenon theflightdeckattime,briefly the aircraftrapidlylostpressurization. ing theoutflowvalve,flightengineeractuallyopened it and PACK TRIPchecklist,whichisfairlylengthy.Also,insteadof clos- not refertothePACKREINSTATEMENTFOLLOWINGAUTO outflow valve.”Whentakingtheseactionstheflightengineer did the rightpackonandthen“wenttomanualACclosed the engineer toreinstateit.Theflightstatedthatheturned that thesecondpackwasoff.Thecaptaininstructed flight cruise altitudeof33,000ft.(10058.4meters),thecaptainnoticed a cabinaltitudewarningthatsoundedrightbeforereaching their (NTSB, 1998a).Whileshowingtheflightengineerhowtosilence and performedanemergencylandingatIndianapolis,Ind. On May12,1996,aB-727experiencedrapiddecompression Examples fromrecentaccidents tent withandcomplimenteachotherisparticularlyimportant. which theproceduresandchecklistsofdifferentpartiesareconsis- outcome ofemergencyandabnormalsituations.Thedegreeto nation betweenthesevariousgroupsisessentialforthesuccessful ations areresolved.Ahighdegreeofcommunicationandcoordi- and evenpassengerseachplayimportantpartsinhowthesesitu- personnel, airportrescueandfirefightingMedLink, the situations.Airtrafficcontrollers,dispatchers,maintenance crews tothesesituationsbutalsothe normal situationsarehandled. can haveonworkingrelations,communication,andhownon- on situationresponseandtheinfluencethatcompanymergers (two-person crews,three-personcrews)andcabincrewsizehas mal situations.Weareinterestedintheeffectthatflightcrewsize and experiencelevelsmayhaveonpilots’responsetonon-nor- influence thatbackground,previousexperiences,initialtraining, to stressisalsobeingconsideredwithinthiscategory. and heavyworkload.Flightcrews’emotionaloraffectiveresponse ability torecallproceduralanddeclarativeknowledgeunderstress interested intheconditions,factors,andcuesthataffectpilots’ attention, retrievalfrommemory,andproblemsolving.Weare tions. Weareexaminingtheeffectsofstressandtimepressureon man Performance for dealingwithemergencyorabnormalsituations. amount of“blindfaith”crewsplaceinchecklistsorprocedures accessed butnotcompliedwith.Wearealsoconsideringthe lists orprocedureswerenotusedatallinwhichthey the specificsituation.Weareexaminingsituationsinwhichcheck- ing directions,orcompletingthewrongconditionalbranchfor such asinadvertentlyskippingchecklistitems,misunderstand- how wellcrewsareabletomanagenon-normalsituations. ity toconfusion,fixation,distraction,andoverloadmayaffect sidered. Wearealsoconcernedwiththewaysinwhichvulnerabil- (CRM) withinandbetweenflightcabincrewsarealsocon- During thisevent,thecaptain, flightengineer,andlead We arenotjustconsideringtheresponseofflightandcabin Under the As researchpsychologists,weareparticularlyinterestedin Under Checklist Use Personnel Issues underhigh-stressandhigh-workloadcondi- weareinterestedinerrorsmadebycrews categoryweareexaminingthe Roles ofOthers ISASI 2003 Proceedings involvedin Hu- • 29 ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS approaches used fordealingwiththesesituations duringboth examining. Wearelookingat varioustrainingtechnologiesand gency andabnormalsituations isalsoanimportantareaweare The Issues relatedtotraining reported thathehadfeltvery overloaded. During aninterviewfollowingtheincident,flightengineer was toldtheidentifieratleasttwice,heapparentlyneverheard it. obtain landingdatafromanonboardcomputer.Although he letter identifieroftheairporttheyweredivertingtosohe could five timesoverthespanof6minutes,heaskedfor three- monitor thestatusofaircraftpressurizationondescent, and The flightengineermissedchecklistitems,didnotadequately marks tothecrewoverATCfrequencyonseveraloccasions. completion ofthechecklistsandmistakenlytransmittedhis re- completed. Hedidnotadequatelymonitortheflightengineer’s sion todivertandlandnevercalledforanychecklistsbe was evidentinmanyways.Thecaptaindelayedmakingthedeci- ence ofthegreatstressandoverloadcrewwasexperiencing jumpseat ridersbarelyescapedtheburningaircraft.Theinflu- crew’s evacuation. maining partiallypressurized,whichimpededanddelayedthe list itemsanduncompletedchecklistsresultedintheaircraftre- for figuringthelandingspeeds.Uponlanding,missedcheck- normal approachandlandingchecklistsobtaindataneeded gency checklistshedidaccess,wasalsotryingtocompletethe the flightengineerwasattemptingtocompletetwoemer- evacuation checklistswereneveraccessedorcompleted.While procedures todoso.Theemergencydescentchecklistandthe did notcallforanychecklists,althoughrequiredbycompany flight. Thecaptaindidnotnoticetheflightengineer’serrorsand the firstofficer,whoremainedpilotflyingthroughout progress ofthefireandcoordinatingdiversionwithATC checklist heconducted.Thecaptainwasbusymonitoringthe SMOKE LIGHTILLUMINATEDchecklist,whichwasthenext the fire-detectionsystem,andothercockpitduties. to thishewaspreoccupiedwiththewarningindicators,testing half minutesafterthewarninglighthadfirstilluminated.Prior to theclosestairportbutonlysomewhatlater—three-and-one- The captaindidcommandanemergencydescentanddiversion ditional branchonthechecklistfor“IfDescentisNOTRequired.” from thecaptain,flightengineerchosetocompletecon- checklist andcontinuedtoexecuteit.However,withoutinput FIRE checklistandthentheflightengineeraccessedprinted the crewcompleted,memoryitemsfromSMOKEAND a realfireontheirhands.Theflightengineerannounced,and light illuminatedandthecrewveryquicklyrealizedthattheyhad at 33,000ft.(10058.4meters)thecabin/cargosmokewarning sues relatedtohumanperformanceunderstress.Duringcruise 5, 1996(NTSB,1998b),providesseveralgoodexamplesofis- the firstofficerdidimmediatelydonhismask. tude warningsounded,asrequiredbyprocedures—fortunately, tain didnotputhisoxygenmaskonimmediatelywhenthealti- pression checklistortheemergencydescentchecklist.Thecap- not completeanyemergencychecklists,includingthedecom- 30 This wasaveryseriousemergency,andthecrewtwo The flightengineermissedtwostepsontheCABIN/CARGO An accidentinvolvingaDC-10nearNewburgh,N.Y.,onSept. • ISA Training SI 2003 P that flightandcabincrewsreceive regardingemer- r oceedings freezing. Onlift off,clearicebrokeoffthewings andwasingested The daywassnowyandwindy,withtemperatureshoveringaround failure shortlyafterdeparturefrom Stockholm(Martensson,1995). Sweden, andinvolveda situations. Oneaccidentoccurred onDec.27,1991,atGottrora, mation issuesfactorintheresolution ofemergencyandabnormal Two recentaccidentsprovide examplesofhowaircraftandauto- Examples fromrecentaccidents hand-flying skillsarealsobeingconsidered. sues involvingrevertingtomanualflyinganddegradation of tween highlyautomatedaircraftandlessaircraft. Is- paring proceduresforemergencyandabnormalsituations be- rectly toanon-normalsituation.Wearealsointerestedin com- automation mayleadcrewstomisdiagnoseorrespond incor- We areexploringthedegreetowhichuncriticalacceptance of mation aremostappropriatetouseandunderwhatconditions. ations. Weareinterestedinlearningwhatlevelandtypesofauto- response. pit warningsandwarningsystemsmayfacilitateorimpedeacrew’s these situationsarehandled.Wealsointerestedinhowcock- that systemswithflightprotectionenvelopesmightplayinhow named the outcomeofnon-normalsituations.Inacategorywehave We arealsolookingatissuesrelatedtotheaircraftthatinfluence Issues relatedtotheaircraft ing performancewasproperlymonitoredandmanaged. and companyrecordsdidnotprovideenoughevidencethattrain- equately addresstherecognitionofanenginefailureatlowpower determined thattrainingprovidedbythecompanydidnotad- an enginefailure.Duringtheaccidentinvestigation,itwasalso flight crewstoalwaysassociateanilluminatedignitionlightwith was discoveredthatthecompanyhadincorrectlytrainedtheir the flightuntilimpact.However,inaccidentinvestigationit mined thatbothengineshadfunctionednormallythroughout anomaly” (Commutercaptain,1996,p.8)anditwaslaterdeter- sengers survivedtheaccident. tain lostcontroloftheaircraftanditstruckterrain.Threepas- an enginefailure.Duringamissedapproachprocedure,thecap- undertake anyabnormaloremergencyproceduresassociatedwith pilot flying,didnotfeatherthepropeller,secureengine,or that theleftenginehadflamedout.Thecaptain,whowas flight idle,anilluminatedignitionlightledthecaptaintobelieve scending through2,100ft.(640meters),withthepowerleversat 1994 (Commutercaptain,1996).Atthefinalapproachfixde- approach totheairportatRaleighDurham,N.C.,onDec.13, be foundinthecrashofaBAeJetstream32thatlostcontrolon A particularlytragicexampleofsomethesetrainingissuescan Examples fromrecentaccidents in whichmultipleproblemsoccurconcurrently dard” orambiguousproblemsandhowtorespondsituations ested intrainingtohelppreparecrewsdealwith“non-stan- are rarelypracticedorevendiscussed.Weparticularlyinter- to skillacquisitionandretention,especiallyofproceduresthat initial andrecurrenttraining.Weareinterestedinissuesrelated Automation Issues The illuminatedignitionlightwasactuallya“minortransient Critical AircraftSystems, are alsoimportanttoconsiderinthesesitu- n MD-81thatexperiencedadual weareinterestedintherole - engine in waysthatdid nothelpthem.Duringnormal operationstheuse manually; theycontinuedtotry useautomation(i.e.,AltitudeHold) at leastinitially.Additionally,the crewdidnottrytoflytheaircraft autopilot andautothrottlescontributed greatlytotheirproblems, and indicatorsweregreatlyconfusing tothecrew,andcenter mation ontheprimaryflightdisplays. Thecontradictorywarnings by andcontinuedtotryuse (and beconfusedby)airspeedinfor- was displayingcorrectairspeedinformation,theyweredistracted was nospecificairspeeddiscrepancywarningontheB-757. generally includedinchecklistsrelatedtothesemessages. There indicated airspeedsisfarfromintuitiveandnotinformation ticular messagestothedifferenceincaptain’sandfirstofficer’s would haveprovedusefultothem.Therelationofthesetwo par- unlikely thattheyhadtimetodosoortherelatedchecklists RUDDER RATIOorMACH/SPDTRIMadvisories;however, itis autopilot waserroneous.Thecrewdidnotattempttoclarify the by theleftADCtocaptain’sairspeedindicatorandcenter completely blockedandconsequentlytheinformationprovided formation totheleftairdatacomputer(ADC)hadmostlikelybeen into theoceanashorttimelaterandalllivesonboardwerelost. selection ofAltitudeHoldwasineffectual.Theaircraftcrashed were attoolowofapowersettingtomaintainaltitudesothe to leveloffandstabilizetheaircraft.However,throttles selected AltitudeHoldonthemodecontrolpanelinanattempt power andthenremoveditmorethanonce.Thefirstofficer tried tosortoutthetruenatureoftheirproblem,theyapplied autopilot andautothrottlesautomaticallydisengaged.Asthecrew bewilderment, thestallwarning“stickshaker”activatedand in anattempttobringthespeeddown.Addingcrew’s attitude andtheautothrottlesmovedtoaverylowpowersetting tionally, thecenterautopilotcommandedan18-degreenose-up high speedandtheoverspeedwarningclackersounded.Addi- tem (EICAS)display:RUDDERRATIOandMACH/SPDTRIM. sages appearedontheengineindicatingandcrewalertingsys- nate airspeedindicators.Afewsecondslatertwoadvisorymes- ancies appearedbetweenthecaptain’s,firstofficer’s,andalter- working oncetheaircraftbegantoclimbbutsignificantdiscrep- that hisairspeedindicatorwasnotworking.Itappearedtostart Sumwalt, 2000).Duringthetakeoffroll,captainindicated coast oftheDominicanRepubliconFeb.2,1996(Walters& trated byanaccidentinvolvingaB-757thatlostcontroloffthe pany hadnoknowledgeoftheATRfeature. the accidentinvestigation,itwasdiscoveredthatairlinecom- the surgingandcontributedtofailureofengines.During feature. Thisincreaseinenginepowerincreasedtheintensityof increased automaticallybytheautomaticthrustrestoration(ATR) all 129onboardsurvived. ments. Despitetheaircraftbreakingintothreepiecesonlanding, crew toattemptanemergencylandingusingonlyback-upinstru- electronic flightinstrumentsystem(EFIS)wentblank,forcingthe both engineslostpower.Greysmokedfilledthecockpitand lift off,theleftone39secondslater.At3,194ft.(973.5meters) engines tosurge—therightonebegansurge25secondsafter by theengines,damagingfanstages.Thisdamagecaused Although thecrewagreedthatalternateairspeedindictor Investigators laterdeterminedthatthepitottubeprovidingin- By thistime,thecaptain’sairspeedindicatorshowedavery Issues relatedtoautomationandwarningsystemsareillus- On climboutthecrewdidnotnoticethatenginepowerwas By identifying latentvulnerabilitiesanddelineating issues,the domain, morethancanbeanswered definitelybyanyoneproject. all thoseassistingthem.Many issuesareinvolvedinthislarge skilled performancebyflightcrews andclosecoordinationamong These situationsareinherently challengingandrequirehighly abnormal situationsoccuron flightsacrosstheworldeveryday. are infrequentinairlineoperations. However,emergencyand Fortunately, seriousaccidents,suchastheoneswehavereviewed, Conclusion cluded communication. cupied withotherdutiesorifthereinforcedsecuritydoor pre- crew didnotheartheflightattendantbecausetheywerepreoc- currently underinvestigationsoitisstillunknownifthe flight banging northeloudtalkingthroughdoor.Thisincident is ing loudly.Theflightcrewreportedthattheyheardneither the tion oftheflightcrewbybangingoncockpitdoorandspeak- After landing,theleadflightattendanttriedtogetatten- make announcementsandcontactthecockpitwasinoperative. ing smellinthecabinanddeterminedthathandsetusedto in thecockpit.Theforwardflightattendantsalsonoticedaburn- gan flashingrandomly.Theflightcrewnoticedaburningsmell display unitsandstandbyinstrumentswentdarkthenbe- of aB-717-200indicatedthattheleftgeneratorhadfailed.The ing thefinalapproachfix,engineandalertdisplay(EAD) 2003 inFlushing,N.Y.(NTSB,2003;preliminaryreport).Near- crews likelyinfluencedtheevacuationofaflightonMarch26, the oxygenmaskmicrophone. kept slidingaroundandcommunicationwasverydifficultthrough mask onherheadandtofacebecauseitwassolargethat after theeventoccurred,firstofficerhadtoholdoxygen 1989). AlthoughthecrewwasabletolandatMaui13minutes (7315.2 meters),causinganexplosivedecompression(NTSB, 727-200 asitwaslevelingoutatacruisealtitudeof24,000ft. 28, 1988,whenan18ft.sectionoffuselageseparatedfromaB- emergency. OnesuchaccidentoccurredinMaui,Hawaii,onApril gency equipmentimpededaflightorcabincrew’sresponsetoan There havebeenmanyaccidentsinwhichproblemswithemer- Examples fromrecentaccidents nication betweenthecabinandcockpit. with issuessuchasthedecisionwhethertoevacuateandcommu- dressing thefullrangeofevacuationissues,weareconcerned error byflightattendants.AlthoughtheEASProjectisnotad- configurations andtypescreatevulnerabilitytoconfusion differences inemergencyequipmentamongdifferentaircraft use ofemergencyequipment.Wearealsointerestedinwhether crews receiveadequatetrainingandpracticetobeproficientin goggles thatdonotfitovereyeglasses.Anotherissueiswhether mal situationscanbeproblematictouse,forexample,smoke ment providedtoflightandcabincrewsforemergencyorabnor- involving Finally, theEASprojectisaddressingaselectedsubsetofissues and evacuations Selected issuesrelatedtoemergencyequipment ditions, theseproblemsarelikelyexacerbated. of automationcanbeconfusingforcrews;undernon-normalcon- Problems withcommunicationbetweenthecabinandflight Emergency EquipmentandEvacuations. ISASI 2003 Someequip- Proceedings • 31 ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS National TransportationSafetyBoard.(1997). National TransportationSafetyBoard.(1996). National TransportationSafetyBoard.(1989). Martensson, L.(1995).TheAircraftCrashatGottrora:Experiencesofthe Klein, G.A.(1998). Fuel ImbalanceCitedinLearjet35AControlLoss.(2000). Commuter CaptainFailstoFollowEmergencyProceduresAfterSuspected References 3 2 1 Footnotes situations frombecomingaccidents. In thisway,wehopetohelppreventemergencyandabnormal EAS Projectwilllayafoundationforestablishingbestpractices. 32 Allaccident-relatedinformationincludedinthispaperhasbeentaken Forsimplicity,theterm“non-normalsituations”willoccasionallybeused TheEmergencyandAbnormalSituationsProjectisfundedthroughthe PB96-910407). Nashville, Tennessee,January7,1996. Flight/Hard Landing,ValuJetAirlinesFlight558,DouglasDC-9-32,N922W, Author (NTISNo.PB89-910404). 737-200, N73711,NearMaui,Hawaii,April28,1988. Cockpit Crew. MA: MITPress. dation—Accident Prevention,57(2). (1996). Engine Failure,LosesControloftheAircraftDuringInstrumentApproach. from thereportsofinvestigativebodiesinvolved. to referbothemergencyandabnormalsituations. Training ElementofNASA’sAviationSafetyandSecurityProgram. • ISA SI 2003 Flight SafetyFoundation—AccidentPrevention,53(4). The InternationalJournalofAviationPsychology,5(3)305-326. P r oceedings Sources ofpower:Howpeoplemakedecisions. Washington, D.C.:Author(NTISNo. N Aloha AirlinesFlight243,Boeing In-flight FireandImpactwith Ground SpoilerActivationin Washington, D.C.: Flight SafetyFoun- Cambridge, ful commentsintheirreviewofanearlierdraftthispaper. Berman andTomChidesterwhoprovidedthoughtfulhelp- the preparationofthispaper.AppreciationisextendedtoBen the EASProjectTeam,assistedinlocatinginformationused Richard Geven,ToddKowalski,andSeanBelcher,membersof Acknowledgements Walters, J.M.,&Sumwalt,R.L.,III.(2000). Transportation SafetyBoardofCanada.(2003). National TransportationSafetyBoard.(2003). National TransportationSafetyBoard.(2002). National TransportationSafetyBoard.(1998b). National TransportationSafetyBoard.(1998a). Reports A98H0003). IWF, Peggy’sCove,NovaScotia5nmSW2September1998(TSBReportNo. lision withWater.SwissairTransportLimited,McDonnellDouglasMD-11HB- www.ntsb.gov/ntsb/brief.asp?ev_id=20030409X00464&key=1. NYC03FA067) B717-200, N957AT,Flushing,NY,March26,2003(NTSBIDNo. ington, D.C:Author.(NTISNo.PB-2002-910401). About 2.7MilesNorthofAnacapaIsland,California,January31,2000 Pacific Ocean,AlaskaAirlinesFlight261,McDonnellDouglasMD-83,N963AS, York, September5,1996. ing, FederalExpressFlight1406,DouglasDC-10-10,N68055,Newburgh,New ntsb/brief.asp?ev_id=20001208X05709&key=1. CHI96IA157 Boeing 727-290,N775AT,Indianapolis,IN,May12,1996(NTSBIDNo. PB97-910406). Near MiamiFlorida,May11,1996. Terrain, ValuJetAirlines,Flight592,DouglasDC-9-32,N904VJ,Everglades, . Chap.4,pp.75-87.NewYork:McGraw-Hill. ). AbstractRetrievedJuly16,2003,fromhttp://www.ntsb.gov/ Gatineau, Quebec:Author. . PreliminaryReportRetrievedJuly16,2003fromhttp:// Washington, D.C.: Washington, D.C.:Author(NTISNo.

Author (NTISNo.PB98-910403). Aircraft AccidentAnalysis:Final Loss ofControlandImpactwith In-Flight Fire/EmergencyLand- American TransAirFlight406, Air TranAirwaysFlight356, In-Flight FireLeadingtoCol- . Wash- SESSION II ISASI 2003 Proceedings • 33 ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS 34 • of parts,etc. • Some ofthethingsyoumaybe lookingforcaninclude the causesofevent.Itcan eliminatecertainideasaswell. can bethekeyfordeterminingexistenceofcluesleading to struction isanexcellentinvestigativetoolwhenusedproperly. It initial sortingprocess. the actualrecoveryprocessofphysicalwreckage,oreven the what happensafterthepartsarerecovered—itwillnottalk about nized fashion. and afteraperiodofinvestigation,layingitoutinsome orga- from landorunderwater,sortingit,usuallybyairplanesection, same. Theyarepartofthesystemcollectingphysicalwreckage reconstruction about. Inmycorrespondence,Ihaveusedthewords for newideas. value indoingareconstruction.Finally,itwilllooktothefuture suggest certainstepstheinvestigatorcantaketoensurethereis mockups ofmajorportionsanairplane.Inaddition,itwill simple, minorpartslayoutupthroughcomplex2-Dand3-D tices andthevariouslevelsofaccidentreconstruction—froma and should,beimproved.Thispaperexaminesthecurrentprac- Sometime mockupsorreconstructionsjustseemtohappen. process forreconstruction investigators anddiscovered,nottomysurprise,thatthedecision quest forinformation,Iapproachedmanyoftheworld’sleading reconstruct wreckageaspartofanaccidentinvestigation.Inmy as theseeminglyhaphazardindustryprocessfordecidingwhento D tions totheindustry. ISASI’s prestigiousJeromeF.LedererAwardforoutstandingcontribu- Institute andisamemberofitsAdvisoryBoard.In2001,heearned (SSI). HeiscurrentlyinstructingattheSouthernCaliforniaSafety • patterns Why doareconstructioninthefirstplace?Clearly,recon- This paperwilldiscussthemockupanditsdecisionprocess— Before wegetintothedetails,let’sdefinewhataretalking There seemedtobegeneralagreementthatthesystemcould, pr structural inflightbr missile orgun pr • ISA ogr SI 2003 ess andeffectsoffir at thebackofmymind.OnethoseiswhatIperceive tion experience,severalissueshaveincreasinglygnawed uring mymorethantwodecadesofaccidentinvestiga- interchangeably.Inmymindtheyareoneinthe P Accident Reconstruction—The r oceedings consulting business,SafetyServicesInternational partner, KevinDarcy,formedanaviationsafety When heretiredattheendof1998,andhis commercial airplaneinvestigativeteamfor17years. he workedattheBoeingCompany,leadingBoeing’s in largeaircraftaccidentinvestigations.For41years, John Purvis ojectile, meteor hits,spacedebris,etc. eak up—br is By JohnW.Purvis(MO3002),Partner,SafetyServicesInternational isaninternationallyrecognizedexpert verylooseandpoorlydocumented. e, smok eak e, andheat—fir up patterns,sequence,loss Decision Process mockup e orsmok and e layout canbevery lowsinceitcouldbeaccomplished withexist- with thegroupchairmanmaking thedecision.Thecostofsucha The mockupmightbedone as,say,astructuralgroupactivity, Basic, simplelayout— entire section.Let’sexamine theseonebyone. of alimitedsectionorsystem oftheairplaneorarebuildan out onthefloor.Finally,there isthecomplex3-Dmockup,either be amorecomplex2-Dlayoutwhereoftheairplane islaid as thelavatorywejustmentioned.Graduatingfromthere would flat onthehangarfloor,orevenasimple3-Dconstruction such amine alimitedareaoftheairplane.It’sbasic2-Dlayout, done wreckage. Thiswouldtypicallybedoneonahangarfloor toex- be assimpleareassemblyofjustfewsignificantpieces of eral, eachhavingincreasedcomplexity.Thereconstruction can control theoverpressure. you coulddevelopstrongercargocontainersorothermeans to the overpressurewon’tcausecatastrophicstructuraldamage.Or airplane oritsoperatingprocedurestocontrolventingsothat sequence andhelpyoudeterminewheretomakechangesthe a cargohold,themockupmayleadyoutolookatbreakup similar situation.Forexample,ifyoususpecttherewasabombin lished, themockupcanplayanimportantroleinpreventinga • • • • need todetermineoriginandprogresspatterns • • inhalation • reconstruction. ture tomoreexactlydeterminetheeffectsofanexplosion. mini-mockup ofthelavatoryareaandnearbysurroundingstruc- analyzed. Ontheotherhand,onemaywishtoatleastbuilda take chemicalswabsfromthelavatorycomponentsandhavethem tory, withoutdoingamajormockup,maybepossible.Onecould occurred inalavatory,recoveringjusttheportionsofthatlava- to establishcauses.Forexample,ifyoubelievethatanexplosion ers themajority. • • • • • What kindofmockupsarewetalkingabout?There sev- Even ifthecausesofaccidenthavealreadybeenestab- Let’s lookatthetechnicalissuesthatmaysupportaneedfor Sometimes thereareotherwaystogettheinformationneeded No doubtmoreitemscouldbeaddedtothislist,butcov- Evidence derivedfr Evidence fr Major missingparts P Evidence offir Sear Evidence fr interactions betweendiffer missing pieces chemical r overpr mid-air collision arts foundsomedistancefr ch fore essur esidue ofane om fullbodyX om othersystemsanalysis xplosives e, suchasfr e andsmok om abasic,simplelayout The decisionprocessislikewise simple. om abomborothere xplosive device -rays andautopsies,burns,smok e onthestructur ent airplanesystems om thewr eckage sites es andsystems—the xplosion e what wasmissing, andwherethebasicparts oftheairplanehad River. Asthisdeveloped,wewere able tovisualizewhatwehad, out thepartsofairplane astheywererecoveredfromthe D.C., wehadanentirehangar flooravailableandwereabletolay the airplaneendedupin PotomacRivernearWashington, manufacturer, oftheareaunderstud y. technical information,you’llneedadiagram,probablyfrom the the expectedlengthofinvestigationandweathe r. For be indoorsoroutdoors.Theneedforaroofwoulddriven by some chalk,acleanfloor,andbasictechnicalinformation. Itcan make alarger2-Dmockupcanbetapemeasure,masking tape, and canbeextensivebutstillquitecosteffective.Thetools to Comprehensive 2-Dmockup— wreckage recovered. have theaddedbenefitofprovidingavisualinventory of the ing pointforamoreformaldecisiontogofurther.Allmockups determining probablecause.Thesimplelayoutisoftenthestart- cases, thismaybeallthatisnecessarytoassisttheinvestigatorin witness marksortoexamineburnsmokepatterns.Inmany laying outafewpiecestovisualizetheirrelationships,lookfor this isdoneinmostaccidents.Potentially,canbeasbasic ing personnelandinexistingspace.Somesortofalayoutlike Aft pressurebulkhead3-Dmockup(JALB-747,1985). (UAL DC-10,SouixCity). Limited 3-Dmockupapplyingtoaspecificareainquestion For example,intheAirFlorida 737accidentin1982,where These arealsocommonlydone issue willonceagainsurfaceat thenextICAOAIGmeeting. to helpwithcostsofextraordinary investigations.Iexpectthis Indeed, theNTSBhasproposed toICAOthatmeansbefound surance orafundsupported bygovernmentsoraservicetax. ered extraordinary.Thesesolutions mightincludeworldwidein- could beappliedtomockups, sincetheycouldalsobeconsid- The sameideasthatwereoutlinedtoassistwithrecovery costs Should BeartheBurdenforExtraordinaryInvestigativeCosts?” years ago(ISASIBoston1999),Ipresentedapaperentitled “Who have thefinancialresourcestopayforsuchanunusualeffort. plish amajorreconstruction.Moreimportantly,theymaynot poorer Statesmaynothavethetechnicalcapabilitiestoaccom- terns orcurlingandbendingofparts. easier tovisualize,especiallythoseinvolvingfireorsmoke pat- ization ofmissingpieces.Three-dimensionalrelationshipsare or confirmpotentialcriminalactivity.Itcangiveagoodvisual- it mayallowaninsightthatdidn’texistbefore.Iteliminate help reduceoreliminateoutsidepettheories.Ontheotherhand, will createsightlinesthatcouldprovideotherclues.Thesemay or absenceofcauses,suchaspenetrationmissileimpact.It that nootherinvestigativetooloffers.Itcanshowthepresence millions ofdollars. dollars. Theoverallcostforamajormockupcanrunintoseveral reconstruction, theframealonecancosttensofthousands database totrackthepartsbeinghungonframe.Onamajor age theprocess.Further,significanteffortwillbeexpendedona formation ofaseparate“reconstructiongroup”tostaffandman- tion andpreserved.Somelarge3-Dmockupsmayrequirethe cause ofthelengthtimemockupwillbeunderconstruc- and manpower.Thephysicalfacilitieswillbeinuselongerbe- The costsriseastronomicallybecauseofthedemandsforspace is notforthefeintofheart,anditcomeswithpoliticalovertones. tion, dependingontheirextent.However,goingtoa3-Dmockup 3-D mockups— mockup, ifneeded. two surfaces.A2-Dmockupcaneventuallybeconvertedtoa3-D though fordetailedlayoutsseparateareaswouldbeusedthe laying outtheupperandlowersurfacesinsamearea,al- remove otherevidence.Further,itwillprovideextraspacefor will notruboneanotheranddamagethefracturesurfacesor moving thepartsintoposition.Itwillalsoensurethattornedges tween thepiecestovisualize/examinethemaswellfacilitate for thelayoutbyupto20percent.Thisallowsyouwalkbe- tional spacebetweenthepiecesbyenlargingallotted up” theareabeinglookedat.Scaling-upmeansprovidingaddi- will helpwiththevisualization.Anothertrickwouldbeto“scale outline oftheareatobemockedup,puttingdownmaskingtape chairmen. proceed wasmadebytheIICafterconsultingwithhisgroup vestigation, itwasaprudentfirststeptotake.Thedecision accident andtheunknownsituationinearlydaysofin- and thiseffortwashalted.However,giventhehighprofileof such astherecordersbegantosupercedeneedforalayout separated. Eventually,otherinformationfromtheinvestigation One factstands outloudandclearinmydata collecting.Inmy What istheanswertothiscostsandresourcesdilemma?Some However, notallareasoftheworldarecreatedequal.Some On thepositiveside,a3-Dmockuphasdistinctadvantages Once thetapemeasureandchalkhavebeenusedtomark These canbetheultimateinphysicalreconstruc- ISASI 2003 Proceedings • 35 ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS • examples wherelargemockupswereused. tions. Itisfarfromacomprehensivelistbutratherpresentedas As oneofmyexpertssaid,“Theyarehelltodust.” expensive edificeoncetheinvestigationisnolongernecessary? are onlypartoftheconsideration.Whatwillyoudowiththis be moreproductiveforotheruses.Cost,space,andmanpower standing iftheplanneedstobealteredorreversed. gation. Atthesametime,theremustbeflexibilityandunder- mined aheadoftimebytherightpeopleinvolvedininvesti- how muchoftheairplaneyouREALLYneedshouldbedeter- investigation changes.Thedecisionprocessofwheretostopand it isdifficulttostopthemomentum,evenifthrustof More aboutthislater. mands, besidestechnical,thatmayswayadecisionformockup. and appreciate—therecanbegoodvalidreasonsde- nition, budget,ormanpower.Thisisimportanttounderstand political supportfortheinvestigativeagencyinaquestrecog- for themediaandpoliticians.Itcanprovidebothpublic “sexy” fromthemediastandpointandmakesexcellentfodder investigation. Rathertheyareimportantforshowandtell.Itis think thatanextensivemockupisrarelyrequiredforatechnical keep themsimple,safe,anduncluttered.Myexpertsseemto add muchtothetechnicalunderstanding,anditisdifficult search forinputs,Ifoundthatlarge3-Dmockupsoftendonot 36 • • • and digitizethe wreckagethroughaphotographic orlaserpro- Consider this:wetakeourdigitizing devicestotheaccidentsite mockup, bypassingtheneed foraphysical2-Dor3-Dmockup. job—because oftheireffortthe futureishere,rightnow. their paper;thisisexcitingstuff. TheASChasdoneanamazing simulating theinflightbreakup sequence.Playcloseattentionto integrating severalsetsofdatatoaccomplishothertaskssuch as in theCI611747accident.Itwillalsodiscusscapabilities of Software ReconstructionandPresentationSystem(3D-SWRPS) wan. ItwillcovertheuseoftheirimpressiveThreeDimensional pared bypeoplefromtheAviationSafetyCouncil(ASC)of Tai- ensure absolutedatasecurity. ability tomanipulatedigitizeddataandtheneedforsystems that but itscostandprosconsarestilltobedetermined. may bemissing.Itprovidesanotheroptionfortheinvestigator, is goodforcataloguingtherescuedpiecesanddeterminingwhat mockup istypicallybeingdoneafterthe3-Dinplace.It understand whatiswaitinginthewings.Atthistime,avirtual of thesoftwareisstillbeingdeveloped,butinvestigatorsneedto currently usedtodesignpartsormanufacturingprocesses.Much mockup, alongthelineofcomputer-aideddesign(CADprograms) tizing objectsimproves,thereisagrowinginterestinthevirtual Virtual mockups • • Below isalistofsomeaccidentsthatinvolved3-Dreconstruc- Quite often,thefloorspaceoccupiedbya3-Dmockupcould Along withthis,itseemsthatonceamajormockupgetsstarted, Eventually, itwillbepossible tojumpdirectlyavirtual During thisconferenceyouwillhearanexcellentpaperpre- One ofthepossibledownsidesthisnewtechnologyis the Swissair 111,MD United SiouxCity TW Air F V P • aluJet, DC an AmLock ISA A 800,747(centerfuselage/centerwingtank) SI 2003 rance, Concor P -9 (car r oceedings erbie, 747(centerfuselage) —As computerpowergrowsandmethodsofdigi- -11 (for , DC go compartment) de (fueltanksandwings) -10 (centerengine/tail) war d fuselage) Canada). Jig/frame forSwissairMD-113-Dmockup(photocourtesyTSB eas oftheairplane • • for addinginteriors,carpets, seats,systems,etc.,ifnecessary • • • visualization, analysis,andaccess out largerthantheoriginaldesigntoallowroomforinteriors, • • • • • Let’s summarizesomeoftheresourcesrequired. nation madethattherearenobetterwaystousetheseresources. time, cost,andmanpowerneedtobeunderstoodadetermi- investigative teamasprioritiesduringtheplanningstages.The estimated byinvestigativeteams. age building.Thesearemajordrivingfactorsandusuallyunder- obtaining budget,anditcouldevenplayaroleinpersonalim- their work.IthasPRvalue.Thegrand3-Dmockupvaluein be adesireonthepartofinvestigativeagenciestoshowoff whether tobuildamockupandhowfartakeit.Thereseems pressures fromthepublic,families,andmediaallplayrolesin for technicalinformation,butitseemstobetrue.Politicsand to hearthattheprocesscanbedrivenbyanythingbutneed especially a3-Dmockup.Asinvestigators,weprobablydon’tlike many non-technicalreasonsforbuildingsomeformofmockup, to theframeseparately. ture surfaces,burnpatterns,andotherfeaturesapplythem Further itmaybepossibletodoverycloseupdigitizingoffrac- place, thuseliminatingtheneedofbringingthemtosurface. der waterusingremotelyoperatedvehicles,digitizingpartsin and digitized.Eventually,thetechnologymayallowustogoun- land-based accidentsbecausethepiecescanbeeasilyaccessed to avirtualframe.Thetechniquewillfirstbecomecommonwith are thenmanipulatedontoa3-Dviewofeachpartandapplied cess bytakingmultipleviewsofeachpiecewreckage.These All ofthesemaybevalidreasons—iftheyareacceptedbythe Let’s notfoolourselves.Asmentionedpreviously,thereare Multiple mock Knowledgeable, pr A safephysicalframe(3D)designed forthejobandwithspace Management support T Consider an“e Labor Money Hangar spaceandr Additional partsmaybeneeded ime , andlotsofit ups (2Dor3D)tocoverseparate systemsorar- xpanded” mock ofessional help(saytobuilda goodframe) oofing needs up—wher e youmak e thelay - rules anddecisionprocessleadingtoa“go”or“nogo”ofrecon- organization withanexistingformalprocesssettingouttheground construction decisionprocess.Inmysearch,Icouldnotfindany This shouldalsoinclude,asadistinctandseparateitem,there- the peopleandtheirassignments,budget,travel,research. management oversightincludestheresourcesandtheirallocation, of theoverallmanagementaninvestigation.Typically,project The decisionprocessforconstructingamockupisjustsmallpart Summary expensive whiteelephantlookingforahome. do withit?”Itcanbecomeamonument,museumpiece,oran technical value.Noonehasthoughtabout“whatarewegoingto they cantakeonlivesoftheirownbeyondtheconsideration ceeds technicalneeds.Finally,whenthemockupsarecomplete, or changedirection.Theextentofthereconstructionsoftenex- sion hasbeenmadetoconstructamockup,itisdifficultstop well-thought-out, formalplansforthefullprocess.Onceadeci- cupy. Iwassurprisedtodiscoverinmyresearchthatthereareno money, andeffortspenttocreatethemthespacetheyoc- hoped, especially3-Dmockups,whenconsideringthetime, retribution. sion tohalttheprocesswouldbeacceptableatanypointwithout duce valuableresults.(Itwould“solve”theaccident.)Thedeci- a simple,2-Dor3-Dmockupwouldprovetobewiseandpro- lic pressureswouldbeminimal.Thedecisiontomoveaheadwith investigators. Economicswouldnotplayarole.Politicsandpub- be technicallybasedandmadeprimarilybythetechnicalteamof a designthatallowsittoberelocated • But weliveinarealworld.Mockupsmaynotbeasvaluable So, wherearewe? Know wher e themock

In theidealworld,decisionprocesswould up’s finalr esting placewillbe—andhave ISASI seminar,wecanlearnfromeachother. quickly thanwecantrackindividually.Thankstovenueslikethis onstrates, thisisarapidlychangingworldprogressingfarmore that ultimatelymaybeunnecessary.AstheASCpresentationdem- before launchingoffonanexpensiveandtime-consumingproject tempt togetinvestigatorsthinkaboutwheretheyaregoing reconstruction decisionprocessbutratherasonestepinanat- over theworld.Itshouldnotbetakenasultimatewordon by generousinputsfromnumerousexpertinvestigatorsall garnered over45yearsofaviationexperienceandsupplemented • • • and doitinalogical,well-plannedway • requirements or,atleast,understandthem • • one facets oftheinvestigationproceed;itmayturnoutyoudon’tneed • accomplish you canreasonablyexpecttoget),andwhataretrying • • earlier, sometimesmockupsseemto—justhappen. somewhat haphazardlyintheheatofbattle.Asmentioned cess, handeddowninformallyovertheyearsandaccomplished struction. Mostly,Ifounditwasa“seatofthepants”decisionpro- The precedingdiscussionhasbeenadistillationofmythoughts The bottomlineisthesame:Haveaprocessandfollowit. To improvetheprocess,considerfollowing: Have anendpointinmind Have aplanandstickwithit A P Attempt tominimizetheeffectsofpolitics,cost,time,andspace Involve alltheinter Consider delayingthemock Think aboutwhatyouhave,need(andofthat It isamajordecision;actaccor r void makingitanopen-endedr oceed upthechain,startingwithabasic,2- ested partiesinthedecisionpr up decisionpr dingly esear . ch pr ISASI 2003 N oject ocess whileother D construction Proceedings ocess • 37 ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS 38 Accident sitescould beanywhere,insidean airport,onamoun- I. Introduction issues willbediscussedinthis paper. the associatedcostsarealsohigher thanthelandoperation;both plex thanthelandoperationin theaccidentinvestigationprocess, underwater wreckagerecovery operationoftenisfarmorecom- as welllessonslearnedfromthosetwooperations.Since the the wreckageandrecorderrecoveryoperationsforbothaccidents, recovery ofrecorders,werequitedifferent.Thispaperpresents covery operationsandenvironmentalconditions,including the around thesameproximityofPenghuIslands,wreckage re- able causalfactorsareidentified.Althoughbothaccidentsoccurred tigations ofthosetwoaccidentsarestillunderwayandno prob- of Franceasthesupportingparty.Attimethiswriting, inves- tigation, BEAofFranceistheaccreditedrepresentativewith ATR porting partiestotheNTSB.IncaseofGE791accident inves- ited representativewithBoeingandPrattWhitneyasthesup- two accidents.InthecaseofCI611,U.S.NTSBisaccred- vestigation agencyoftheROC,isinvestigationauthority flight datarecorderandcockpitvoicerecorder. mately 60meters.Bothaccidentaircraftwereequippedwitha an areaof160metersby260withthedepthapproxi- perished fromtheaccident.Thewreckagesweredistributedwithin 10 nmsouthwestofMakung,PenghuIsland.Bothcrewmembers members onboard,crashedintotheoceanlocatedapproximately TransAsia FlightGE791,anATR72cargoaircraft,withtwocrew- 21, 2002,approximately0137Taipeilocaltime(UTC0937), (230 ft).All225peopleonboardFlightCI611perished.OnDec. nautical mileswithanaverageoceandepthofabout70meters west fromMakungonPenghuproper,coveredanarea30square Islands. Thecrashedsite,locatedapproximately15nmnorth- inflight breakupandcrashedintotheTaiwanStraitnearPenghu enroute toHongKong.Itwaslaterfoundthattheaircrafthadan from theATCradarscreen.TheCI611departedTaipei with 209passengersand16crewmembersonboard,vanished 2307), ChinaAirlinesFlightCI611,aBoeing747-200aircraft, On May25,2002,approximately1507Taipeilocaltime(UTC Abstract The AviationSafetyCouncil,anindependentgovernmentin- • Recovery Operations—Comparisons ISA SI 2003 P r oceedings man ofCI611AccidentInvestigation. Taiwan, andtheWreckageRecoveryGroupChair- David Lee By DavidLee CI611 andGE791Wreckage isanAviationSafetyInvestigator,ASC, And LessonsLearned *1 , StevenSu *2 , andKayYong important thing atthattimewastheplanning oftheunderwater positions fromTaiwan’sCoastGuardwereavailable. Themost radar datafromTaipeiAreaControlCenterand floatingdebris part oftheinvestigationteam. Whentheaccidenthappened, NTSB, Taiwan’sCAA,andtheoperator, ChinaAirlines,tobe age debris.ASCcalledtheState oftheaircraftmanufacturer, started torescuethefloatingvictims andrecoveredfloatingwreck- experience forthiskindofaccident.InitiallytheCoastGuard people inthewater.Wehavelimitedknowledgeandalmostno undergone asearch-and-rescueoperationformorethan 200 ter recoveryintheocean.ThiswasalsofirsttimeTaiwanhad CI611 wasthefirstaccidentASCdealtwithregardingunderwa- 2.2 Factorsofunderwaterrecoveryplanning is adversetoanunderwaterrecoveryoperation. to 6meters.Generally,theenvironmentinwinteratTaiwanStrait coral reefsontheseabed.Thewaveheightvariedfrom2meters knots. Thedepthofthewaterisabout60meters.Therearemany knots, guststo55knots.Thecurrentspeedvariedfrom37 wreckage recoveryperiod,thewindspeedvariedfrom13to33 the weatherandmarinemeteorologyarepoor.DuringGE791 From OctobertoFebruary,theTaiwanStraitisusuallyverywindy; accident siteisabout62kilometersfromtheCI611site. GE791: friendly tounderwaterrecoveryoperation. speaking, theenvironmentinsummerandfallatTaiwanStraitis meters to70meters.Theseabedissandyandflat.Generally speed variedfrom2to5knots.Thedepthofthewateris50 4 days,butitdidnotseriouslyaffecttheoperation.Thecurrent Each typhoon’svisitinterruptedtherecoveryoperationforabout The underwaterrecoveryteamalsoencounteredtwotyphoons. period, thewindspeedvariedfrom8-21knots,guststo40knots. for thevisitingtyphoon.DuringCI611wreckagerecovery meteorology atTaiwanStraitinsummerisusuallygood,except the weatherinTaiwanisgettinghot.Theandmarine CI611: 2.1 Weatherandmarinemeteorology II. Underwaterrecovery meteorology arequitedifferent. ter. Theaircraftsize,breakupsituation,weathe,andmarine is aboutsevenmonths,oneinsummerandtheotherwin- accident onDec.21,2002.Thetimebetweenthesetwoaccidents accidents aretheCI611accidentonMay25,2002,andGE791 and twoPart121aircraftaccidentsinTaiwanStrait.These sites onwater,whichincludetwohelicopteraccidentsinrivers existence. Thereare21accidentsitesonlandandfour tain, oratsea.ASChasinvestigated25occurrencesin5yearsof **, Aviation SafetyCouncil,Taiwan,ROC TheCI611accidentoccurredonMay25,2002.In TheGE791accidentoccurredonDec.21,2002. fective distance ofthepingerreceiverinhand. Thepingerre- dia, andtheenvironmentalcondition. Wemaynotknowtheef- sitivity ofthereceiver,output ofthepinger,conductingme- pinger signalfrom1nmaway. Thedistancedependsonthesen- detected bythepingerreceive r. Agoodreceivermaydetect the in sand,theoutputwouldbe attenuated. TheULBsignalcanbe However, ifthepingerwasinsand orpartofthepingerwas sand orsoil.Usuallythestandardoutputisabout160db mersed inthewater.Thesignalisconductive,butnot 37.5khz supersonicsonareverysecondasthepingerwas im- is installedwiththeflightrecorder.Thepingerwouldtransmit porting vessels.Theunderwaterlocatebeacon(ULBorpinger) Recorder searchrequiresproperdetectionequipmentand sup- 2.3 Recordersearchandrecovery adding costtotherecoveryoperation. ally requiresmorepowerfulvesselsandequipment,therefore tions arefavorable.Badweatherormeteorologyconditionsusu- ment andvesselcanbeoperatedonlywhentheweathercondi- rology alwaysaffectunderwaterrecovery.Usuallythesurveyequip- Weather andmarinemeteorology: can helptoverifythewreckage. ery. AnROVwithhigh-intensitylightandcolorvideocamera mote operatedvehicle(ROV)isnecessaryforunderwaterrecov- used tocheckthedepthofwaterwithhighresolution.Are- area withhighresolution.Multi-beamsonardetectionwasalso technology couldsurveyalargeareawithlowresolutionorsmall seabed. Dependingontheequipmentandsurveymethod,this Equipment: the searchplanwasmodifiedaccordingly. Updated sonar/ROVdata:Astheunderwaterwreckagewasfound, become lessandeffectiveforreferencewhentimepassedby. position offloatingdebris.Theobjectswould and directionatthetimeofaccidentwhenwewantedtouse affected bythecurrentandtide.Weconsideredspeed was notdifficulttofindthefloatingdebrisandoiltrail,which Floating debrisandoiltrail: dar wasalsousedtodetecttheairbornedebris. data intoamastermap.Inadditional,thedopplerweatherra- for theintegrationofradardata,salvageposition,andflight radar trackprecisely,theWGS-84coordinatesystemwasselected, viation, andaltitudeofradarstation.Inordertocalculatethe the primaryradardataaretoobtainposition,magneticde- evant signalinthesky.Themostimportantfactorsprocessing target; thesetargetscouldbetheinflightbreakupcluesorirrel- the azimuthangleandslantrangetodeterminepositionof Mode-C altitude,andposition.SSRdatawereusedtocompute Only theSSRcouldtrackaircraftwithtranspondercode, signal returns(sometimestheyaredenotedasPSRandSSR). In fact,therawradardatacouldbeprimaryorsecondary tude, longitude,mode-Caltitude,groundspeed,andtrackangle. of missingaircraft.Thecontentsradardataincludetime,lati- Radar data: considered duringtheplanningprocess: recorders andwreckagerecoveryplan.Thefollowingfactorswere mary referencepoint,andwecontinuedtodeveloptheflight tion obtainedfromthesecondaryradardatawasusedasapri- sible wreckagedistributionpattern.Thelasttransponderposi- recovery. Wecalculatedtheradartrackandestimatedpos- Radardataarethemostusefulrecordforsearch Side-scansonarwasusedtodepictthepictureof After theaccidenthappened,it Weatherandmarinemeteo- ? Pa@1m. time foreverydive. Itwasalsounreliable,because thediverswere surface. Althoughthenoiseeffect wasmuchbetter,ittooklonger tried todetectthesignalbydiving 10metersunderthewater’s Since itwasverydifficulttouse thepingeronwatersurface,we fortable. Thewindnoisedirectly interferedwithhumanhearing. rocky current,andtheworkers onboardbecameveryuncom- we shutdowntheboatengine. Theboatwasthenmovedbythe That inducednoiseonthepingerreceiver.Toreducenoise, and thereforemovedthesonardrumoutofwaterfrequentl y. teorology, the4-meter-highwavemovedboatupand down submerged inwaterabout1-2metersdeep.Inbadmarine me- very useful.Usuallythesonardrumofpingerreceiver was environmental conditions,thecommonpingerreceiverswere not and marinemeteorologywereverybadalmosteveryday.In bad GE791 crashedinTaiwanStraitonDecember21.Theweather able. Werecoveredbothrecorders25daysaftertheaccident. The tember, theweatherandmarinemeteorologywasgeneralfavor- international investigationparties.Fortunately,fromMaytoSep- nology, wehadgreatassistancefromournationalresourcesand difficulties. Althoughwewerenotfamiliarwithunderwatertech- tered notonlytechnicalproblemsbutalsoresourcecoordination CI611, ASChadnounderwaterrecoveryexperience.Weencoun- search operationwereexperienceandmarinemeteorology.In finding therecorderatvicinityofpinger. finding therecorder.However,thereisahigherprobabilityof recorder. Weshouldunderstandfindingthepingerdoesnotequal vessel withdynamicpositingsystemarevitaltotheoperation. ter isdeeperthan50meters,saturationdivingandaworking difficult tofindthepingerwithnon-saturationdiving.Ifwa- searching. Whenthewaterislessthan50metersdeep,it’snot few hundredsmetersoftheprobablepositionwithwatersurface cate thepingerasclosewecan.Usuallyiswithina signal aredetected,wecanusethetriangulationmethodtolo- 5. Underwatersearching:Whenseveralpositionsofthepinger mental noise. possible, theengineshouldbeshutdowntoreduceenviron- hearing, theenvironmentalnoiseshouldbeaslowpossible.If operation isessential.Becausethesignaldetectedbyhuman detect thesignalbyhumanhearing.Experienceindetection 4. Watersurfacesearch:Thecommonpingerreceiverisusedto or againstthecurrent,ratherthanacrosscurrent. on themarinemeteorology;supportingvesselshouldgowith current toshiftthegridrelatedreferencepoint.Depending X 10km.Sometimeswemayconsidertheradartrack,wind,and as thecenteringrid.Thesearchareamaybesetabout10km ceiver andreferencepoint,wecouldmaketheposition 3. Plansearcharea:Oncewehavethedistanceofpingerre- sponder returnisagoodreference. 2. Selectreferencepoint:Usuallythepositionoflasttran- of thepingerreceiverordetectionequipment. 1. Equipmenttest:Useapingertotestthefunctionanddistance plified intofivesteps. to moveandstartshutdowntheengine. Under goodmarinemeteorologyconditions,asmallboatiseasy for planning.Thesupportingvesselshouldnottobetoobig. ceiver shouldbetestedwithapingerinsimilarwatercondition The majordifferencesbetweentheCI611andGE791recorder If theimpactforcewashigh,pingermayseparatefrom The approachinsearchingforthepingersignalcouldbesim- ISASI 2003 Proceedings • 39 ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS size andmaterial reflectdifferentintensity. Comparingthedif- nar andreceivesthereflection fromtarget.Targetswithdifferent relies onsonardetection.Detection equipmenttransmitstheso- Underwater sitesurvey: 2.4 Wreckagesurveyandrecovery recovery ****Diving neverlaunchedduetohighcurrentinGE791recorder *** ELACUT2000isaunderwatercommunicationsystem. equipment requiredstringentcontrol. **Fifty metersisathresholdfordiving,alldivers,procedure,and *Diving requirescurrentlessthan2knots. Recovery Comparisons. Table 1:CI611andGE791RecorderSearchingand parison betweentheCI611andGE791recordersearch. pinger signalbywatersurfacesearching.Table1showsthecom- pinger receiver.After20daysofsearching,wefoundonlyone tually thepingersignalwasfoundandthenconfirmedby noise andgainedaprecisepositionofthedetectionpoint.Even- system to37.5khz.Withthisapproach,weisolatedtheoutside adjusted thecarrierfrequencyofunderwatercommunication multi-beam sonar,ROVandadynamicpositioningsystem.We vessel, whichhasmultiple-functions,includingside-scansonar, The underwatercommunicationsystemisbuiltintheresearch use theunderwatercommunicationsystemtofindpinger. minutes inwater).Tosolvethisdifficulty,thesearchteamtriedto current wouldmovethedivers1nmfrominitiallocationafter20 moved bythecurrentandcouldnotstayatafixedlocation(the 40 • ISA SI 2003 P r oceedings Underwaterwreckagesearch primarily tion arethelack ofaprecisewreckageposition andalackof than ROVsanddivers.The disadvantages ofatrawlingopera- wreckage canbepickedup by trawling.Thecostismuchless may stillbesomewreckageto berecovered.Thesmallpiecesof Trawling operation: make properplanningtoreduce anyrisk. gerous totheworker;payattention tothewindandcurrent very easytodamagethewreckageagain.Theprocessisalso dan- tion. Duringthetransportation,especiallyfromshiptoship, itis transported toatemporaryareaorhangarforfurtherinspec- how towrapuplargewreckagepieces.Allneeds tobe required toassistthedivercheckpositionsofwreckage and confirmed beforemoving.Theengineersfrommanufactures are tant positionssuchasthelock,latch,andjackscrewneed tobe age needstobedocumentedbyvideothoroughly.Someimpor- pieces requireacrane.Beforemovinganywreckage,thewreck- an ROVmaypickupthesmallerwreckages.Thelargewreckage Wreckage recovery: CI611 andGE791. Table 2:WreckageSiteSurveyComparisonBetween CI611 andGE791. wreckage fromthemonitor. manufacturer andoperatorarerequiredtoidentifythepieceof high-intensity, colorcameraisrequired.Structureengineersfrom are installedonanROV.Duetolowvisibilityindeeperwater,the 5. Visualcheckusingunderwatercamera:Usuallythecameras know thepositionofwreckagehigheraccuracy. depict theseabedcontourin1-meterresolution,butweneedto in thisarea.Sometimesthemulti-beamsonarcouldbeusedto age isfound,wemayusethehighestresolutionside-scansonar 4. Finescanning:Afterthehighestconcentrationofwreck- sonar tocoveralargeareainshorttime. 3. Coarsescanning:Uselow-resolutionandwide-rangeside-scan higher resolutionequipmentinthemostprobableaccidentarea. scan speedofthedetectionequipment.Weshouldassign 2. Plansurveyarea:Considerthescanrange,resolutionand erence. tected/recovered, wecouldselectmanypointsforplanningref- positions fromprimaryradardata,oiltrail,andwreckagesde- transponder returnasthereferencepoint.Dependingon 1. Selectreferencepoint:Initiallyweusedthepositionoflast for largeareasitesurvey. ward scansonar.Theside-scansonarsystemiscommonlyused bed contourincludeside-scansonar,multi-beamandfor- pect wreckage.Thecommonsonartoolsusedtosurveythesea- ference tobackground,asonarspecialistcouldidentifythesus- Table 2showsthewreckagesitesurveycomparisonbetween Wreckage searchcouldbesimplifiedintofivesteps. Whenallfoundwreckageisrecovered, there After checkingwreckagevisually,diversor by anROV. enough, whichresultedinlow efficiencyofwreckagevisualcheck sitioning systemofside-scan sonarsystemwasnotaccurate good foralongperiodtimeafter theaccident.However,po- costly. Fortunately,theweather andmarinemeteorologywere for wreckageandrecorderrecovery ofthisaccidentarehugeand This caseisaninflightbreakup accident.Theresourcesrequired two accidentsasthefollowing. From theunderwaterrecoverypointofview,weconcluded these 3.1 Conclusions III. Resultsanddiscussion was animpactbreakupcase. pattern, theCI611wasaninflightbreakupcase,andGE791 the CI611andGE791accidents.Fromwreckagedistribution rent andcontouroftheseabed.Table4showscomparison aircraft impactswater,thedistributioncouldbeaffectedbycur- last altitude,andspeedifitwereaninflightbreakup.Whilethe The wreckagedistributionwouldseriouslybeaffectedbywind, age distributionisverydense,itmayresultfromimpactbreakup. widely distributed,itmayresultfrominflightbreakup.Ifthewreck- sequence, wind,current,altitude,andspeed.Ifthewreckagewas Wreckage distributionwastheresultfromaircraftbreakup 2.5 Comparisonbetweenthetwoaccidents GE791 WreckageRecovery. Table 3:TrawlingOperationComparisonofCI611and age recovery. the trawlingoperationcomparisonofCI611andGE791wreck- area trackingsystem,andtrawlingtoolselection.Table3shows tioning systemofships,shipspositiontrackingsystem,trawling includes trawlingareaplanning,selectingships,posi- against current.Theplanningofatrawlingoperationatleast operation; thetrawlingshipscouldoperateonlywithcurrentor during December,January,andFebruarydidnotallowthecross employed. However,themarinemeteorologyinTaiwanStrait increase thecoveragerate,crosstrawlingindirectioncanbe to recover.Thecoveragerateoftrawlingisdifficultcontrol.To GE791 wasbytrawlingoperation.Onlytheenginesweredifficult tion wasuseful.Mostoftheunderwaterwreckagerecovery the GE791accidentareinsmallerpieces.Thetrawlingopera- large wreckagepieceswithatrawlingoperation.Theof CI611 accidentlargerthan10meters.Itisimpossibletorecover net isabout10meterswide.Therearemanywreckagesofthe rock, thenetmustbecutandleftbehind.Usuallytrawling transfer, etc.Whenthetrawlingnetbecomesstuckinseabed include limitedwreckagesize,trawlingnetand cause secondarydamage.Thelimitationsofatrawlingoperation visual checkbeforerecovery.Thetrawlingoperationmayalso The wreckagedistributionoftheCI611accidentspreadwidely. Figure 1:CI611radardataandwreckagedistributionpattern. Table 4:ComparisonofCI611andGE791accidents. is currentlyproposing aspeciallawforaviation accidentinvesti- cost, thatwouldadverselyaffect theaccidentinvestigation.ASC Taiwan. Iftheoperatorwasnot willingtopayfortherecovery to prescribegovernment,operato r, owneoranyonetopayin the expenseofwreckagerecovery isstillanissue.Therenolaw the efficiencyofvisualcheck by anROV. tioning systemofsidescansonarwasimproved,whichincreased rology wasbadandnotsuitableforwreckagerecovery.Theposi- is animpactbreakupaccident.Theweatherandmarinemeteo- age wasquitesmallexceptenginesandlandinggears.This case Underwater recoveryiscostl y. Whoshallberesponsible for The wreckagedistributionofGE791wasverydense.Allwreck- ISASI 2003 Proceedings • 41 ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS • • underwater recovery. • process, wehavelearnedthefollowinglessons: Through theCI611andGE791accidentunderwaterrecovery 3.2 Lessonslearned sponsibility ofthegovernment. gation thattheexpensesofunderwaterrecoveryshallbere- 42 • reference points. Figure 2:GE791wreckagedistributionpattern. Accuracy ofradartrackplaysamajorr Good planningisamust. W Adequate equipmentisvital. • eather andmarinemeteor ISA SI 2003 P r oceedings ology ar e theprimaryfactorsfor ole inplanninginitial • or IRS). • In wreckagerecovery • more userfriendlytodivers. • • survey, ifweatherisbad. • survey. • In searchofrecorders ter recovery,wehavethefollowingrecommendations: To anyagencythatmaytakepartinaviationaccidentunderwa- 3.3 Recommendations ** *2 *1 Footnotes ment areamust. • efficiency ofrecovery. • impact breakup,flightpath,speed,wind,andcurrent. • age recovery. • • ManagingDirector InvestigationLab/Chief Investigator P P Use pingerr F P Use lar Contour ofseabedaffectsther P W Floating wr Site surveybefor Use smallandquietboatwithpingerr or lar owerful RO ositioning systemofS r inger r r ecise side-scansonarsurveyisr eckage distributionpatterncanbeaffectedbyin-air-br ge wr ge vesselwithequipmentsimilartoUT2000forsurface eceiver withboneconductionphoneandcompassis eckage haslesssignificanceinunder eckage r eceiver withanRO V isr N e r equir ecovery isvitalbutmaynotbeaccurate. ecovery S ed ifcur S, RO , skillfuldiverswithpr ecognition ofwr V , diving,andvesselsaffectsthe r V whendivingisnotallowed. ent ishigh. equir ed (towfishwithbeacon eceiver forsurface eckage bysonar water wr oper equip- eak eck- up, . factual evidencecollectionand thefollowingtaskofanalysis. [3], wreckagereconstructionbecomes animportantmethodtohelp (TWA800, 747-141)[2],orSwissair Flight111(SR111,MD-11) Flight 103(PA103,747-100)[1], TransWorldAirlinesFlight800 however, foraninflightbreakup accidentsuchasPanAmerican probable causes,andproposingsafetyrecommendations,etc. ering relevantfactualdata,draftinganalyticaltopics,finding out searching forandsubsequentreadoutoftheflightrecorders, gath- When anaircraftaccidenthappens,investigationbeginson scene: I. Introduction tant roleinsupportingfutureaccidentinvestigation. quence. Itisbelievedthatthistechnologywouldplayanimpor- simulation programtoassistintheanalysisofbreakup se- bine radarreturnsignal,wreckagesalvagedata,andaballistic gation ofstressbreakup.Inaddition,the3D-SWRPScancom- cess, andcanbeusedtodeterminefracturebehaviorpropa- It canprovidesub-centimeteraccuracyinthereconstructionpro- a genericCATIAengineeringmodelofthesametypeaircraft. covered, structureframeofaBoeing747-200cargoaircraft,and puter 3-D-graphictechniques,laserscanningofwreckagesre- determined. The3D-SWRPSutilizesacombinationofthecom- going, andprobablecausesofthisaccidenthavenotyetbeen dent. Atthemomentofthiswriting,investigationisstillon- serious incidentinvestigation,hasbeeninvestigatingtheacci- ment organizationresponsibleforallcivilaircraftaccidentand accident. TheAviationSafetyCouncil,anindependentgovern- wan Strait.All225peopleonboardtheaircraftperishedinthis sengers. TheaccidentoccurrednearPenghuIslandintheTai- aircraft wasaB-747-200carrying19crewmembersand206pas- breakup accidentthatoccurredonMay25,2002.The port theinvestigationofChinaAirlinesFlightCI611inflight sentation system(3D-SWRPS).Thiswasdevelopedtosup- plication ofathree-dimensional-software-reconstructionandpre- The purposeofthispaperistopresentamethodologyandap- Abstract Wreckage ReconstructionTechnology There areseveral relevantapplicationsassociated withwreck- The aboveworkprocessisfamiliar toeveryoneinthisfield; At theAircraftAccidentInvestigation Application ofthe3-DSoftware mance Groupmember,CI611AccidentInvestigation. laboratory, ASC,Taiwan,andRecorders&Perfor- Victor Liang Prepared byWen-Lin,Guan* isanengineerintheinvestigation Aviation SafetyCouncil.PresentedbyVictorLiang. 1 ,VictorLiang* mational helpful inthedeterminationofcausal factors.Inthe breakup accident,wreckagereconstruction becomesveryinfor- water recoveryoftherecorders andwreckages.Foraninflight and logisticisanoverwaterinvestigation, whichrequiresunder- tors, systems,andlogistics. trol, weather,airport,maintenance, survivalfactors,humanfac- clude groupsofflightoperations,recorders,airtraffic con- verity oftheoccurrence.Atypicalinvestigationteamshould in- verification, analysis,findings,andsafetyrecommendations. six phases:on-sceneinvestigation,factualdatacollection, tion ofsimilaroccurrencesfromhappeningagain. factors, andforproposingsafetyrecommendationsthepreven- dictates theinvestigationproceduresfordeterminationof causal information system,remotesensing),etc.ICAOAnnex13generally recovery, spatialremotesensing(globalpositionsystem,geographic space, avionics,humanfactor,flightoperation,weather,underwater Aviation accidentinvestigationintegratesthetechnologiesofaero- 2.1 Characteristicsofaviationaccidentinvestigation wreckage reconstruction II. Aviationaccidentinvestigationand of wreckageandterrain. or insideanairport,todeterminethedistributionalrelationship eas ofthree-dimensionalsitesurvey,insecludedmountainarea, here canalsobeusedforfutureaccidentinvestigationinthear- semble themintocompletea3-Dmodel. these planeswithselectedreferencepointsofalignmenttoas- scanner atanobjecttomeasureitstangentplanes,thenaligns so-called 3-Dsurveyingtechnologyaimsthepreciselaser the efficiencyandcostsavingsofwreckagereconstruction.This surveying technologyprovidesanevenbetterwaytopromote Concorde’s phenomenonofcombustiblestabilization[6]. searches attheUniversityofLeedsappliedCFDtoanalyze gears andsupportstructures[4][5]. and EMB-145,FEAwasusedtoexaminethestressoflanding lators. Forexample,inhardlandinginvestigationsforaMD-11 chanics interfaceandflightcontrolsbyengineeringsimu- (CFD), forverificationofflowfields;evaluationhuman-me- of structurestressandmetalfatigue;computationalfluiddynamics age reconstruction:finiteelementanalysis(FEA),fortheresearch The mostdifficultaccidentinvestigation intermsofbudget Size oftheinvestigationteamdependsonnatureand se- The investigationofanaviationaccidentasawholecontains The methodologyandapplicationof3D-SWRPSpresented Thanks totheadvancedtechnology,applicationof3-D After theAirFranceConcordeaccidentin2000,academicre- 2 ,PhilTai* 3 ,andKayYong**, ISASI 2003 Proceedings • 43 ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS oramic images oftheflightdeck,whichprovide wreckagein- reconstruction, theTSBofCanada wasthefirsttoproducepan- electrical wiring.Incontrast to pastaircraftwreckagehardware tion sectionswerethenfocused onthefuselage,flightdeck,and emphasized findingthesource ofsparkandsmoke;reconstruc- smoke problem.”Therefore, wreckagereconstructionofSR111 corder stoppedrecording,flight crewswerediscussing“cabin spread alongthedownwindside.Beforecockpitvoice re- dar dataindicatedthatwreckagesfollowedtheflightpath and was focusedonthefuselageandcentralfueltankssections. sized findingthesourceofexplosion;hence,reconstruction corder. Therefore,wreckagereconstructionofTWA800empha- ity andhigh-energysoundwavesrecordedinthecockpitvoice re- such as“fireballanddescending”appeared.Therewasabnormal- path spreadalongthedownwindside.Ininterviews,testimonies mary radardatadisplayindicatedthatwreckagesfollowedthe flight ing conditions,andfracturesurfaces,etc. clude failureconditionsatdifferentsectionsofthefuselage,burn- characteristics ofsalvagedwreckages.Thosein- be radartracks,flightrecorderdata,relatedtestimonies,and during theon-sceneinvestigationphase.Thesecluescanusually be madeaccordingtocluesandfactualinformationcollected wreckage reconstructionisanimportantissue.Adecisionshould 8) Safetyconcernsofpersonnelatwork,etc. cut upandcraneoperation;7)Accessibilitytothemockup; age reconstruction;5)Designofframeormockup;6)Wreckage wreckage reconstruction;4)2-Dlayoutor3-Dwreck- 2) Identificationandtaggingofwreckages;3)Partialorwhole wreckage reconstruction:1)Evaluationofthereconstructionsite; or forcebetweenmainstructures. condition offorces;andthird,tostudythepropagationstress out sourceofstructurefailure;second,tojudgetheendurance did notrequirereconstructionofthewreckages. by groundcollisionmarksandwreckagedistribution,hence Oct. 31,2000,itsstructurefailuresequencecouldbedetermined runway atChiangKai-ShekInternationalAirportinTaiwanon Singapore AirlinesFlight006(SQ006),whichcrashedonthe dents needwreckagereconstruction.Forexample,inthecaseof nation ofthefailuresequenceisrequired.However,notallacci- able method.Ingeneral,toevaluateprobablecauses,determi- mined, thenreconstructionusingthewreckagecollectedisavi- of forceanddestructivedirectionstructurecouldnotbedeter- and examinethewreckagepiecesonebyone.Whensource sea, aftersalvagingthewreckage,investigatorsneedtoidentify Irrespective ofwhetheranaircraftcrashedonlandorintothe 2.2 Methodsofwreckagereconstruction the determinationoftheirprobablecauses. marized inTable1,wreckagereconstructionswereconductedfor SR111, AirFrance’sConcorde,andChinaAirlinesCI611.Assum- there weresixinflightbreakupaccidents:PA103,AA811,TWA800, struction butwithoutsignificantdevelopment.Inthelast15years aggressively seekingforanefficientmethodofwreckagerecon- past decade,theaviationaccidentinvestigationagencieswere 44 Furthermore, aftertheinflightbreakupofSR111,primary ra- For example,aftertheinflightbreakupofTWA800,pri- The determinationofwhethertoundertakeapartialorwhole Several preparationconsiderationsarerequiredpriortothe Wreckage reconstructionservesthreepurposes:first,tofind • ISA SI 2003 P r oceedings on thereliable operationalsafetyofalaser,itcanachieve preci- OPTECH (modelILRIS)and RIEGL(modelLMSZ420).Based scanner. Table2summarizesthefunctions of3-Dlaserscanners gram toassistanalyzingthebreakup sequence. turn signal,wreckagesalvage data,andballisticsimulationpro- the breakup.Inaddition, 3D-SWRPScancombineradarre- used todeterminefracturebehaviorandpropagationofstress of centimeter accuracyinthereconstructionprocess.Itcanalso be engineering modelofthesametypeaircraft.Itprovides sub- frame ofanidenticalB-747-200aircraft,andagenericCATIA techniques, laserscanningofwreckagesrecovered,thestructure The 3D-SWRPSutilizesacombinationofcomputer3-D-graphic 2.3 Architecturesof3D-SWRPS SWRPS). software wreckagereconstructionandpresentationsystem (3D- TSB inSR111investigation,ASCwentontodevelopthe3-D method fromthetraditionalhardwarereconstruction. to establishSR111wreckagereconstructionisanimproved burning duringthebreakupsequence.The3-DCADmodelused wiring diagrams,TSBwasabletodemonstrateevidenceofwire with wreckagesketches,3-DCADmodels,andrelevantpower spection technologysimilarto“virtualreality.”Inconjunction (b) outermodeloffrontfuselage. cargo aircraft,(a)innermodelofreararea, Figure 2:3-DscanmodelofB-747-200 reconstruction andpresentationsystem. Figure 1:Architectureof3-Dsoftwarewreckage The 3D-SWRPSprojectuseslong-range andprecise3-Dlaser From theexperienceofusing3-DCADmodeldevelopedby may causedata loss. • tion tooltoreducethe3-Dmodel’s size. computer’s memoryallocation, thisstepusesthepolygonreduc- • ner andopticalsystems. nized pointclouds,asproducedbymostplane-of-lightlaserscan- dividual piece,thepiecewasthendigitized.Itprocessesorga- • aircraft wreckagereconstructionthroughfollowing: and modelingprocesstooknearly1month. smaller cargofloorbeampieceswereignored.The3-Dscanning crane handlingtobescanned.Wreckageslessthan1minsizeor eled into3D-SWRPS;amongthem,50wreckagepiecesneeded Taiwan. Intotal,161piecesofwreckageweredigitizedandmod- and modeledatHangarIIofTaoYuanAirForceBase(TAFB), aircraft’s innerandouterbody. ILRIS todothejob.The3-Dscanningwasdonewhole of aChinaAirlinesB-747-200cargoaircraft.ASCselectedthe collected, aBoeing747-200CATIAmodelandscanned 3-D wreckagemodelontothe“referencemodel,”twomodelswere sion to3mmwithmaximumrangeof2km.Inorderalignthe side viewof3-Dmodel. of (A)and(B)arepresentasinnerouter 640 andsideviewof3-Dmodel.Marks Figure 3:Craneoperationalphotoofitem • ing overlappedmeasurements. procedure reducesthenoise in theoriginal3-Ddatabyaverag- from thesametypeofaircraftandBoeing’sCATIAmodel. This 3-D scansofwreckageintoareferencemodel,whichwasobtained tion allowsinvestigatorstoautomaticallymergeasetofaligned • bringing thesescansintothesamecoordinate. different 3-dimensionalcoordinatesystem.Thisstepconsists of the digitizingprocessproducedseveral3-Dscansexpressed ina scanner inordertomeasureallwreckagesurfaces.Asa result, gators needeithertorotatethewreckageormove3-D laser 3D-SWRPS representsadifferentprocessingmethodforthe Sections 44,46,and48oftheCI611wreckagewere3-Dscanned Manually editsurfaces:Especially withunevensurfacesthat Aligning multipledatasets:Duringdigitizingpr 3- Mer P olygon editingandr D objectdigitizing:Oncethelaserscannerscannedeachin- ging multipledatasets:A3- eduction: Inor D -graphic virtualr der tocontr ocess, investi- econstruc- ol the and betweenS-07L andS-11R,respectively. number ofstationsandstringers arebetween1960and2100, the craneoperationandside viewofitem2136.Thereference stringer numberbetweenS-24L andS-50R.Figure4illustrates item 640is260”x200,”station numberbetween1920and2180, 3(B) istheouterviewof3-D modelofitem640.Thesize Figure 3(A)isaninnerviewof the3-Dmodelofitem640;Figure card, and80GBharddrive. station, equippedwitha1024MBmemory,AGP4xgraphics data processingplatformisaPC-basedhi-levelgraphics work- to 70,000innumbers,andfrom3MB120filesize. The wreckage, each3-Dmodelconsistsofpolygonsrangingfrom 30 gons. Accordingtotheconditionsofcrookedandfractured shape. Thebasicelementofa3-Dmodeliscomposed poly- age. Eachpiecerequiresthreetoeightscansdepending onits ASC spent2monthsscanningandmodeling161piecesofwreck- 3.2 ResultsofCI611wreckage3-Dmodel flight deck,andL1-doorframe,etc. of thefuselagemodel,includingregistrationnumber,nosewheel, recorders. TherightsideofFigure2indicatestheouterportion ments ofsection46,floor,duct,andinstallationplatformforflight 200 cargoaircraft’sinnerleftaftfuselage,includingframeseg- tion ofthemodel.Figure2illustrates3-DmodelaB-747- ket wasremoved.ASCspent30hoursscanningtheinnerpor- sure bulkhead,anddoorframes. fuselage, horizontalandverticaltails,innerframe,duct,aftpres- The referencemodelofaB-747-200cargoaircraftincludesnose, 3.1 Resultsofcargoaircraft3-Dmodel III. Resultsanddiscussion include size,station,section,damagephotos,3-Dmodel,etc. wreckage attributesthroughasecuredIntranet.These database oftheSystemsGroup,whereinvestigatorscouldaccess structures. Inaddition,the3-Dwreckagemodelalsolinksto boring wreckagepieces,metalfatigueandstresspropagationof different angles,tofurtherexaminethefractureconditionsofneigh- the 3D-SWRPStoviewinnerandoutersideoffuselagein stringer numberoftheoriginalaircraft. onto thereferencemodel,basedupontheirframestationand ASC investigatorsspent3monthstoalignandattach62pieces 3D-SWRPS. whole as3D-SWRPS.Figure1showsthedetailedarchitecturesof of ASCTaiwandevelopedotherprogramsandintegratedthe TSB ofCanadadevelopedRAPS.TheInvestigationLaboratory The U.S.NTSBdevelopedtheballistictrajectoryprogram. tion, andtherecoveryanalysispresentationsystem(RAPS). trajectory, polywork,multigencreator,polytrans,rationalreduc- in timehistory. listic trajectory,windprofiledata,andwreckage3-Dmodeldata late theinflightbreakupsequencebycombiningradardata,bal- • models fromthedigitizedcolor3-Ddata. • Figure 3showsthecraneoperationandsideviewofitem640. During aD-checkofthecargoaircraft,heatinsulationblan- The resultgivestheinvestigatorscapabilitytointeractwith After all161piecesofwreckagewerescannedandmodeled, The 3D-SWRPSconsistsofsixstand-aloneprograms:ballistic Br T e eak xtur up animation:Amajorfunctionofthismoduleistosimu- e mapping:Investigatorscancr ISASI 2003 eate te xtur Proceedings e-mapped • 45 ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS lage structure. Figure5indicatesthatthesevere damageofthe it showsthattheitembelongs tosection46oftherightaftfuse- ally aligningthispieceontothe referencemodelofaB-747-200, could notbedonejustasinactual hardwarereconstruction. gaps intheconnectingsurfaces ofwreckages,orthealignment even selectionofthereference pointofalignmentcouldcause appropriate selectionofreferencepointalignment.An un- ware reconstructionisthecomputer’smemoryallocationand the ment andaligningsurfaces.Thegreatestdifficultyin3-D soft- fracture surfaceswereselectedfirstasthereferencepointof align- nates ofstationsandstringers.Wreckageswiththeleast-deformed The entire3-Dwreckagemodelisalignedwithreferencecoordi- 3.3 Resultsof3-Dsoftwarereconstruction 46 reference framemodel. Figure 5:3-Dmodelofitem640and and sideviewof3-Dmodel. Figure 4:Craneoperationphotoofitem2136 By usingthereferencecoordinates ofitem640,whenmanu- • ISA SI 2003 P r oceedings results ofmetalfatiguelocatedatstation2100. located atstationof2160(12.5cm);(B)SEM Figure 7:(A)“N”shapecrackofitem640 relevant stationnumbers. Figure 6:Wreckagephotoofitem640and from theright). aft pressurebulkhead(upside-downview 3-D softwarereconstructionatsection46and Figure 8:Comparisonof2-Dlayoutand crack atstation 2160.Figure7(B)showsthe electron microscopic reference model.Figure7(A) indicatesthe“N”shaped12.5cm to markorcomparethedifferences betweenthewreckageand distance, includingcurveangles onthesurfaceofwreckages,then ate ormeasurefracturebehaviors. Itcouldeasilymeasurethearc ducted. Infact,itisveryuseful toadoptthe3D-SWRPSevalu- blue-dotted lineisthearea where SEMexaminationwascon- Figure 6showsthe3-Dmodelsoffuselageanddouble r. The on Figure6.) lar cutisasampleofmetalfatigueexamination.(Detailshown 1920 and1980.BesidethelowerleftofFigure5witharectangu- also illustratessignificantfractureconditionsbetweenstations formed butintactwithrelevantframesofthefuselage.Figure 5 inner viewofitem640;theaftcargo-doorframeisslightly de- shaped fractureexistedatstation2180.Figure5alsoshows the outer blendislocatedatstations2060and2180,alarge “V” Precision 3-DLaserScanners. Table 2:ComparisonofLong-Rangeand for InflightBreakupInvestigationsinRecent15Years. Table 1:AircraftWreckageReconstruction Upper Figure6showstherepairdoublerofitem640.Lower ing withsimulation programforbetteranalysis support. compared tohardwarereconstruction; d)flexibilityincombin- for otheraccidentinvestigation; c)onlyone-halfofthecostas lem; b)reusability,oncedeveloped, themethodologycanbeused termine fracturebehaviorand aircraftbreakuppropagation. vide sub-centimeteraccuracy scan quality,andcanbeusedtode- generic engineeringmodelofthesametypeaircraft.Itcan pro- 3-D-graphic techniques,laserscanningofwreckagepieces, plus 3D-SWRPS wasdevelopedbyutilizingacombinationofcomputer IV. Conclusions tween 3-DhardwareandsoftwarereconstructionofCI611. from theinvestigationpointofview. choice if3-Dhardwarereconstructionisnotreallythatnecessary out togetherwith3-Dsoftwarereconstructionmightbethe best Advantages ofthe3D-SWRPS area)nowreckagedisposalprob- Table 3(page48)summarizesandmakesthecomparisonbe- neers. Useof2-Dwreckagelay- ner, andlaborcostoftwoengi- operation, rentalofa3-Dscan- was US$91,500,includingcrane The costof3D-SWRPSforCI611 US$143,000, onlyforsection46. construction forCI611was ement analysis. and inconjunctionwithfiniteel- trajectory analysisandsimulation, advantages inreusability,ballistic struction. 3D-SWRPShasgreat less thanfor3-Dhardwarerecon- quired for3D-SWRPSismuch power, budget,andspacere- ware reconstruction.Infact,man- generally betterthan3-Dhard- The utilizationof3D-SWRPSis software reconstruction 3.4 Comparisonof3-Dhardwareand outer rightside. pressure bulkheadviewfromthe struction atsection46andtheaft 2-D layoutand3-Dsoftwarerecon- CATIA model.Figure8showsthe age alignedonthereference ing, thereare62piecesofwreck- stringer numbers.Forthetimebe- respective fuselagestationand the hangarflooraccordingtotheir all wreckageswerearrangedon sketching, andinitialexamination, struction. Afterfinishtagging, base hangarforhardwarerecon- tified andplacedattheAirForce wan Strait,whichweretheniden- had beensalvagedfromtheTai- station 2100. shows metalfatiguecrackaround examination ofitem640,which The costof3-Dhardwarere- Totally, 1442piecesofwreckage ISASI 2003 Proceedings N • 47 ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS 48 Hardware andSoftwareReconstruction. Table 3:Functionalcomparisonbetween3-D • ISA SI 2003 P r oceedings ** ManagingDirector *3 Investigator *2 InvestigationLab/Engineer *1 InvestigationLab/DeputyChief [8] AviationAccidentReport,“ExplosiveDecompression—LossofCargoDoor [7] VittorioCasella,“AccuracyAssessmentOfLaserScanningData:ACase [6] J.T.,QinetiQ,“FatalConcordeFireExplained,”LeedsUniversity(CFD [5] FranciscoK.Arakaki,etl.“EMB-145NoseLandingGearDoorStress [4] Har,J.,“MD-11CompositeSpoilerAnalysisUsingMSC/NASTRAN& [3] “AviationSafetyRecommendationsMaterialFlammabilityStandardsIn- [2] AviationAccidentReport,“inflightBreakupOvertheAtlanticOcean [1] AviationAccidentReport,“Boeing747-121,N739PAatLockerbie, References LULU, HAWAIIFebruary24,1989.”NTSBReportNumber:AAR-92-02. in FlightUnitedAirlines811Boeing747-122,N4713UHONO- neering, ViaFerrata,Pavia,Italy,2001. Study,” UniversityOfPavia,DepartmentofBuildingandTerritorialEngi- Center), Leeds,Yorkshire,Hampshire,UK.,2001. ment, 12227-901SaoJosedosCampos—SP—Brazil,1996. Analysis,” EmpresaBrasileiradeAeronautica,StructuralAnalysisDepart- RR-90-011, 1990. PATRAN PLUS,”KoreaInstituteofAeronauticalTechnology(KIAT),KIAT- ber: A98H0003,2001. vestigation IntotheSwissairFlight111Accident.”TSBOccurrenceNum- 2000. Moriches, NewYorkJuly17,1996.”NTSBReportNumber:AAR-00-03, Trans WorldAirlinesFlight800Boeing747-141,N93119nearEast 90 (EW/C1094),1990. Dumfriesshire, Scotlandon21December1988.”AAIBReportNumber:2/ recorders…did notrevealpositiveevidence oftheexplosion official reportintotheaccident conceded“analysisoftheflight related insomewaytothedetonation oftheexplosivedevice, aircraft. Whileitisreasonably inferredthatthe“loudsound”is improvised explosivedeviceled directlytothedestructionof material arisingoutofLockerbie indicatedthatdetonationofan tic clearancefromShanwickATC. Averylargevolumeofforensic The soundoccurredwhilethecrewwascopyingtheirtransatlan- pit areamicrophone(CAM)trackattheendofrecording. cember 1998identifiedaloudsoundlasting170msonthe cock- The AAIBreport[1]onthePanAmLockerbieaccident in De- Introduction to locatetheseatoffailure. delity andthatwouldprovideinvestigatorsinthefuturewith tools systems thatmaybeabletorecordtheseeventswithgreater fi- why thisisso.Weshallalsoputforwardpracticalsuggestions for explosions andstructuralfailuredecompressions,weexplain CVR recordingsaregenerallyunabletodiscriminatebetween aircraft fromtrialsinseveralcountries.Inparticularweshowthat cordings ofexplosionsandrapiddecompressionsonavariety sults. ThispaperreportsonaprograminvestigatingCVRre- dent recordingshavenotbeenabletoproduceconclusivere- major accidentsinthepast30years.Comparisonsbetweenacci- to interpretsuchrecordingshasbeenseverelytestedinseveral events withgreatfidelity,andtheabilityofaccidentinvestigators important. Yettheserecordersarenotdesignedtorecordsuch failure decompressions,therecordingsareseenasevenmore catastrophic lossofanaircraftthroughexplosionsorstructural corder (CVR)andflightdatarecorder.Intheeventofsudden, tion isoftenfocusedonthetworecorders,cockpitvoicere- the immediateaftermathofamajorcivilaircraftaccident.Atten- Rapid identificationofthecausefailureisahighpriorityin Abstract and hastwochildren,Georgina(4)Imogen(2). of Acoustics,andaCharteredEngineer.DyneismarriedtoHelen he isamemberoftheRoyalAeronauticalSociety,FellowInstitute Am Lockerbieaccidentin1988.InadditiontohisISASImembership, ISVR arisingoutofarecommendationintheAAIBreportonPan He firstbecameinvolvedinairsafetybyleadingtheresearchactivityat undergraduates, post-graduates,andprofessionaldevelopmentcourses. Structural FailureDecompressions CVR RecordingsofExplosionsand By StuartDyne(MO4779),Manager,ISVRConsulting,InstituteofSoundandVibrationResearch, a universitylecturer,healsocontinuestoteach shock andvibration,signalprocessing.Formerly the UniversityofOxfordandworksinacoustics, in theUK.HehasaB.A.engineeringsciencefrom Vibration ResearchattheUniversityofSouthampton consultancy divisionoftheInstituteSoundand Stuart Dyne ismanagerofISVRConsulting,the University ofSouthampton,UK structure ataspeed of4,000to5,000m/s.The CAMissensitiveto the structureand ashockwaveisthentransmitted throughthe ferred. Firsttheblastwavefrom theexplosionsourceimpingeson following explanationoftherecords describedabovemaybein- explosive devicesatmanylocations onseveralaircraft,andthe mences at0.025s. The pressurerecorddiffersfromtheothersinthatitonly com- CVR 2,suggestingapossiblesaturationofthetapedynamic range. high leveluntil0.06sandforthewholeofrecord for after 0.035swhereastherecordforCVRsystem1remains ata is similaralthoughthevibrationresponseamplitudefalls soon (the graphbecomesmorespiky).Interestingly,thevibrationrecord at around0.025sincreasinginamplitudeandfrequency range range (thegraphisfairlysmooth).Allofthemchangecharacter mencing before0.01swithalowamplitudeandfrequency identical. Thefeaturesthattheyshareincludearesponse com- CAM signatureshavesomesimilarfeaturesbutarecertainly not epochs withintherecording. recording butishelpfulhereinthedeterminationofcause obviously thisreferencewouldnotbeavailableonanaccident Time zeroisthetimeofdetonationexplosivedevice— the explosionwasapproximately9.4maftofsensorposition. sensors wereincloseproximitytoeachotherthecockpitand (pressure sensor)installedclosetotheCAMfortrials.All gether withanaccelerometer(vibrationsensor)andamicrophone the CAMchannelofeachthreetape-basedCVRsystemsto- the groundinanex-serviceBAeTridentaircraft.Theplotshows strumentation signaturesforanexplosioneventconductedon Figure 1(reproducedfromreference[4])showstheCVRandin- CAM/CVR recordingsofexplosions locate theseatofhullloss. tween explosionsandstructuralfailuredecompressionsto to considerwhatsystemswouldbenecessarydiscriminatebe- why theserecordingsdonotleadtousefulforensicevidenceand fore oflittlediagnosticvalue.Thisaimthispaperistoexplain NTSB butdidnotrevealthecauseofaccidentandwasthere- cording wassubjecttodetailedandmeticulousanalysisbythe [4]. MorerecentlyaloudsoundattheendofTWA800re- EUROCAE ED-56,atconferences[2,3]andtoanISASIseminar been reportedwidelytoworkinggroups,suchasISASIWG50, structural failurerapiddecompressionshastakenplaceand report, astudyintotheCVR/CAMresponsetoexplosionsand aircraft’s flightrecordersystems.”Sincethepublicationofthis positive andnegativepressurepulses,preferablyutilisingthe vestigation wasthat“amethodshouldbedevisedofrecording event.” Moreoverasafetyrecommendationarisingoutofthein- Similar resultshavebeenobtained fromverymanytrialswith Several featuresarestrikingaboutthisFigure.First,thethree ISASI 2003 Proceedings • 49 ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS 50 Formulae forthis weregivenin[4]takinginto accountthepos- this differenceusingvaluesfor thetwopropagationvelocities. borne andairborneshockwaves andcomputeddistancefrom One simplytookthedifference inarrivaltimesofthestructure- the seatofanexplosionevent shouldberatherstraightforward. tion isolationispoorlydesigned. of afewg’sarethemselvesquitehigh,notthattheCAMor vibra- from thestructure.Thereasonissimplythatvibration levels box investigations,yettheCAMisintentionallyvibrationisolated phenomenon hasbeenexploitedinthepastwithhelicopter gear- It isinterestingthattheCAMquitesensitivetovibration. This Vibration sensitivity and isdesignedtohaveverylowvibrationsensitivity. graph) respondsonlywhenthepressurewavearrivesat CAM CAM isalsosensitive.Theinstrumentationmicrophone(bottom further localvibration(asseenbytheaccelerometer)towhich produces bothapressureresponsefromtheCAMand large andyieldsarequitesmall.Onarrivalinthecockpit,blast speed ofsoundinair340m/swheredistancesarerelatively dependent upontheyieldofexplosionbutcanbetakenas and eventuallyarrivesinthecockpit.Thespeedofthiswaveis wave. Meanwhiletheair-blastwavetravelsthroughfuselage vibration andrespondstothearrivalofstructure-borneshock an explosivedevice. Figure 1:CVRandinstrumentationsignaturesfor The resultsofallthesetrials appearedtoshowthatlocating • ISA SI 2003 P r oceedings the time takenforsoundtotravel fromtheseatofexplosionto the aircraft.ClearlyCAMdoes notshowatransitionatt2.The mately 8.1 on apressurisedBoeing727 aircraft.Thechargewasapproxi- probably highlydistorted. sensors areoverloadedandthattherecordingissaturated and tion isgreaterthanthedynamicrangeofrecorderorthat the the explosionyield.Thisimpliesthatphysicalparameter varia- only greaterthanforthephaset1t2,i.e.,responseafter the However, accidentrecordingsdidnotappeartoshowthese cockpit isapproximately 0.024s,andtheresponse clearlybe-

m aftofthecockpit;explosion rupturedtheskinof series ofcurves onanaccidentrecordingspectrogram withthe retical basis.Operationally,the methodrequiredplacementofa tiple transmissionpathsseverely limittheapplicabilityoftheo- However, thenatureofexplosion sourceandcomplexmul- sive, i.e.,differentfrequencies propagateatdifferentvelocities. of themethodisthatstructural shocktransmissionisdisper- explosion usingthespectrogram oftheCVRrecording.Thebasis described abovetolocatethesourceforaccidentrecordings. guishable fromtt2wouldbeindistin- There wouldthenbenodiscriminationbetweenthetwophases of and therecordingswouldbesaturatedaswascaseinFigure 4. t1t2 One extensivestudy[5-8]hasattemptedtolocatetheseatof an The yieldsofalltheexplosionsanalysedinFigures2and3 are • • erything theinvestigatorwould seeknamely: ings ofpressurefromeithersideaneventappeartooffer ev- been explainedbyreferencetoone-dimensionalflowmodels [2]. tively constantpressureproducingastep-likeappearance) has curve (aseriesofpressuredropsbetweenshortperiods rela- which cabinpressureisventing.Thepreciseformofthe The rateofdepressurizationindicatesthesizeholethrough conditions. Thisisduetothebreachofpressurizedfuselage. pressure falltoavaluesignificantlybelowtheoriginalambient the devicetowithin0.5m.Secondly,bothtransducersshow a two pressurerisescanbeusedtodeterminetheaxiallocationof earliest andgreatestpressurerise.Thetimedelaybetweenthe explosion source.Thetransducerclosesttothechargeshows times becausethetransducersareatdifferentdistancesfrom is widelytabulated[10,11].Thepressurerisesoccuratdifferent is afunctionofexplosionyieldanddistancefromthecharge with apressurerise.Themagnitudeofanair-blastrise these timehistoriesarenoteworthy.Firstbothrecordscommence ally) ofanexplosioninapressurizedfuselage.Severalfeatures 5 showstheoutputoftwotransducersplacedoneitherside(axi- microphones withcorrespondinglow-vibrationsensitivity.Figure of pressuretransducers.Theseareeffectivelyverylowsensitivity Among theinstrumentationdeployedinsometrialswerearrays decompressions Other transducerstodetectexplosions/structuralfailure explosions andstructuralfailuredecompressions. compression andseemstobeunablediscriminatebetween CVR/CAM combinationisunabletolocatethesourceofade- dium andsensorareboth(considerably)exceeded.Thus,the and (ii)becausethelimiteddynamicrangeofrecordingme- be frequencydependent,non-linear,anddirectionallydependent) tion sensorbutmerelyexhibitsvibrationsensitivity(whichmay recordings ofvibrationas(i)theCAMisnotdesignedavibra- sions anddecompressions,theCVRrecordsarenothighfidelity been observedinvariousdecompressiontrials.Forbothexplo- polarity. TheirpropagationvelocityandarrivalattheCAMhas speed ofsoundaswithblastwavesbutareobviouslyopposite pression waveattheCAM.Decompressionwavestravel in thecaseofastructuralfailuret2isarrivaldecom- explosion, t2isthearrivaltimeofblastwaveatCAMand failure andexplosion-generatedconditions.Inthecaseofan is likelytobeproducedatveryhighlevelsunderbothstructural that themethodshouldnotbeusedinaccidentinvestigation. mendation arisingfromareview[9]attheendofstudywas were notabletoconfirmthevalidityofapproach.Arecom- spectrograms ofseveralaccidentswereanalysedinablindtestbut spectrogram leadingtoambiguousresults.Inonepartofthestudy, problematic asseveralsetsofcurvescouldbedrawnonanygiven source totheCAM.Investigatorsfoundthisaspectparticularly intention thattheircurvaturewouldindicatedistancefromthe systems maybe widelyapplicable,trialsareneeded toconsiderthe • sure rise) Interpretation oftheresultsinFigure5indicatesthatrecord- The intervalt1Seattle,September1995. N Fundamentals ofProtectiveDesignforConven- Explosive ShocksinAir Identification oftheCauseFailure Time-frequency AnalysisoftheCVR’s , ProceedingsoftheInstitute , InternationalSocietyof , AirAccidentsInvesti- : Developmentand , Springer-Verlag, , International , Labo- SESSION III ISASI 2003 Proceedings • 53 ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS 54 and it’sjustas true fromthevantagepoint oftheFederalAvia- and dataanalysis.Isaidthis whenIheadedtheSafetyBoard, grams, high-gradematerials, andelectronicallygenerateddata sues involvinghigh-techsafety systems,integratedcomputerpro- the world,accidentinvestigations areincreasinglydrivenbyis- Airlines 587hereintheUnited Statesandfromaccidentsaround numbers ofaircraftonourrunwaysandtaxiways. launches. And,ofcourse,weallrecognizetherisksgreater private aircrafttojumbojetsfromhelicopterscommercial space with differentratesofspeedandflightpatternsfromthesmallest are theincreasingnumbersofaircraft,aswelltypes machinery withhundredsofcomplicatedsystems.Addto this of partsflyinginformation.Theyarehighlycomplexpieces of “kicking tin.”Buttoday’saircraftaremuchmorethanthousands of afatalaccidentaboutoneinevery16millionflights. on thejob.Today,anairlinepilotinUnitedStatesfacesarisk to protectthefuselage—oneinsixU.S.airmailpilotsperished Air MailService—tellingpilotstocrashlandbetweentreessoas three dayslater. brothers identifiedtheproblem,fixedit,andflewintohistory elevator. Henosedup,stalled,anddivedintothedunes.The most madeit.Buthewassurprisedbythesensitivityofplane’s through, WilburfirsttriedtocoaxtheFlyerintoair.Heal- investigators. OnDecember14,threedaysbeforethebreak- sire tofindthesolutionsafeflight. tion pioneer,OttoLilienthal,inaglideraccidentfiredtheirde- brothers interestedinaviation.The1895deathofGermanavia- that itwasthechallengeofsafetyinpartgotWright founder oftheFlightSafetyFoundation—JerryLedererhassaid been thebedrockofaviationsafety.Asourhonoredguest— find theanswerfromlittlebitsofmetal.” and-personal, hesaysis“constantlyamazedatyourabilityto gators beforehestartedhiswork.But,aftersixyearsofup-close- USAir Flight427crashinvestigation,admiredairsafetyinvesti- toughest jobsintheworld.BillAdair,whowroteabookon investigation authoritiesrepresentingfourcontinents. by thepresenceheretodayofsomanyseniorofficialsfrom And, asweallknowfromUSAir 427,TWA800,andAmerican And that,inlargepart,isbecauseofwhatwehavelearned When JerryLedererissuedhisfirstsafetybulletinattheU.S. One couldcallWilburandOrvilleWrightthefirstairsafety And, forthefirstcenturyofflight,accidentinvestigationhas Thank youallforeverythingdointakingononeofthe • ISA SI 2003 P r oceedings Aviation safetyhasnoborders,asdemonstrated to seesomanypeoplefromcountries. together fromallovertheworld.And,it’sgreat really isaremarkableforumthatbringspeople the FederalAviationAdministration(FAA).ISASI President Bush,SecretaryMineta,andallofusat Good morning.Ibringgreetingsonbehalfof By MarionC.Blakey,FederalAviationAdministratorKeynoteSpeaker Learning from ‘Kicking Tin’ dation training materialsintoSpanishandPortuguese. Morethan deserves specialrecognition for translatingFlightSafetyFoun- tral andSouthAmerica.ThePan AmericanAviationSafetyTeam ments. Similareffortsareunder wayinAsia,Europe,andCen- 70 percentwhenweimplement theagreed-uponsafetyenhance- reduce theriskoflosscontrol orCFITaccidentsbymorethan accidents. We’remakingreal progress.CASTestimateswecan well andisaperfectexampleofteamworkgettinginfront of where weneedtotakeadditionalaction. tinuous processthatallowsustoevaluateresultsaswell tosee risks, prioritizeactions,measureanddocumentresults.It’s acon- system safetyapproach.Weidentifyhazards,assessandanalyze And thatisexactlywhattheFAAcommittedtodoing with a them, anduseharddatatodetectproblemsdisturbingtrends. our reactivenature.Wemustgetinfrontofaccidents,anticipate one ofthebiggesthistoricalcharacteristicsaviationsafety— time notlivingoutofasuitcase,homewithyourfamilies. spend moretimeteaching,training,maybeevenspendingsome lenge. Butwewantaviationtobesosafethatinvestigatorscan out ofbusiness.YouandIknowthatthisisaformidablechal- expects, buttheydeserve,thesafestformoftransportation. mitted toanevenstrongersafetyrecord.Thepublicnotonly tors, flightcrews,mechanics,andmanufacturers—mustbecom- ments. Howevergoodtheyare.Allofus—government,opera- a strongsafetyrecord.Butwecan’t,andwon’t,restonourachieve- ful infiguringoutwherethedebrislanded. course, NTSBwithitsstoreofexpertisewastremendouslyhelp- west Regionhelpedrecovertruckloadsofshuttledebris.And, central roleinfoamimpacttesting.AndDonDayofourSouth- failure rates.FAAdebris-reentryspecialistPaulWildeplayeda ing NASA’sdecision-making,completewithstandardsofriskand vestigator DanDigginsfromtheFAA…workedoncharacteriz- plex ways.”Theshuttleinvestigationwastrulyateameffort.In- pivotal roleorganizationandculturecanplayinanaccident. age …extremelyhightemperaturesofspacetravelandthe field fromParistoMoscow…foamthatcausedcatastrophicdam- debris fromCaliforniatoeastTexas,theequivalentofa tigation Boardfacedwiththeshuttleinvestigation…plotting gation SteveWallaceandtherestof have totacklenow.LookatwhatFAA’sheadofaccidentinvesti- cause accidents.It’sthehigh-techandsystemfailuresthatwe failure ofpilotstosettheirflapsduringtakeoff. as likelytobeanerrorinalineofcomputercodeitisthe tion Administration:thecauseofnextmajoraccidentisjust CAST—or theCommercialAviationSafetyTeam—isworking And toreachthatpoint,theaviationcommunityischanging Our goalattheDOTandFAAistoputaccidentinvestigators And it’sthatteamapproachenablesaviationtoenjoysuch As AdmiralHalGehmansays,“Complexsystemsfailincom- We havegottensogoodatsolvingandpreventingthesingle- Columbia Accident Inves- portant exports,andit’soneofourmostimportantimports.We I stronglybelieveaviationsafetyisoneofournation’smostim- Global AviationInformationNetwork,orGAIN,initiative. among safetyprofessionals.Andthisistheentirepointof we needtomakesurethatthesafetydataaresharedworldwide mination forASAPdatashortly.Lookingbeyondourborders, FOQA datawillbeprotected.Weexpecttoissueasimilardeter- disclosure. that protectvoluntarilysubmitteddataandinformationfrom safety dataonanationallevel.TheFAAhasalreadyissuedrules across airlines,wemustmoveforwardwiththeaggregationof the dataremainatlocaloperatorlevel.Toidentifytrends grams isvaluable,itsfullpotentialwillnotberealizedaslong pilots. more than80,000ASAPreportshavebeensubmittedbyairline port possibleviolationsthroughtheirlocalASAPoffice.Todate, vides enforcement-relatedincentivesforemployeeswhoself-re- limitations, aircraftsubsystemmalfunctions,andcountlessmore. problems, includingunstableapproaches,exceedingoperating terns andtrendscanemergethatallowustoidentifyahostof able fromanyothersource.Andit’sthroughthisdatathatpat- objective informationaboutdailylineoperationsthat’snotavail- equipped tocollectandanalyzeFOQAdata.Thisdataprovides grams. Bytheendofthisyear,almost1,800airplaneswillbe FAA-approved flightoperationalqualityassurance(FOQA)pro- ing sohardtosupportASAPandFOQAprograms. possible tomakeinformeddecisions,whichiswhytheFAAwork- on informationgatheringandsharing.Weneedasmuchdata This data-drivenapproachiswhywe’replacingsomuchemphasis 12,000 pilotsreceivedapproachandlandingsafetytraining. Aviation isthemostinternationalformoftransportation,and And thishasbeenastumblingblock.We’vedeterminedthat But whiletheinformationbeingcollectedthroughthesepro- Under AviationSafetyActionPrograms(ASAP),theFAApro- There arecurrently12majorandregionalU.S.airlineswith That’s thepowerofdata,disciplinedanalysis,andfollowthrough. Congratulations, andthankyou,Jerry. that IpresentthisspecialawardtoJerryLederer,Mr.Aviation. takes acalculatedrisktodefygravityandreturnsearthsafely on behalfofmillionsairtravelers,everyonewho with greathonoronbehalfofthemenandwomenFAA, safety recordinthefirstcenturyofflight,itisJerryLederer.It if thereisonepersonwhocanbecreditedfortheoutstanding first accidentinvestigationcourse.AndIthinkit’sfairtosaythat aviation safer.In1948,heorganizedtheFlightSafetyFoundation’s three-quarters ofacenturyfindingtherightsolutionstomake here’s arolemodelforyou,JerryLederer,manwhohasspent tribution totheprofessionofairsafetyinvestigator. pared withtoolsandtraining.Thiswouldbeanenormouscon- not havethesameleveloftrainingavailabletothem—arepre- that yourmembers—especiallyairlinememberswhomay is goingandleadthedevelopmentofmoretrainingtoensure tioned totaketheleadlookatwhereaviationandtechnology investigators. Astheinternationalsociety,youareideallyposi- you tobuildandgrowmaketheseavailableevenmore of standardizingtrainingcoursesforinternationalstudents. investigation schoolinOklahomaCityandwe’retheprocess lighting. Trainingissoimportant.TheFAAhasitsownaccident ing, andawarenessaswellimprovedairportmarkings drivers. Thisiswhywe’resofocusedonincreasededucation,train- greater awarenessbypilots,controllers,andairportvehicle know wehavemoretodo.Improvingrunwaysafetydependson (AMASS) with37totalinstallationsplannedfor34airports.We has commissioned31AirportMovementAreaSafetySystems resolution alerts. sign andcertificationprocesstostandardizingresponsesTCAS many, theUK,andTaiwanonabroadrangeofissuesfromde- is takingactiononrecentrecommendationsfromCanada,Ger- learn somuchfromourinternationalpartners.Infact,theFAA As youthinkabouthowcanbecomeevenmoreprepared, I applaudISASIforitsinternationalseminars.Andchallenge And we’readdressingsafetyontheairportsurface.TheFAA N ISASI 2003 Proceedings • 55 ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS The BEAwasnotified oftheaccidentbylocal civilaviationau- Conduct oftheinvestigation with onepersonwhowasinthe house. near ahouseandcaughtfire. All oftheoccupantsperished,along with asteepbankanglethen dive towardtheground.Itcrashed pass, severalpeople,including theAFISagent,sawitturnleft Runway 10.Thatwastheirlastcommunication. passing thePaindeSucrereportingpointforafinalapproach to Fourchue reportingpoint.Shortlyafterward,theyannounced Barthélemy informationfrequencythattheywerepassing the Barthélemy. Afewsecondslater,theyannouncedontheSaint cuit located3nauticalmilesnorthwestoftheisland Saint of theislandFourchue,entrypointaerodromecir- left theSaintMartinJulianaaerodromefrequencywhenabeam the pilotflying. undertakes commercialoperationsontheroute.Thecaptainwas ated byCaraïbesAirTransportonbehalfofCaraïbes,which Saint Barthélemy,19nauticalmilesaway.Theaircraftwasoper- flight planbetweentheislandofSaintMartinand was carryingoutthescheduledserviceTX1501underaVFR On Saturday,March24,2001,theDHC-6-300registeredF-OGES Presentation oftheevent cial pilot’slicensewithaninstrumentratingandATPLcertificate. joined theBEAasasafetyinvestigatorin2000.Heholdscommer- held aBeech200typerating. a commerciallicenceandmultiengineinstrumentrating.Healso Flight SafetyFoundation,inArlington,Va.Heisthecurrentholderof aeronautical engineeringin1987,includinganinternshipatthe National CivilAviationSchool(ENAC)withamastersdegreein 56 When theaircraftbeganitsshortfinalbeforeLaTourmente Cruise wasperformedbetween1,000and1,500feet.Thecrew • ISA SI 2003 For DHC-6TwinOtterAccident, P r Investigating TechniquesUsed the FrenchCivilAviationAuthority(DGAC),he Studies andResearchProgramsinAeronauticswith University. Afteranappointmentasmanagerof graduate degreeinhumanfactorsfromParis School (ENAC)in1995.Hereceivedhispost- engineer fromtheFrenchNationalCivilAviation Gérald Gaubert oceedings Supervision Division.HegraduatedfromtheFrench as deputyheadoftheFlighttrainingOrganizations civil aviationauthority)for8years,including4years to joiningtheBEA,heworkedforDGAC(French Safety AnalysisandStudiesDivisionin1996.Prior Department Head.HejoinedtheBEAasheadof Stéphane Corcos , 30,graduatedasanaeronautical , 39,istheBEAInvestigations By StéphaneCorcosandGéraldGaubert,BEA,France March 2001 left nochance ofsurvivalforanyitsoccupants. Theinitial been subjecttoaviolentimpact, followedbyanintensefire,that • flight crewsshortlybeforeor shortly aftertheaccident. • during theflight. • had beengrantedawaiverduetoitsinitialdateofcertification. rect evidencecouldbefound. pants inwhich,unliketheaverageinvestigation,verylittle di- were facedwithapublictransportaccidentkillingall19 occu- This wasthestartingpointofanunusualinvestigation,where we Findings andchallenges planes carryingmorethanninepassengersonrevenueflights. dation wasissuedonthecarriageofflightrecordersforallair- the investigationwaspublished.Atthattime,asafetyrecommen- technical adviserfromthepropellermanufacturer,Hartzell. portation SafetyBoardwasattachedtotheinvestigation,witha Canada. Subsequently,acorrespondentoftheNationalTrans- the manufacturers,deHavillandBombardierandPratt&Whitney March 27,accompaniedbytwotechnicaladvisersrepresenting tative. Thelatterjoinedtheinvestigator-in-chargeonTuesday, pate intheinvestigationbynominatinganaccreditedrepresen- counterpart, theTransportationSafetyBoard(TSB),topartici- craft wasofCanadianmanufacture,theBEAinvitedits site intheFrenchWestIndiesfollowingday. thorities anddispatchedateamofinvestigatorstotheaccident Aerial viewofthesiteandaerodrome. Additional initialfindingswerethat First ofall,therewerenoflightrecorderssincetheairplane On Aug.6,2001,apreliminaryreportdetailingprogressin In accordancewithinternationalagreements,becausetheair- the e no unusualweatherphenomenon wasr no abnormaleventswer xamination ofthecrashsite r e r eported bythecr evealed theairplanehad eported byanyother ew atanytime • necessary totrytakeadvantageofanypossibleclues. to impact.However,furthertechnicalexaminationwasdeemed usual attitudesoranyabnormalsituationonboard1minuteprior stroyed; whenplayedbackthesecondrevealednosignofun- • powerplants) showednoevidenceofanymalfunction. examination ofthewreckage(flightsurfaces,flightcontrols, Descent path12percent. Crash site. • was available. • is setat12percent). from it(inparticular,thefinaldescentpath rating isrequiredforoperationstoand local windconditionsaresuchthatasite surrounding terrain,environment,and this airplanetypeatairport,wherethe • could havebeenafactorintheaccident. to theaccident, andfrompeopleinthe eyewitnesses ofthefinalsequence thatled rely onsignificanttestimony,both fromthe tions. teardown revealednomechanical malfunc- wreckage andapowerplant propeller The investigationcould,therefore, only the A two camcor the autopsyr further e the captainlack TIS r xamination ofpartsfr ecor der tapeswer evealed nopr der wasunserviceable,andnoradartracking ed r ecent e e foundinthedebris:onewasde- xperience on oblem that om the Spectral analysis ofthesoundtrackontakeof f. with asecondcamcorder,synchronizedthefirstone,and the samepassengerseat.Wealsorecordedengineparameters operator’s airplane). flights fromSt.MartintoBarthélemy(onboardalocal board thesametypeofairplane,duringcommercialscheduled ter apprisedofnormaloperationsinsuchanenvironment. spectral analysis,toestimatethetrackofF-OGES,andbebet- found duringtherecordedpartofflight. use. and right-handpropellers,whichisconsistentwiththeirnormal was madepossiblethroughanalysisofthevideo. nized, andthediscriminationbetweenleftrightpropeller During theapproach,propellerswerenolongersynchro- over 5secondsthenstabilizedat1,685RPM(synchronized). nized). Duringcruise,theysloweddownfrom1,825to1,690RPM the propellerrotationspeedthroughfrequencymeasurements. the accident. the northwesternpartofSt.Barthélemy,around1minutebefore rotation. of theisland. • • • age showedthreepartscontaininguseableimages: The playbackandtheanalysisofavideotapefoundinwreck- Analysis ofavideotape steep nose-downattitude. turn, thenpitcheddowntowardtheground,crashedina the airplanebankedtoleft,inasharperandleft left withapuzzlingtriggeringeventdescribedbyeyewitnesses: working environmentofthecrewmembers.Theinvestigatorswere The filmwasshotwiththesamemodelofvideocamerafrom The methodologyusedwasacomparisonwithrecordingson Three otherobjectiveswerepursued:tovalidatethesoundtrack To summarize,noevidenceofanymalfunctionormisusewas An asymmetryaround9percentwasfoundbetweenleft-hand During takeoff,bothpropellerswereat2,120RPM(synchro- Spectral analysisofthesoundtrackallowedustodetermine The filmendedapproximatelywhentheairplanewasabeam In allofthesequences,leftpropellershowedevidence on appr during cruisewithaviewofthetipSt.Martinisland, during initialclimbfr oach toSt.Barthélemywithaviewofthenortherntip om St.Martinaer ISASI 2003 odr ome, Proceedings • 57 ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS • merged imagesfrombothsourcesintoasinglefilm. 58 vantage ofsuch amethod,inadditiontoits beingsimple,wasto ponent ofADREP(initslatest edition“ADREP2000”).Thead- were alsotobesoughtinthe SHELL model,whichisasubcom- tified eventsthathave,atone time,ledtoaccidents.Thecauses exploration, sinceitcontains anorganizedcollectionofalliden- Therefore, theuseofADREP modelwasanaturaltoolfor was veryremoteandthisassumption wasdeemedacceptable. investigation, thelikelihoodthata“discovery”wasbeingmissed previously described.Withoveracenturyofhistoryinaccident did notresultfromunknownphenomenathathadnever been already documentedininternationallyacceptedtaxonomy and ing assumption:thisaccidentwastheresultofidentifiedevents, to becompleted.Theworkundertakenwasbasedonthefollow- The worktoidentifytheeventsandpotentialcauseswas still Determination ofthescenario runway track. • Sucre, whichisratherinthelowerrangeofusualaltitudes; • low pitch; • image, andwasoperatedinitsnormaloperatingrange; Propeller rotationduringapproachsequence. Photos copiedfromthevideotape,asanillustrationofcalculationnumberpassagesbladespersecond. We wereabletoestablishseveraltechnicalfacts,amongwhich the enginewasdeliveringpowerfr the airplanewaspr the altitudewasbetween1,000to1,100ft.acr at theendofsequence,pr • ISA SI 2003 P r oceedings obably indescent,andslightlyrightofthe opeller wasselectedonfull om tak eoff uptothelast oss theP ain de there-itis, theimpact ofastrongauthority gradient ofthecap- tion: thelackofrecentexperience forthecaptain,alikelyget- Other factorswereanalyzed during thecourseofthisinvestiga- Elements fromtheoperational context the lossofyawcontrol,thenroll control,ontheaircraft. ment whencomingoutofthereverse betarangewouldhavecaused control ofhistrackonshortfinal.Athrustasymmetryatthe mo- the reversebetarangeforpropellersbycaptaintoimprove of probability.Themostprobableisthatadeliberateselection of probability. reverse betarange,with,nevertheless,totallydifferentlevels of crew—loss ofcontrolduringago-aroundorinflightuse of the speed reduction—andtworelatingtodeliberateactions by the sudden incapacityofacrewmemberorstallduetoexcessive could beretained:tworelatingtoinvoluntaryactionbythecrew— which wecategorizedallpotentialeventsthathadlookedat. tion. Wethencameupwithafaulttreeforthehypotheses,in third: whetheritwastheresultofanormalactionorviola- crew related,second:whetheritwasinadvertentorvoluntary, sons ofsimplification.Firstdividingline:whethertheeventwas endeavored tomergetwoofthemintothemainthreeforrea- The treewasinitiallybasedonfivelevels(dividinglines),butwe them intoamoresyntheticapproachwiththeuseoffaulttree. Of thefourhypothesesretained,firstthreehavealowdegree Among thevarioushypothesesthatwereanalyzed,onlyfour needed tocondensetheresultsandsort tive probabilitiesofcausalfactors. tion—is stillneededtoassesstherela- testimonies andotherfactualinforma- expert judgment—sometimesbasedon not provideadefinitivescenario,and before beingdiscarded.Finally,itdoes needing athoroughlyarguedrefutation many irrelevantcombinations,eachone tage ofthissolutionisthatitcanleadto grounds andexperience.Thedisadvan- the resultissodependentonback- consequence ofbrainstormingwhere tuitive assertions,whichisacommon eliminate thebiasesresultingfromin- fective prevention.Italsohelpedus avenues gavemoreofachanceforef- tential factors.Exploringallpotential ensure thatwewouldnotmissanypo- After thisworkwascompleted,we fr.org orwww.bea.aero. the circumstances. eration thatenabledustorunthisinvestigationsosmoothlyin Finally, wemustemphasizetheoutstandinginternationalcoop- of thetoolsdevelopedbyICAOforaccidentdatabasecoding. of theaccidentasrigorouslypossible.Wetookgreatadvantage use originalsolutionstotryandgetabettergraspofthecauses investigating authoritytouseweremissing.We,therefore,had killed 20,andalloftheclassictoolsthatonecouldexpectan This eventaffectedascheduledrevenuepassengerflightthat Conclusion the difficultyofapproachtoRunway10atSt.Barthélemy. tive flightsonpotentialdeviationsfromstandardprocedures,and tain overhiscopilotCRM-wise,theinfluenceofshortandrepeti- The finalreportisavailableontheBEAwebsiteatwww.bea- N ISASI 2003 Proceedings • 59 ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS • document. etc.) usedwithtemplatesandforms,suchasaword-processing • a minimumofsupportfromotherpeople. riences, andtoolstoenableahighlevelofjobperformance with grated, on-demandaccesstoinformation,advice,learning expe- cilitating responsiveness,dataaccess,andcollaboration. complexity, eliminateduplicateeffort,andreducecostswhilefa- and resourcestoincreaseproductivity,shortencycletimes,lower that provideinvestigatorsandassociatedagencieswiththetools information technologytodeliverintegratedend-to-endsolutions Electronic PerformanceSupportSystems(EPSS)employexisting transferred manuallytotheproperformatand/ordatabase. result induplicatedeffort;dataarecollectedthefieldandlater workload. Theseinvestigationsgeneratereamsofpaperandcan sis ofallpertinentdata,contributetotheinvestigators’ tion ofdataandinformationfromnumeroussources,analy- ous governmentagenciesandpartiestotheinvestigation,collec- acterized byarequirementforconstantcoordinationamongvari- many as20investigationsatonetime.Theseinvestigations,char- complete. Itispossibleforaninvestigatortobemanagingas ber ofon-sceneandlimitedinvestigationstheycaneffectively With afinitenumberofinvestigators,thereislimittothenum- 1,700 accidentsingeneralaviationtheUnitedStatesalone. for GECapital. Customer TrainingforSikorskyAircraft,andperformanceconsultant 60 lations andscenarios. as interactivetutorials,well asmultimediatrainingusingsimu- • ing proceduresandmakingdecisions. system, orcoachingfacilitythat guidesauserthroughperform- • and case-historydatabases. on-line helpfacilities,statisticdatabases,multimedia E These arefourattributescommontoanEPSS: An EPSSisdefinedasanelectronicsystemthatprovidesinte- In ordertoenhanceprocessessuchasaccidentinvestigations, EPSS solutions tie togethertheuserinterface functionality, data, Information Base T Learning Experiences Advisor • ools Investigation EnhancementThrough ISA SI 2003 safety investigators(ASIs).In2002thereweremorethan occur requiringvariousdegreesofinvolvementfromair ach yearthousandsofaviationaccidentsandincidents —productivity software(wordprocessing,spreadsheet, —an interactiveexpertsystem, cased-basedreasoning P r oceedings C-5 aircraftSystemSafetyBoard,Managerof experience. Hisbackgroundincludesmanagingthe and 25yearsofleadershipmanagement is aretiredAirForceofficerwith2,500aircrewhours Safety InspectionSystem(OASIS)wasdeveloped.He Scientific’s AtlantaofficewheretheOn-lineAviation Ja Jay Graser —on-line referenceinformation,hypertext Information Technology —computer-based-training (CBT), such is theOperationsDirectorforGalaxy By JayGraser,OperationsDirectoGalaxyScientificCorporation 1

2 FAA hadfielded. by thesafetyinspectorsassinglemostimportanttool produced a20percentincreaseinefficiencyandwasidentified the fieldworldwide.Anin-depthstudyindicatesthatOASIShas than 4,000FAAairsafetyinspectorsandaviationpersonnelin Aviation InspectionSystem(OASIS),whichisemployedbymore gations. applications canbemodifiedforuseinaccident/incidentinvesti- involved anddesiredoutcomesisperformed,field-provenEPSS phy, electronicsketchpad,etc.Onceadetailedstudyofthetasks nents suchasglobalpositioningsystem(GPS),digitalphotogra- bilities canbefurtheraugmentedwithfullyintegratedcompo- tivity, andabilitytooperateindisconnectedmode).Thecapa- mobility (vialaptop,handheldandtabletPCs,wirelessconnec- forms, rules-basedprocessautomation,dataintegration,and specific tasks.Thisincludestoolssuchasautomated“intelligent” information processes,andbusinessrulesneededtoperform known information. through theuseofformsthat automaticallyfillinfieldsbasedon cation ofeffortandimproves consistencyofthedatacollected formation directlytothedatabase. Thisapproachreducesdupli- while ina“disconnected”mode, andlaterdownloadingthein- formation whilestillintheoffice, collectingthedatainfield in thefieldusingalaptopcomputer byuploadingrequiredin- One exampleofanexistingEPSSapplicationistheOn-line The OASISEPSSallowsinvestigatorstoperforminspections ing referencematerial whileperformingother tasks,suchasdriv- thetic voice.Thisallowstheinvestigator tobe“handsfree”learn- tem translatesselecteddocuments intoanaudiofileusingasyn- Text toAudioConversion from theEPSS. needed inaninvestigationis madeavailableinadigitalformat Reference Material recovered. ordinates ofdebrisandphotosthescenewhere was Integrated GPSMapping CE, andPalmOScompatibledevices. ing degreesondifferentdevices,suchaslaptopPCs,Windows Multiple PlatformCompatibility the systemtoprotectdatafromunauthorizeduse. Biometric Security sonnel. sion orattemptedaccessofthePCPDAbyunauthorizedper- Encryption greater datastorageinlessspace. as 4000:1inordertospeedwirelesstransmissionandallowfor Data andImageCompression when awirelessconnectionisavailable. nology byallowingtheinvestigatortouploadanddownloaddata Wireless Mode when theinvestigatorisinfield. needed datawhileconnectedtothenetworksothatitisavailable Disconnected Mode signed specificallyforaccidentinvestigations. The followingEPSSfeatureswouldbeintegraltoasystemde- —Files areencryptedforprotectionduringtransmis- —The systemtakesadvantageofwirelesstech- —A thumbprintscannerisintegratedinto —Any referencematerialthat mightbe —The systemallowstheusertodownload —Area mapsarelinkedwithGPSco- —Much likeane-mailreader, thesys- —Images arecompressedasmuch —The EPSScanbeusedtovary- control personnel, theinitialdatacollection mustbeeasytoac- including lawenforcement, insurance companiesorairtraffic Notification Scenario accident investigation. The followingisascenariowhere anEPSScouldbeusedin wording thatisconsistentwithprescribedandacceptedstandards. Style Guide help developpossiblescenariosofwhatoccurred. It canbeusedtohelpwitnessesdescribewhattheysawand also Flightpath Simulation format mostusefultotheinvestigator. government andnongovernmentdatabases,presented ina The informationiscollectedfrommultiplesources,suchas both stored onthedevice’sharddrivesothatitisavailablein field. investigation iscollected,reducedtoonlywhatneeded,and Integrated andPortableData removable transparencies,withoutdamagingtheoriginalphoto. scene, electronicallymarkingtheimagesusingasmanyfour Sketch Pad per documentcopies. ments, suchaspilotlogbooks,reducingoreveneliminatingpa- cluded intheinvestigator’sequipmentandusedtoscandocu- Document Scanning the appropriatehelp. keeps trackofthecurrentstageinvestigationandprovides rors, andalerttheinvestigatortorecentchanges.Thesystem that prioritizeperishableevidence,avoidcommoncausesofer- Active Help EPSS andformsthroughverbalcommands. Voice Interface viewing theaccidentscene. speech andallowsdictationtothesystem“handsfree”whilere- Voice toTextConversion ing orwalkinganaccidentscene. —The investigatorcanannotatephotographsofthe —Since notificationcouldcome frommanysources, —As theinvestigatoriswriting,systemsuggests —A systemofhints,reminders,alerts,andchecklists —Allows theinvestigatortomovethrough —A small,lighthandscannercanbein- —A 3-Dflightsimulationoftheaircraft. —The system“learns”theinvestigator’s —Data appropriatetothecurrent ISASI 2003 Proceedings • 61 ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS Family Affairswas notcalledoutonthisaccident, itwouldbeup aircraft, theplannedflight,and thesurroundingenvironment.If already listenedtoseveralhours ofaudioregardingthepilot, On-scene arrival them toshareinformationand plantheinvestigation. are used,theycanusethebuilt-in collaborationtoolsthatallow call andfindroomsaconference facility.WhenGoTeams vestigator couldhaveusedthelistoflocalhotelsinEPSS to have involvedotherinterestedgovernmentagencies,sothe in- more seriousaccident,requiringa“GoTeam,”itwouldprobably tact informationisreadilyavailableintheEPSS.Hadthisbeen a local authorities.Police,fire,anddisasterresponseagency con- plied intheEPSSremindsinvestigatortocoordinate with GPS coordinatesandanypossiblehazards.Thechecklist sup- the surroundingterrainandmapsofarea,complete with of theultra-compressedimages,includingsatelliteimaging of the researchthatwasn’tconvertedintoaudiofilesandview some the triprequiresaplaneride,investigatorcanreadsomeof gator cancontactthem,iftheydon’tcalltheinvestigatorfirst.If system. TheircontactinformationisintheEPSS,soinvesti- with you.TheyreceivedautomaticnotificationfromtheEPSS and enginemanufacturersrepswhohavebeenassignedtowork While theinvestigatorarepreparingtotravel,soaircraft synthetic voiceissimilartothoseusedfore-mailreadersystems. would createafilethatcanbedroppedtoCDorMP3.The auto readeroptionforseveralfilesofresearch,andthesystem used tostudytheavailabledata.Theinvestigatorcouldselect would requirealongdriveorotherdowntime,whichcanbe and isitamountainousarea.Insomecasesthetriptoscene area; forexample,isitcoldenoughtorequireweathergear The EPSSchecklisttellstheinvestigatorcurrentweatherin aged intheEPSSsoitcanbeeasilystudiedenroutetoscene. EPSS anddownloadthefileforaccident.Informationispack- travel totheaccidentscene,heorshecansecurelylogon Investigator response of thenavigationalaidsandtraffic. the accident.Additionally,systemwouldcheckforconditions Environment also collected. craft andhelpfulhintsininvestigatingthistypeofaircraftare history, andlimitations.Previousinvestigationsinvolvingthisair- lems, manufacturerdata,productioninformation,maintenance the make,possibleconfigurations,engines,avionics,knownprob- Machine operations personnel,andmaintainers. that mightprovideevidenceregardingtheaircrew,controllers, ing records,inspectioncertifications,andanyotherdata Human factors consistent intheEPSS. information ispassedfromoneinvestigatortoanother,itremains number andanyotherinformationthecallercouldverify.As accident. Thissearchcouldbebasedontheaircraft’sregistration the appropriatedatabasestocollectavailabledataon establishes anewcasefileimmediatelyandbeginsqueriesofall questions inordertoensurecriticaldataisnotlost.Thesystem guides thepersonansweringcallthroughaskingaseriesof cure webinterface,regardlessofhisorherlocation.TheEPSS call regardinganaccidentcanaccesstheEPSSsystemviaase- cess andstraightforwardtouse.Thepersonwhofirsttakesthe 62 • ISA SI 2003 —Aircraft type,images,includingbackgrounddataon —Weather conditionsintheareaduringtimeof P r —Flight planifavailable,medicalrecords,train- oceedings —Upon arrivalatthescene, investigatorhas —As theassignedinvestigatorpreparesto and whenitis projected tobecomplete. results, includeapop-upmessage tellingthatstatusofitem ence topartsoftheinvestigation thatarepending,suchaslab the investigator’sreportfitswith theacceptedformat.Anyrefer- automatic styleguidemakessuggestions alongthewaytoensure the office,investigatorcan beginrefiningthereport.The point. report basedonwhattheinvestigatorhascollectedupto that and thebuilt-inreportgeneratorwithstyleguidesuggestsa press order toprepare,theinvestigatorclicksonPressReport mode phone callsorinmorningand/oreveningpressconferences. In tors. nario, basedonthesamedynamicmodelsusedbysimula- data, theEPSScreatesadesktopanimationofpossible sce- parameters ofweatheranddebrispattern,and,ifthereisenough lects alltheaccountsofseveraleyewitnessesandaddsknown ture theiraccountinasimulationandtext.Theinvestigatorcol- of theaircraftand,basedonwhattheytellinvestigator,cap- saw. UsingtheEPSS,investigatorcanshowthema3-Dimage may realizemostofthemcannoteffectivelydescribewhatthey government agencies,manufacturers,andoperators. cal informationcanbesharedwithfellowinvestigators,other the cellphoneandlaunchesacompressedfiletosystem.Criti- his orheroffice.TheinvestigatorpatchesPDAinto investigator hasenoughdatathatitwouldbeworthsharingwith and dictatesometexttogoalongwiththenewhint.Soon, camera totakeshotsofthearea,downloadthemintotheirPDA, like tocaptureforotherinvestigators,theycouldusetheirdigital the years.Additionally,ifinvestigatorshaveahinttheywould them totapthecollectiveexperienceofallinvestigatorsover counter anunfamiliarsituation,theinvestigation“hints”allow them ofwheretheyareinthechecklist.Ifinvestigatorsen- memory. recovery, becausesomehaveinternalbatteriestopreservethe vestigator mayneedtodeterminewhichboxesprioritizetheir include avionicsthathaveNVM(non-volatilememory).Thein- the systemcouldremindinvestigatorthataircraftmay and viewtheassociatedpictures.Anexampleofanalertisthat Rather thanhearthetext,investigatormayelecttoreadit investigator thatthenextstepinchecklistcomeswithanalert. out everhavingtotouchthePDA.Anaudibletonenotifies vated interfaceallowstheinvestigatortonavigateformwith- or office,butmuchofthetypingiseliminated.Thevoice-acti- can editthevoice-recognition-generatedtextfilesbackathotel headset typicallyusedforhisorhercellphone.Theinvestigator the scene“handsfree”andtalktosystemusingsame need tomake. have dropdownmenus,reducingthenumberofdecisionsyou “N” number,andalmostallofthefieldsthatarenotfilledout filled outfromwhatthesystemwasabletoresearchbasedon “kicking tin,”the“smart”formsinEPSSarealreadypartially while notimpedingtheinvestigation.Asinvestigatorbegins ance intheEPSStoprovidethemsupporttheywillneed to theinvestigatorapplyhisorherexperienceandguid- Once thefieldinvestigation is complete,onthetripbackto The investigatorwillneedtodealwiththemedia,either via As theinvestigatorbeginsinterviewingeyewitnesses,heorshe Even iftheinvestigatorsgetdistracted,EPSSwillremind The voice-recognitionsystemallowstheinvestigatortowalk An EPSSappliedtoaccidentinvestigationswouldenhancethe Conclusion munications. and audiofilessuchasthecockpitvoicerecorderATCcom- sor, includingimages,simulationsofthemostlikelyscenario, plete, theinvestigatorcansenditelectronicallytosupervi- will allowthemtocompletethereport.Oncereportiscom- results theinvestigatorwasawaitingareavailableandthatthis days, theinvestigatorreceivesane-mailreminderthatlab tion totheothersixormorereportsinprogress.Afterafew or herthoughtsarestillfreshandthenturnshisatten- tor makesafewmoreentriesintothemostrecentfilewhilehis Back attheoffice —Upon returningtotheoffice,investiga- 2 1 Footnotes able causeoftheaccident. ciency whilestilldeliveringacredible,reliablereportontheprob- allow investigatorstohandlemultiplecaseswithmaximumeffi- nate paperfeaturessuchascasemanagement,andtrackingwould ties throughtheuseofelectronicfilesandreduceorevenelimi- torial communicationbetweeninvestigatorsandinterestedpar- reduce paper,andultimatelysavelives.Written,audio,pic- effectiveness andefficiencyofinvestigations,savetime,radically Raybould,B.(1990).Solvinghumanperformanceproblemswithcomput- Gery,G.(1991). ers. Weingarten Publications. the workplacethroughstrategicapplicationoftechnology Performance &Instruction Electronic performancesupportsystems:Howandwhytoremake , 29 N (11), 4-14. ISASI 2003 . Boston,MA: Proceedings • 63 ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS 64 colleagues were hiredtoquellthenervousness ofnewflierson from OaklandtoChicagoona BoeingTrimotor.Sheandhereight May 15,1930,thefirst“skygirl,” EllenChurch,workedtheflight country flights.Theyhiredeight nursestofulfillthenewrole.On Airlines begana3-monthstewardess experimentontheircross- dangerous airplanes. way toenticepassengersflyonthesometimesunreliable and load baggageandprovideinflightservicestopassengers.It wasa Europe byBritain’sTheDaimlerAirway.(1)Thesestewards would sengers. In1922,someofthefirstcabinboyswereemployed in contracts, wesawairlinesbegintocarryalimitednumberof pas- attendant roleontheaircraft.WithadventofU.S. mail the 1920sand30s,wefirstbegantoseestartof flight dents orproblemsregardinganevacuation. vide valuablerecommendationsforuseinpreventingfuture acci- flight attendantscanassistinanaccidentinvestigationandpro- mendations forthepurposeofpreventingfutureaccidents.Trained dent anddeterminetheprobablecause,thenmakerecom- accident investigationistolookatthedetailsofaparticularacci- some ofthelessonslearnedfromthem.Theentirepremise on boardtheaircraft,thenlookoverafewhistoricalaccidentsand were madeinthecabin. tigation intheUnitedStatesandsomeofimprovementsthat cifically aboutflightattendantparticipationinanaccidentinves- a historicaloverviewofthecabinenvironment,Iwillspeakspe- versary oftheWrightBrothers’firstcontrolledpoweredflight.As L recurrent training. flight attendantsinadditiontohelpingplanandteachattendant for theAFA.Athercarrier,sheservedinmultiplerolesrepresenting 16-year flightattendantonaleaveofabsencefromhercarriertowork to AFAparticipantsinNTSBaccidentinvestigations.Kolanderisa and handbookchanges.Shealsoprovidesthetrainingassistance office. Shedealswithproposedregulatorychanges,advisorycirculars, liaison betweenthesafetychairsatcarriersandInternational contacts inthedepartmenttodealwithsecurityissues.Kolanderis tional Officers,anditsBoardofDirectors.Sheisalsoonethemain Participation inAccidentInvestigations Then in1930,againstsomemanagement opposition,United Let’s beginwithabriefhistoryoftheflightattendant.During I willreviewtheoriginalroleandhistoryofflightattendant • Historical ReviewofFlightAttendant ISA SI 2003 (ISASI) 2003seminartohelpcelebratethe100thanni- the InternationalSocietyofAirSafetyInvestigators adies andgentlemen,Iampleasedtobeheretodayfor P r oceedings Candace K.Kolander providing servicestotheAFAmembership,Interna- regulatory, andtrainingmissionofthedepartmentby works withtheDirectortoaccomplishresearch, Association ofFlightAttendants(AFA).Assuch,she Washington, D.C.,Internationalofficeofthe Safety, HealthandSecurityDepartmentatthe By CandaceK.Kolander,AssociationofFlightAttendants,AFL-CIO istheCoordinatorforAir specifically permitted bythetermsofweathercompetency letter. uled aircarrierflightsataltitudes above18,000feetareprohibitedunless obstructions toflightandavoid hazardousweatherconditions.Sched- are prohibitedexceptfortheperiods oftimewhicharenecessarytoclear and careforthepassengers.Scheduled aircarrierflightsabove15,000feet sea level,thereshallbeacompetent cabinattendantprovidedtoobserve carrying passengersandoperating atairaltitudeabove15,000feet Maximum altitudeofflightoperations.Inscheduledaircarrieraircraft dent investigation.(3)CivilAirRegulation(CAR)§61.742 read: and responsibilityforeconomicsafetyrulemaking acci- operations. TheCivilAeronauticsBoard(CAB)hadtheauthority a flightattendantseemstobein1941regardingaltitudeof flight keting slogans,“Coffee,tea,orme?”“I’mCindy,flyme!” for companyprofits.Ipersonallycringewhenheartheold mar- cally inlateryearsasourroletookonasexualundertoneexploited utilitarian, yetprofessional.Thisimagewouldchangedramati- were sturdyandlaceup—thiswasano-nonsensestylethat standard attire,replacedwithnurses’uniformsworninflight.Shoes ately fortheworkandenvironment;woolsuitscapeswere aircraft forpracticalandmarketingreasons.Wedressedappropri- under theageof25,andweight115pounds.(2) those original“skygirls”alsohadtobesingle,childlessfemales heated, orairconditioned.Inadditiontobeingregisterednurses, 24 hourstocomplete,inanairplanethatwasnotpressurized, those long,arduousjourneysthatsometimestookbetween18and Sky Girlsbegin. As farasIcantell,theearliestreferencetorequirement for In theearlyyears,flightattendantswereonboardcommercial knowledge theflightattendant couldoffer. comment. Butitwasabeginning inrecognizingthebenefitsof non-survivable accident,therepresentative wasunabletoofferany asked iftherepresentativehad anycomments.Becausethiswasa was recognizedbytheBoardat theendofpublichearingand for EasternFlight642,thesteward andstewardessrepresentative some ofthesepublichearings.Infact,atahearingin 1955 and location.Duringthe1950s,membersofthisuniondidattend the originalWesternUnionTelegramannouncinghearingdate sor tothe the finalreportswerebarely10pageslong. was becauseofourinternalofficememosatthetime.Some of report; theonlyreasonIknewtherewasastewardesson board did notevenmentionthenameofdeceasedstewardess inthe attendants, werelargelyignored.Afewoftheaccidentreports tors, whichaffectedthesafeevacuationofpassengersandflight to, during,andafteranincidentoraccident.Thein-cabinfac- years priorto1960,fewmadementionofcabinconditions dent reportsdonebytheCAB,notNTSB. reference anaccidentreportpriorto1967,Iamreferencingacci- marily aimedatfindingoutwhytheairplanecrashed.WhenI dent investigations.TheCABaccidentinvestigationswerepri- existence inthespringof1967.Priortothat,CABdidacci- them. some ofthesepastaccidentfileswhenwewereaffiliatedwith very scarcein“content.”IamstillworkingwithALPAtoobtain investigation. IwilladmitsomeofthecurrentfilesatAFAare documentation offlightattendantparticipationinanaccident dants begantodocumentandworkonaccidentinvestigations. Association (ALPA).Itwaswithinthisaffiliationthatflightatten- way oranotherwithourbrotherorganization,theAirLinePilots flight attendantunion’sroleinanaccidentinvestigation. by men.Thismaledominationwasalsoevidentregardingthe the labormovement,withrareexceptions,wasalsodominated criminatory towardwomen.Atthetimemyunionwasfounded, sents 50,000flightattendantsat26airlines. ciation ofFlightAttendants(AFA),AFL-CIO.Today,AFArepre- (ALSA) isthepredecessortomycurrentorganization,Asso- formed theirownunion.TheAirLineStewardessesAssociation strips. Withinlittlemorethanadecade,theearlieststewardesses they wereoftenonlysecuredbyflimsythumbnaillatchesorcloth ens, coffeepots,sodacans,andservingutensils.Thatisbecause her withinflyingdistanceofloosegalleyequipmentsuchasov- poor salaries.Forexample,herseatingenvironmentoftenput disillusioned bythepoorworkingconditions,longhoursand held glamorousjobs,therealitywasthatstewardesseswere airplanes of10-passengercapacityormore. shall beprovidedbytheaircarrieronallflightscarryingpassengersin lation §40.265read: specific totherequirementforaflightattendant.CivilAirRegu- As farasIcan tell, thefirsttimemyunion predecessors par- Throughout thoseyears,itwasinterestingtoseethepredeces- In myreviewoftheAFAaccidentinvestigationfilesduring The currentNationalTransportationSafetyBoardcameinto I havesearchedAFA’saccidentfilesfrom1951to1970forany From 1949until1973,thepresentdayAFAwasaffiliatedinone The attitudeofsocietyduringthoseearlyyearswasoftendis- While thegeneralpublicthoughtthatthesenewstewardesses The nextmodestproposalwasinitiatedbytheCABin1953, Federal Register Flight attendant.Atleastoneflightattendant noticeofahearing.Somethefileshave be includedin the actualaccidentinvestigation? new ideawasbeingconsidered: Couldafemaleflightattendant dants whoparticipatedinthese meetings. gan toadmiretheintelligence andknowledgeoftheflightatten- born. Butperhapsevenmore importantthanthat,thepilotsbe- Thus, anunderstandingofthe problemsofflightattendantswas sured thatflightattendantswerealsopresentatthesemeetings. human factors,andphysicalstandards.Thestaffmember en- that hadtodowithcrashworthiness,fireandrescue,training, member assignedtohelpflightattendantsvariousactivities in areassimilartotheflightattendants,beganinvite staff Chairman ofALPA’stechnicalcommittees,whichhadaninterest human factors. tendant experienceaswelladegreewithstrongemphasis in The staffmemberALPAselectedforthepositionhadflight at- and EngineeringDepartmentinmattersthatwerecloselyrelated. of flightattendantsandtoassisttheothermembersSafety ing withthemembersofS&Divisiontoimprovesafety ALPA relatedtoourorganization. the organizationalchange,wewouldseeachangeinway other division,ofcourse,wasthePilotDivision.Inadditionto the StewardsandStewardessesDivision(S&S)ofALPA.The organizational structuredramatically.Wewouldnowbecalled completely separatedivisionunderALPA.ThischangedtheALPA affiliation withALPA.ALSSA,asmentionedabove,becamea many occupants. shattered fuselage,feedingafiercegasolinefireandquicklytrapping the rupturedfueltanksofrightwingflowingdownandunder age cametorestwiththerightwingelevated.Thisresultedinfuelfrom cartwheeled upasmallincline,theleftwingdisintegratedandwreck- in thefinalreportregardingcabinstates: designated assurvivableornon-survivable.Theonlyreference or participatedinsomeportionoftheflight.Theflightwasnot flight, aswellquestioningpersonswhomayhaveseen,heard, of theflightandcrewactivityinfinalphases ALSSA. Thisgroupfocusedonthefactsconcerninghistory the record,thiswasamalewhoparticipatedingroupfor member oftheFlightOperationsandWitnessGroup.Justfor ards andStewardessAssociation(ALSSA)listedasanofficial a memberoftheNationalSafetyCommitteeAirLineStew- pilots survived.Theaircraftwasdestroyedbyimpactandfire.”(4) 1956. “Thehostessand21ofthe33passengerswerekilled;both following takeofffromtheGreaterPittsburghAirportonApril1, Flight 400,aMartin404thatcrashedandburnedimmediately ticipated inanaccidentinvestigationitwasTransWorldAirlines As flightattendantscontinued toassistinthesemeetings,a The ALPASafetyandEngineeringDepartmentthe ALPA hiredastaffmemberfortheexpresspurposeofwork- In 1960,myunionwentthroughfurthertransitionswithour Although wewerestillunderthewingofALPA,wouldhave ”(5) ISASI 2003 “As theaircraft Proceedings • 65 ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS both divisions. Itcontainedinformationabout contacts,whatto vestigation manualthatwasfollowed bysafetyrepresentativesof coordinate withALPAheadquarters. ALPAhadanaccidentin- cide whoneededtogothe accidentinvestigationandwould the appropriateALPArepresentative andthatpersonwouldde- dure wouldgosomethinglike this:Thecompanywouldnotify mittee membersandweregladtogetit. seniority, theycouldusetheguidanceofPilotSafety Com- days andasmostflightattendantsdidnotholdmanyyears of the timethisworkedquitewell.Theairlinesweresmallerin those to workcloselywiththePilotDivisionSafetyChairmen.Most of the ALPAS&Division.SafetyChairmenweresupposed attendants assignedtotheaircarrierinvolvedweremembers of Division toeachsurvivableaccidentinvestigationwherethe flight ing bothpilots,werefatallyinjuredinthisaccident. seat panwasdeformed.Sevenofthe43personsaboard,includ- from theirattachments.Mostseatswerethrownfreeofthewreckage.” lowing theprincipalimpact,all20doublepassengerseatsweretornfree the centerwingpanel,sustainingonlyinteriordamage.Duringandfol- and compressedmetal.Thecentersectionremainedintactattachedto report reads: not cabinissues.Thereferencetotheinteriorinfinal ones duringthetime:Itfocusedoncauseofaccidentand on theaircraft,finalCABreportwassimilartomanyother sive examinationanddocumentationofseatdamagetoevery the pilotsandotheraccidentinvestigators. had shifted.Herknowledgeandherprofessionalismimpressed pilot investigatorstookpicturesoftheseatandluggagethat the callsherushedouttoaccidentscene.Sheandoneof ardess” waslivinginNewYorkandavailable,sowhenshegot across eachshelfforrestraintofthebaggage.TheUnited“stew- back andnexttoacarry-onbaggagerackwithonlyonestrap Human FactorsInvestigativeGroup. accident investigationandwaslistedasanofficialmemberofthe July 2,1963.UnitedflightattendantIrisPetersonhelpedinthe the MohawkAirlines,Martin404accidentatRochester,N.Y.,on part oftheaccidentinvestigationteamwasat the majorproblemswasflightattendantseatsandrestraints. lems thatwererelatedtoflightattendantsandtheirjob.Oneof possessed necessary,firsthandknowledgeofmanysafetyprob- that webelievecandealwiththetaskathand. for thetask,aswellcommonsenseandanemotionalcharacter ensure thattheyhavetheuniqueknowledgeandtrainingrequired guidelines, beforewesendflightattendantstoacrashsite lieve thisisgenderspecific.Forthatreason,underourcurrent withstand theemotionalimpactofacrashsite,butIdonotbe- the flightattendantswerewomen. be toohorribleforanywomantosee,andatthetime,mostof nalistic andfeltthatsomeofthegrislyaccidentsceneswouldjust accident investigators,andcompanypersonnelwereverypater- 66 In theidealsituationforanaccident investigation,theproce- From thenon,ALPAtriedtosendamemberoftheS & S The flightattendantseatonthisaccidentdidnotfail,butthe While theHumanFactorsInvestigativeGroupdidaveryexten- The Martin404hadaflightattendantseatallthewayin As farasweknow,thefirsttimeafemaleflightattendantwas There wasalsoagrowingrecognitionthattheflightattendants Today, wedorecognizethatnotallindividualswillbeableto The ideawasmetwithresistance.Manypilots,government • ISA SI 2003 “The forwardsectionwasreducedtoamassoftorn,twisted, P r oceedings (6) Equipment would comelooseandcouldcause aninjuryorblock latches usedtorestraintheequipment werenotstrongenough. valuable. participated. Overtheyears, thiswastoprovemoreand made withthegovernmentaccident investigatorsandotherswho ally participateintheaccident investigationwasthecontactthey problems andhighdeathratesofthestewardesses. around todealwithanyofthosethingsconsideringallthe cabin was averygoodchancethatthestewardessnotgoing tobe emergency andevacuationpurposes.Yetduringthoseyears,there of it,andtalktotheflightattendantabouthowitworked. the accidentsitewaslookatflightattendantseat,takepictures craft. Sothefirstthingflightattendantsdidwhenthey gotto flight attendantswereseatedinloungeareasonthedifferent air- McDonnell DouglasDC-8wasinacloset,alsosidefacing.Other dant wassidefacing.Oneoftheflightattendantseatson but littleconsiderationwasgiventothefactthatflightatten- faced amirror.Nowthemirrorwassupposedtobeshatterproof, facing, withnoshoulderharnessandwheretheflightattendant for galleyequipment. tion ofthepassengers,andincreasedrestraintmechanisms tions, theneedforbetterevacuationtraining,educa- tions regardingflightattendantjumpseatdurabilityandloca- ternal documents.Ofcourse,wehadwrittenrecommenda- loose duringanemergency. ley itemsthatenabledtheseheavyordangeroustocome Other concernswerethedeficientlatchingmechanismsongal- shoulder harnesses,nordidtheyhavepaddingforthehead. Mohawk accident.Infact,mostflightattendantseatsdidnothave just plainawfulplacessuchastheflightattendantseatin ing, ortuckedintostorageareas.Manyofthemwerelocatedin dant seatstuckedincorners,attachedtocockpitdoors,sidefac- pact onthem—eveninnormalflight.Therewereflightatten- attendant seatsthatliterallyfelloffthewallwithlittleornoim- on flightattendantseatsandrestraintsystems.Therewere our earlyemphasisduringanaccidentinvestigationwasmainly deaths wastheflightattendantseatingarrangement.So,ofcourse on thelandingphase.Theprimarycauseoftheseinjuriesand severely injured,and48percentwerekilledorinjured volved insurvivableaccidentsontakeoffwereeitherkilledor between 1964and1970,43percentoftheflightattendantsin- cording totheNationalTransportationSafetyBoardstatistics, fect theflightattendant’ssafety.Andtherewerequiteafew.Ac- dants werestillpartoftheALPAinvestigationteam. rival attheaccidentlocation.Thebottomlineisthatflightatten- was thepersontowhomflightattendantreporteduponar- the ALPAstaff,andS&divisionwereconcerned;this site wasinchargeofthataccidentinvestigationasfarALPA, investigation. However,theALPArepresentativewhowasat staff wouldgototheaccidentsiteandparticipatein dent investigators,howtodealwiththepress,etc. do whenyougotthere,howtodealwiththegovernmentacci- For example,flightattendants discoveredthatsomeofthe One ofthemajorbenefitshavingflightattendantsactu- We allknowthataflightattendantisonboardanaircraft for There wasoneflightattendantseatthatinthelavatory,side During theseearlyyears,otherconcernswerenotedinin- At first,theflightattendantsfocusedonthingsthatcouldaf- Sometimes amemberoftheALPASafetyandEngineering authorities that thecabinconditionspriorto and afterasurvivable ant investigationcausedacomplete realizationbyappropriate knowledge anduniquequalifications. Thistragedyanditsresult- gation ofthissurvivableaccident andprovedthevalueoftheir one fromUnitedand Braniff,participatedintheinvesti- It wasperhapscoincidentalthat twowell-qualifiedflightattendants, cluded inthegovernmentreportsonaccidents. port onthataccident.Sometimestheinformationwasalso in- the overallreportofALPACrashworthinessCommittee re- flight attendantinterviews.Thisinformationwouldbepart of ment. Theyalsobegantoputmoreandemphasison the lish thingssuchasthelocationofsafetyandemergencyequip- they begantowalkthroughthecabinofairplaneand estab- about thenumberofflightattendants?Whatslides? needed? Whatabouttheproceduresthatairlinesused? What airplanes. Weretheseplanesactuallysafe?moreexits about wheretheindustrywasgoingwithlargerandlonger tion demonstrationswasbasedontheconcernofalotpeople changed.) Ibelievethereasonforthisrequirementevacua- some oftherulesregardinginstallationinflatableslideswere in 120seconds.(Thiswaschangedto90seconds1967when evacuation demonstrationswheretheaircraftmustbeevacuated craft evacuations.In1964,theFAAstartedrequiringfull-scale themselves. out ofairplanesandmanywerejumpingfromdoorshurting thing elsewasneeded.Peoplewerehavingahardtimegetting held them.Itwasfastbecomingobvioustoeveryonethatsome- they weremeninthosedays)climbeddowntheslideandthen Remember thatfornon-inflatableslidesanable-bodiedman(yes, gers, andtheyaskedthequestionsabouthowslidesworked. to interviewthesurvivors,especiallyflightattendants. the accident.Theywouldgowithgovernmentinvestigators attendants andtheirknowledgecouldbetotheinvestigationof colors. Soaccidentinvestigatorsrealizedhowimportanttheflight that iftheairplanewasdark,noonewouldbeabletosee fastened tothesamecolorof“eyes”onairplane.Nevermind them. Therewerestrapsthatcolor-codedhadtobe aircraft werenotinflatable.Itwasquiteanoperationtoinstall dures were,theycouldhelpfigureoutiftheseproceduresworked. attendants assignedtotheinvolvedcarrierknewwhatproce- pened duringtheactualaircraftevacuation.Becauseflight accident investigationteamtogetmoreinvolvedinwhathap- one flightattendantwouldgotothesite.ThisenabledALPA trained. Theoperatingmechanismwasjustnotthesame. open didnotlikeanyofthedoorsonwhichshehadbeen model ofthistypeaircraft,andyetthedoorthatshetriedto adequate ortheirtrainingwasnotsufficient. attendants involvedintheaccidentfeltprocedureswerenot these discoveries.Insomecases,itbecameapparentthattheflight times, therestraintofequipmentwaschangedbecause ment wereimpossibletoremovefromitsstowagearea.Many an exit.Conversely,itwasalsodiscoveredthatsomeoftheequip- In 1965,UnitedAirlineshada tragicaccidentinSaltLakeCity. As theflightattendantsgainedexperienceataccidentsites, In theearly1960s,therewasalotofworkbeingdoneonair- They helpedto“map”theseatsandegressroutesforpassen- In theearlyyears,slidesthatwereusuallyinstalledon As thingsprogressed,therewereaccidentswheremorethan In onecase,theflightattendanthadonlybeentrainedon emergency lighting systemberewiredsothat alossofelectrical ceptible tocombustion.The final reportalsosuggestedthatthe requiring neweraircrafttobe fittedwithnewermaterialslesssus- grams underwaybe“pressed withvigor.”Thisshouldinclude They alsorecommendedthat crashfire-preventionresearchpro- the positioningofstewardesses nearexitsshouldbereviewed. adopted byUALasstandardprocedureonallB-727flights.” near emergencyexitsforalltakeoffsandlandings.Thispracticehas been lowing theaccident,stewardessesrecommendedthattheybe seated stewardess fromreachingherassigneddutystationforevacuation. Fol- with whichthepassengersprogressedtowardexitsthatprevented the cluded theescapeofeverypassenger.Oncontrary,itwasspeed crash didnotproduceanytraumaticinjurieswhichwouldhave pre- that hadburnedthroughtheaftcabinwall. trol thefire.Theyeventuallywererescuedthroughalarge hole tually afirehosewaspassedintothethreesurvivorstohelp con- pounding onthefuselageandyellingtofiremenoutside.Even- another exitbutwereblockedbyflamesandsmoke.Theybegan She andtwootherpassengerstriedtoreturnthecabinseek could beusedforevacuation,butitopenedonlyseveralinches. it. Bothwereoutsidetheaircraftattime. heading Unitedstewardesswasabletoinstructamanactivate gers. However,thegalleydoorslidewasnotinflateduntiladead- second officer. gers pressingintothearea.Thedoorwasfinallyopenedby the forwardmainloadingdoorwerehamperedbypassen- to beconsumedbyfire.Theattemptsofthestewardessopen The fuelignited,causinganentiresectionofroofandcabinarea landing gearassemblyandrupturedfuellinesontherightside. ing 41occupantswereovercomebydensesmoke,heat,orflames. accident.(7) gers whosuccumbedinthehospitalseveraldaysafter gers andsixcrew,therewere43fatalities,includingtwopassen- ing inSaltLakeCity,Utah,onNov.11,1965.Ofthe85passen- as aresult. changes weremade,butnooneknowsthenumbersoflivessaved mendations regardingtheMohawkaccident.Weknowthese team’s recommendations,aswellsomeoftheinternalrecom- procedures andequipmentareineffecttodayasaresultofthis accident hadadirectbearingonthenumberofsurvivors. United AirlinesFlight227. At theendofCABreport, theBoardrecommendedthat The aftstewardesstriedtoopentheventralstairwayseeifit The galleydoorandoverwingexitswereopenedbypassen- The impactproducedalargeholefromtherightmain Fifty occupantssuccessfullyevacuatedthisaircraft;theremain- United AirlinesFlight227crashedduringanattemptedland- Perhaps itwasalsocoincidentalthatmanychangesincabin ISASI 2003 “The impactofthe Proceedings (8) • 67 ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS organization was jumpseatsplacedingalleys withlargecontain- tation oftheproblems. proved throughouttheyearsresulting fromcontinueddocumen- survival, theflightattendant seat, thereareotherissuesthatim- protect ourmembers,wewould getbetterjumpseatprotections. floor levelemergencyexit.Finally,after16yearsoffighting to harness wouldberequired,aswelllocatednearanapproved on cockpitdoorsillegal.Seatswithasafetybeltandshoulder until August1975. tion, andmoreletters,aproposedrulechangewasnot issued that side-facingseatsshouldnotbeusedbecauseofpoorprotec- request toCAMIforseattesting,a1972U.S.Armyreport saying flight attendantseats,grievancesregardingside-facing a after additionalaccidents,arequestforrulemakingregarding with thecurrentregulations.Andclockkeptticking.Even be addressedbutthatingeneraltheseatswerecompliance against thewallinFairchildHillerFH227Baircraft. prohibiting furtheruseofallaft-facingstewardess’seatsmounted After theaccident,FAAissuedanairworthinessdirective other eventinwhichoneofourflightattendantslostherlife.(10) while thoseintheirjumpseatswere. signed toapassengerseatintheforwardcabinwasnotinjured, rarily renderingthemunconscious.Theoneflightattendantas- rafts andgalleyequipmentfellontheflightattendants,tempo- flight attendant’sclothtometalseatbeltopenedonimpact.Life under herandshebecameentangledinseatbelt.(9)Another age, Alaska,in1970,oneflightattendant’sjumpseatfoldedup the CapitolDC-8accidentthatoccurredontakeoffinAnchor- other accidentsinwhichjumpseatdeficiencieswereanissue.In decided tolookatourjumpseatconcerns,therehadalreadybeen flight attendantshadtositfortakeoffandlanding. handles, looseliquorcarts,andoverheadcoffeeurnsinareaswhere safety belts,harnessandbulkheadproblems,projectingdoor potential hazardsof“snapping”jumpseats,submariningunder for helpwiththisunsafecondition.Wehadsubmittedreportsabout flight attendantseatsin1959.Tenyearslater,wewerestillasking that webeganrequestingimprovementstotheunsafeaspectsof But wehadn’t.InreviewingtheAFAhistoricaldocuments,Inoted dant seatissue,orshouldIsayapoorexcuseforso-called“seat.” fulness ofemergencyequipmentandrestraints. survivors todetermineescaperoutes,anddeterminingtheuse- the pilotstoestablishlevelofcrewcoordination,interviewing there wasaprobleminflightattendanttraining,workingwith the slides,interviewingflightattendants,helpingtodetermineif the flightattendantmanuals,helpingdetermineefficiencyof accident investigatorsaskedthemtodo.Thisincludedreviewing the ALPArepresentativesandbasicallydidwhatgovernment tives toparticipateinaccidentinvestigations.Theyaccompanied accident. lighting. Thiswasbecausethelightshadfailedtoactivateinthis power sourcefornormalcabinlightingactivatedtheemergency 68 Years ago,oneofthemainconcerns oftheflightattendant Although Ionlyconcentrated ononeaspectofaircraftcrash The rulewouldfinallymakeside-facingseatsandmounted The 1972FAAstudystatedtherewereproblemsthatshould The MohawkAirlinescrashinAlbany,N.Y.,1972wasan- By 1972,whentheFederalAviationAdministrationfinally I wouldliketosaythatby1970wehadresolvedtheflightatten- By 1970,itwasroutineforflightattendantsafetyrepresenta- • ISA SI 2003 P r oceedings 2. 1. UnitedAirlineswebsite;united.com;FlightAttendantHistory,TheEx- Footnotes flight attendant.” vestigation becausenooneknowstheaircraftcabinbetterthana tage toeveryonehaveaflightattendantparticipateinanin- investigation oftheMohawkaccident,“Itis,afterall,anadvan- not onlyfornewaircraftbutalsocurrentlyinservice. We mustcontinuetolookaheadfuturesafetyimprovements, mendations canhelpprovidebettercabinsafetyforeveryone. an accidentinvestigationandtheirimportantlife-savingrecom- cific emergencies. the trainingtheyweregivenbyoperatorindealingwithspe- every motionofthataircraftonanormalflight,andtheyknow the aircraftatspecifictimeofaccident.Theyknowalmost ticular segmentofaflight,andwheretheywouldhavebeenin ment. Theflightattendantknowstheservicerequiredforapar- utilized ontheaircraft,andlocationofemergencyequip- worn bytheflightattendants,amountofgalleyequipment and menwhoflytheseaircraftdaily.Theyknowtheuniforms flight attendantsatanaccidentlocation.Theyarethewomen in ensuringthesafetyofalloccupantsonboardanaircraft. concerns willneedtobedocumented.Wemustremainvigilant agement, emergencyresponse,andevacuationscenarios.These new challengesforflightattendantsregardingpassengerman- proaches fallintothiscategory.Thesecouldveryrealisticallypose long stretchaircraft,flyingwings,andotheryet-unimaginedap- Double ortriplepassengerdecks,extremewidebodies,extra- ward intoconceptsthatarenotyetontheproductionschedule. These airplanesarenowbeingdeveloped,andwemustlookfor- craft havethecapacityofcarryingmorethan500passengers. in thejumpseat. back “wall”ofgalleycartsandovensfacingdirectlyatmeasIsit my ownworkingenvironmentonanaircraftthathasentire ers andcartsfacingthem.Thisresearchmademethinkabout 3. 10. MohawkAirlines,Inc.;Fairchild HillerFH-227B,N7818M;Albany,New 9. CapitolInternationalAirways, Inc.;McDonnel-DouglasDC-8-63F, 8. 7. 6. 5. 4. perimental Years. Making ofaUnion Docket No.SA-431, ExhibitNo.6A. York, March3,1972; HumanFactorsGroupChairman’s FactualReport, No. SA-423,Exhibit6A. ber 27,1970;HumanFactorsGroup Chairman’sFactualReport,Docket N4909C; AnchorageInternational Airport,Anchorage,Alaska,Novem- June 3,1966,Released7,1966. ber 11,1965;CABAccidentInvestigation Report,FileNo.1-0032;Adopted June 3,1966,Released7,1966. ber 11,1965;CABAccidentInvestigationReport,FileNo.1-0032;Adopted 13, 1964,ReleasedMay20,1964. 1963; CABAccidentInvestigationReport,FileNo.1-0008;Adopted May September 11,1956,Released14,1956. Airport; CABAccidentInvestigationReport;FileNo.1-0070;Adopted September 11,1956,Released14,1956. Airport; CABAccidentInvestigationReport;FileNo.1-0070;Adopted Organization andthePublicLaw,8/13/93.

United Airlines,Inc.;Boeing727,N7030U;SaltLakeCity,Utah,Novem- United Airlines,Inc.;Boeing727,N7030U;SaltLakeCity,Utah,Novem- Trans WorldAirlines,Inc.,Martin404,N40403;NearGreaterPittsburgh Trans WorldAirlines,Inc.,Martin404,N40403;NearGreaterPittsburgh Mohawk Airlines,Inc.;Martin404,N449A;Rochester,NewYork,July 2, Panter Nielsen,Georgia 8300.10 Change9,Chapter6,TheFAAandFlightStandards:History, As IrisPetersonsaidtomeregardingherparticipationinthe I believeflightattendantsarewellqualifiedtoparticipatein Along thesamelines,itisimportanttorecognizevalueof And whataboutfutureconcerns?Theverylargetransportair- , 1982. N , FromSkyGirltoFlightAttendant:Womenandthe as severalinternationalaccidentsduringhistenurethe Recorder GroupChairmanonallmajoraccidentsinCanadaaswell laboratory, whichhedevelopedfortheBoard.HewasFlight TSB, hewastheheadofflightrecorderandperformance tion SafetyBoardofCanada.Forthelast15yearshiscareerat in 1977andhasworkedformorethan20yearswiththeTransporta- tine flightdata monitoring(FDM)programs (Note:Flightdata aid ofdetailedinvestigation.Today,withairlinesembracing rou- were theonlypeoplewhoexamined flightdataingreatdetail, beginning daysofaviation.T raditionally, accidentinvestigators Flight datavolumeandavailability hascomealongwaysincethe Introduction ing fromeachother. the importanceofairlineandinvestigationcommunities learn- leading. Thepaperwilloutlinesomeofthesepitfallsand stress the numerousassociatedpitfallswherebyanimationscanbe mis- tremely beneficial,investigatorshaveconsiderableexperience with ence onthepartofoperator.Whileflightanimation isex- tendency towantautomatictoolsthatrequirenoorlittleexperi- demand forflightanimationsystemswithintheairlines and a cent trendtoroutinelyanalyzeflightdata,thereisanincreasing in particulartheprosandconsofflightanimation.With re- valuable experiencesrelatedtothelimitationsofflightdataand over manyyears.Thisrelativelysmallgroupofpeoplehasgained lyzing dataforyearswithhighlyspecializedtoolsthathaveevolved authorities withsubstantiveflightrecorderlabshavebeenana- the numeroussimilaritiestoaccidentinvestigation.Investigation analyze flightdataonaregularbasisforaccidentpreventionand This paperwilldiscussthegrowingtrendofairlineswantingto Abstract expertise totheFlightscapeteam. flight recording,andhistechnicalknowledgebringsignificant investigation experience,leadrolesoninternationalcommitteesin for handlingflightdata.Hishands-ondataanalysisand Flightscape maintainsandsupportsRAPSotherproductlines specializing inflightsciencesanddataanalysissystems. joined FlightscapeinFebruary2002,aflightsafetycompany recorder labsandwaseventuallycommercializedbytheTSB.Poole Presentation System(RAPS)thatisusedbymanyStates’flight project thatledtothedevelopmentofRecoveryAnalysis& L aboratory Head.In1985hewasresponsibleforinitiatingthe Accident InvestigationWithout started inthefieldofaircraftaccidentinvestigation Aviation Organization’sFlightRecorderPanel.He expert panelmembertotheInternationalCivil analysis. HerepresentedCanadaasthenational a leadingexpertinthefieldofflightrecorder current pilot’slicenseandisrecognizedworldwideas Mike Poole By MichaelR.Poole(M03278 isaprofessionalengineerwith The Accident R ecorder ) , P.Eng.,ManagingPartnerFlightscape ture oftherecording system.Thisinspiredthe firstgenerationof fraught with“dropouts”orbit errorsduetothemechanicalna- cases openedandrecertified.Copy processestookhoursandwere The recorderhadtoberemoved fromtheaircraftandinsome meant onlypursuingthedata inreactiontoasignificantevent. datory FDRwasbynomeans aneasyprocess.Formany,this was availableordone,andtheparametersetswerefewinnumber. recorders lookingforproblems.Inthesedays,noflightanimation sole jobwastositinfrontofamagnifyingglassandreadout foil operational anomalies.Infact,SASevenhadtwopeople whose routinely analyzingflightdataformaintenance,prevention, and rial bitstreamthatwassenttotheFDR. and databussesontheaircraft“packaging”theminto ase- data sourcefortheFDR,acceptinginputsfromvarioussensors The digitalflightdataacquisitionunit(DFDAU)provided the tional FDR/CVRstandardsatthetime(EurocaeED55&ED56). take advantageofearlychipdesignsthatdidnotmeetinterna- ability requirementsascivilianstandards,therebybeingableto data thancivilianaircraftanddidnothavethesamecrashsurviv- civil aviationbecausethemilitaryrecordedtypicallymuchless civilian aircraft.Themilitarywasabletousesolidstatebefore early 1990sbeforesolid-statememorywasacceptableforusein data storedonmemorychips)inthe1980s,itwasn’tuntil anymore). Althoughthemilitaryintroducedsolidstate(digital digital FDR(therearenoanalogFDRstodaysotheDisnotused netic tapeandtheFDRnamewaschangedtoDFDRdenote (early 1970s),whereflightdataweredigitallyrecordedonmag- tion, andVHFkeying,onatimebase).Thencamethedigitalera (airspeed, magneticheading,pressurealtitude,verticalaccelera- recorder, whichrecordedanalogtracesoffivebasicparameters data isworthwhile.Intheearlydays(1960s)camemetalfoil Before exploringthisissue,abriefrecapoftheevolutionflight History offlightdata “accident investigation”toolsarenotneededforFDM. quire differenttoolsandthereisperhapsamisperceptionthat and accidentinvestigationarenotthesamethereforere- investigation community. ing someofthelessonslearnedfromrelativelysmallaccident dent,” andthattherearesomesignificantbenefitsfromexamin- ways, areperforming“accidentinvestigationwithouttheacci- airline industry.Thispaperwillarguethattheairlines,inmany domain offlightdataanalysisisrapidlybeingdrivenbythelarger for theairlinestouseflightanimationreplaydata, and someEuropeannomenclature)themostrecenttrend assurance [FOQA]isU.S.nomenclature,andFDMCanadian analysis [FDM]isICAOnomenclature,flightoperationsquality Airlines quicklydiscoveredthat toextractdatafromtheman- Some airlines,suchasBritishAirwaysandSAS,werealready A commonstatementIhaveheardlatelyisthatFDMprograms ISASI 2003 Proceedings • 69 ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS fore tendtocater tothelowestcommondenominator. Addition- tity. Parameterrulesmustconsider manyaircrafttypesandthere- accessible andhasnowsurpassed theFDRintermsofdataquan- wrong box!TheQARwasdeveloped becausetheFDRwasnot survived buttheQARdidnot. Thedatawereavailablebutinthe based unit,arguablylessquick accessthantheFDR!TheFDR corder with64words/sec.The QARwasa384word/sectape- is agoodexample.TheSwissair111FDRwassolid-state re- The SwissairFlight111MD-11accidentoffPeggy’sCovein1998 vent ofSSFDRs.Whydidinvestigatorscometodislikethe QAR? thought theQARwouldsimplydieanaturaldeathwith ad- tiating termsinceSSFDRswerealsoquickaccess.Many ofus new namebecause“quickaccess”wasnolongeragooddifferen- early 1990sandIsaidtotheQARmanufacturesthattheyneed a I rememberbeingataEurocaemeetinginWashingtonthe They werealsoreadilydownloadablemakingthem data qualityandFDRreliabilitysincetherewerenomoving parts. introduced. Theadventofsolidstatewasagreatadvancementin were recorded. airline couldreconfigureatwilltodeterminewhichparameters additional voluntaryFDAUwasaddedtotheaircraftthat rally addedtotheQARinstead,andinsomecasesacomplete FDR meantrigorousrecertificationissues.Thedatawerenatu- the airlinebecauseitwasnotaccessible,andanychangesto capacity. IfthedataweretobeaddedFDR,itdidnothelp eter listsisbecauseofalackdataavailability,notFDR lem. Thereasonthatwedidnothavelargermandatoryparam- capacity israrelyanFDRproblem;ratheritacquisitionprob- note (anditisaverycommonmisconception)thattheissueof enough, sotheyaskedifcouldhavemore.Itisimportantto same information. conformed tonostandards(voluntary),andbothrecordedthe one thatconfirmedtorigorousstandards(FDR)and facilitate easyaccesstothedata. identical time. Inthemajorityoftheseearlysystems,FDAUsent could simplypulloutthemediaandsubstituteanotheratany with aremovablemedia(initiallytapeaswell)sothattheairline quick accessrecorders(QARs)intheearly1970s.Theywerebuilt 70 Around thistime,solid-statememorymediarecorderswere The datastreaminthoseearlydayswas,bysomeairlines,not Effectively, airlineshadtworecordersonboardtheaircraft, • ISA SI 2003 datastreamtoboththeFDRandQARsimply P r oceedings “quick access.” Benefits offlight animation the late1980sandthroughout the1990s. past astheybecameincreasingly popularandcontroversialin of animation,andtherehave beenISASIpaperswellintothe commercially available. flight animationisreadilyavailable, andnumeroussystemsare systems atthistimeanditwasrelativelyexpensivetodo.T oday, early 1980s.Airlinesdidnotbecausetherewerenocommercial Accident investigatorshavebeenusingflightanimationsince the Flight animation dent investigation. data forroutinemonitoringofflightthanamajor acci- cording system”’ crash protectedmediumwithoutrequiringrecertificationoftheflight re- recorded data(e.g.,additionaldata,sampleratesorresolutions)inthe protect thecoremandatorylistwhileallowingoperatortochange “… itisrecommendedthatindustryprovideoperatorswithsolutions routine playbackneedsishighlydesirable.” cording systemthatsatisfiesbothaccidentinvestigatorsandoperator’s based wirelessdownloadcapabilities,anintegratedcrashprotectedre- capacities, readilyavailabledataontheaircraft,andaffordableground- ED112—“ tion Committeeistryingtochangehistoryinthisregard. the QAR,anditislongoverduethattheytalktoeachother. those whodealwiththemandatoryFDRand an FDR.Theretendstobetwodifferentgroupsintheindustry, Swissair datagoingtotheQARcouldnothavealsobeen “packaging” issue.Therewasnotechnicalreasonwhyallofthe available totheaccidentinvestigator.Insomewaysitissimplya airlines canroutinelyaccessthedataandsamesetis aircraft, andtheyhavenotreallybeenupdated. developed fromtheolddayswhensensorswerededicatedto from digitaldatabussesareincrediblyreliable;yettheruleswere terpret thisasareasontogroundtheFDRsystem.Parameters not anFDRproblem—itisaircraftproblemyetsomestillin- does notworkontheFDRforanAirbusA320,example,itis simply takingadvantageoftheirreadyavailability.Iftheairspeed exist fortheoperationofaircraft,notFDR.TheFDRis parameters todayarefromadigitaldatabusandthe tory interpretationoftheMEL.Therealityisthat99percent do notwanttoaddparameterstheFDRbecauseofregula- from theFDRbecauseofMELproblems!Operators,stilltoday, 1980s, AirCanadaactuallyremovednon-mandatoryparameters bodies invokedtheMELandgroundedairplane.Inlate rameter becameproblematicduringroutineFDM,regulatory of thepast.IfyouaddedaparametertoFDRandifpa- of theQARdespitelogicdictatingthatitshouldbecomeathing their ownpurposes. hand wenttothetroubleandexpenseofrecordingextradatafor plained aboutthecostsofadditionalparametersandonother meant costlycertificationissues.Airlinesontheonehandcom- affecting themandatorybox.Anychangetobox ally, earlystandards Investigators havelongknown aboutthebenefitsandpitfalls The bottomlineisthatitreallyunacceptabletorecordmore Eurocae ED112andtherecentU.S.FutureFlightDataCollec- It makesmuchmoresensetohaveanintegratedsystemwhereby There weresomeotherfactorsthataffectedthecontinueduse With today’ssolid-statetechnology,significantlyincreased encouraged aseparateboxforfearofadversely • principles • • • • Pitfalls offlightanimation • • • • • • • • • • • collection offlightdataforimprovement intheareasof what istheproblem?First,let’sdefineanFDMprogram. FDM isaprovenconceptandbeingembracedworldwide. So a lotofsensetostudythedata investigate accidents.FDMprogramsareveryvaluableasit makes We allknowandunderstandtheelementsreasonswhy we Accident investigationvs.flightdataanalysisprograms It iseffectively an ITsystemtodistributeobjective information FDM ispartofasafetymanagement system.Itisasystematic Misplaced cr Subjective information F P Lend cr Easy todisseminate Effective trainingtool P Stimulating andeffectivemeansofcommunication F Assimilate comple Risk management T Maintenance Operations Drawing conclusionswithoutunderstandingunderlying abrication acilitate analysis owerful andcompelling raining r etty pictur edibility tofindings e syndr edibility x information ome (seeingisbelieving) before thingsbecomecatastrophic. parameters is such thatoftensignificantjudgment isrequired. limitations ofsamplerate,resolution, accuracy,andnumberof events andunderstanding disseminatingthem.Thecurrent flight animationsareactually quiteusefulforanalyzingcomplex virtually allsoftwareoutthere candothis,itshouldbenotedthat purposes; justhitabuttonand uppopstheanimation.While not needexpertaccidentinvestigators. did notneedtohaveexpertiseanalyzeflightdata,we would one uses,butratheritisafunctionoftheflightdataitself. Ifyou fact isthattheexpertiserequirednotafunctionof“tools” you donotneedtobeanexpertuse“airline”system. The that youhavetobeanexpertuse“investigation”system but automatic toolstoeliminatetheneedforexpertise.Somebelieve necessary forFDMprogramsinthequesttoprovideuserfriendly lieves and/oradvertisesthat“investigation”skills/toolsare not FDM 2isarguablyaccidentinvestigationwithouttheaccident. of thedataleadinguptoeventthatinitiatedanalysis. the aircrafthitsgroundornothasnobearingonanalysis ence inthetechniques,expertise,andtoolsrequired.Whether gued thatonceyouperformthestudy,thereshouldbenodiffer- is anaccident,incident,FDM1event,oraPIREP,itcanbear- • • • FDM 2occurrenceinvestigation • • • • • • FDM 1eventdetection into twodistinctcomponents: accidents. Forclarificationpurposes,IliketobreakFDMdown reliability. Aboveall,itisasystemthatidentifiesprecursorsto to reduceoperationsandsupportcostsimprovedispatch Many airlineswanttoroutinely animateeventsfortraining Unfortunately, thereisacomponentoftheindustrythat be- Regardless astowhetherthestimulistudyaflightsequence Flight animationisveryusefulforr Similar toaccident/incidentinvestigation Examination ofasingleevent(s)ingr Examining dailyflightsinsmalldetail Flight animationnotuseful Outputs statisticaldatabase Until r Automatic detectionofevents R outine monitoringofflightdata ecently , plaguedwithpoorqualitydata outine eventsandcomple eat detail ISASI 2003 Proceedings • 71 x ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS For example,a 64 word/secratewouldrequire thedataprinted be incrementedatintervalscoincident withthedataframerate. tain thetimingresolutionof the originaldata,EUfilemust each otherbyasmuchjust underasecond.Inordertomain- wheel, whilebothsampledonce persecond,willbeoffsetfrom on thewordlocation.Forexample,aileronposition andcontrol the recorded atonesamplepersecondareactuallynotsampled at for analysisisthetimeelement.Twoparametersthatareboth tigators haveaccesstoallofthedatatime. which isarelativelysmallandnicelypackagedfilealready,inves- Since investigationsystemsaccesstheARINCbinarydata file, not liketohaveprejudgewhatIthinkmightbeinterested in. prejudge whatisimportant.AsaformerTSBinvestigator, Ido is compromisedbecausehedoesnothaveallofitand must only wanttolookatacoreset,hisability“investigate”the data you need.Althoughthepersonatotherendmaynormally limitations. Whatistypicallydonetosendonlytheparameters 1,000 columnswide!InfactyoucannotdoitinExceldueto well overathousandparameters.ImagineanExcelspreadsheet pass alloftherecordedparameters.Amodernaircraftmayhave another asaCSVorspreadsheetfile,itbecomesproblematicto actual datarecovery. veloped analysistoolsbutreliedonsomeoneelsetoperformthe poor analysistools.Othercompaniescapitalizedonthisandde- their replaysystems,whiletheywillrecoverthedata,havefairly hundreds ofFDRsforeveryreplaystationtheysell.Consequently, itself. Flightrecordermanufacturesliketosellboxesand ARINC datafromtheaircraftdirectlyisasignificantprocessin analysis/animation system.Thisislargelybecausehandlingthe cessed byanotherapplicationsothatitis“readable”their cannot acceptARINCdataandmustfirsthavethepre-pro- the fly”asrequired.Manysystemsinusebyairlines,however, tions interactwiththesourcebinarydataandconverttoEU“on handle theARINCbitstreamdatadirectly.Thatis,allapplica- totally unawareof.Investigatorsusesystemsthatinteractively data insteadoftherealduetoprocessingthatyoumaybe sis/animation toolmaybeshowingyouanartifactoftherecorded tion systems.TheproblemwithpassingEUfilesisthatyouranaly- ables orspreadsheets)topassthedatatheiranalysis/anima- to useengineeringunits(EU)orCSV(commaseparatedvari- concerns. Oneexampleistheproblematictrendinairlines yond thistraditionalrole. erational decisionsbasedontheirflightdataanalysiswellbe- animation capability.Airlinesaregoingtoincreasinglymakeop- operational analysisofevents,partlyinspiredbyreadilyavailable nance andonlyinrecentyearshavetheybeenusedfordetailed simply lostintheprocess. pertise tousecomesasignificantdangerthatthejudgmentis tion andmarketingofproductsasautomaticrequiringlittleex- of thestepsinvestigatorsperformedmanually.Withthisautoma- in withveryattractivetoolsthathaveinternallyautomatedmany used toperformthejobgrewtogether.Todayairlinecanjump and 80stheirexperienceinflightdataanalysisthetools Accident investigatorsgrewupwithlousytoolsinthe60s,70s, 72 Another moreseriousproblemwithpassingEUfilesaround When youhavetopassEUfilesfromoneprocessorsystem There aremanytechnicalexamplesthatillustratesomeofthe Airline playbacksystemswereoriginallydesignedformainte- • same ISA SI 2003 timewithinthesecond.There isarelativeoffsetbased P r oceedings the EUconversion processiftheysuspectaproblem. TheEUcon- get thepropertimeresolution, anddonothavetheabilitytocheck the analystsatotherenddo notgetalltheparameters,do assistance introubleshootingsomething. TheysendaCSVfileand problem asairlineswillfrequently wishtosenddatathemfor be anditisveryhardtoundo. of howhistoryhasgotustoaplacethatwedonotreallywant to it shouldn’tbe,and,liketheQARdilemma,isanotherexample gent standards,whyisitacceptableattheairlines?Theanswer is tigators arenotwillingtocompromisedataqualityandhave strin- is accidentinvestigationwithouttheand inves- investigation authorityevenhasthedata.Ifwebelievethat FDM airline hasrunincidentsthroughits“automatic”toolsbefore the they havetodothework.Thishasalreadyhappenedwhere an event, itishumannaturethattheteamwilltrytouse tools event. However,shouldtheteamcomeacrossamorecomplex animation isbeingusedtolookatarelativelysimple,routine ing processesgoingoninternally(awholeotherpaper),andthe data areinaccurateinthetimedomainastherelotsofsmooth- will displaythedataatpreciselytimesitwasrecorded. ARINC dataontheflydonotsufferfromthisproblemandthey expertise inflightdataanalysis.Systemsthatcanprocessthe This issimplyunacceptableforaccidentinvestigatorswhohave already reducedtheaccuracywellbeyondlatencyconcerns. second, thereisnopointinworryingaboutlatency—youhave all ofthesesystemsouttherethattruncatethedatatonearest talking aboutlatencieswithinthesecondformostpart.For how importantitisthatwebeabletotracedatalatency.Theyare NTSB andotherinvestigationagencieshavegivenpaperson needs andbytruncatingthedataalltonearestsecond.The prejudging whatparameterstheanalysisoranimationsystem bers getevenmoreimpractical.Instead,shortcutsaretakenby wide! Ifyoumovetoa256or512word/secrecording,thenum- require aspreadsheet5,760,000lineslongand1,000+columns in anEUfileformatmaintainingtherecordedaccuracywould each second.Topass25hoursofalltheflightdatatosomeone hours ofdatausingEUfiles,youwouldneed64linesfor for eachparameter.Thismeansthatifyouwanttolookat25 out in1/64timeintervalstomaintainthesameresolution Aircraft manufacturesarealso becomingawareofthisgrowing While inmanyflightanimations,itwillnotmatterthatthe results. Withtheaccessibilityof“automatic”flightanimationsys- accident investigationistogetitrightbeforedisseminatingthe hard tobacktrackoncepeoplehaveseenit.Thegoldenruleof done veryquickly.Ifitisnotaccurateormisleading, stakes havegoneupasairlinescannowflightanimation lar. WiththeFDMevolution,ICAOwillneedtorevisitthisas investigation. ThiswaswrittenbeforeFDMprogramswerepopu- with significantrecorderlabsfeltthatthiscouldcompromisean corders toairlinefacilitiesafteramajoraccident,andotherStates ICAO FLIRECPanelbecausesomeStatesstartedtakingthere- mends Statesuseinvestigationfacilities.Thiswaswrittenbythe an “airline”facilityand“investigation”’recom- ICAO Annex13AppendixDrecognizesthedifferencebetween Summary data inwhichtheyhavenowayofassessingitsvalidity. that theairlinespleasesendrawdata,notsomeartifactof sense tobepassingEUfiles,andmanufacturersarestartingask smaller tosendandhasnocompromises,itdoesnotmakemuch the EUdataasfactual.SinceARINCfileismagnitudes parameters infrequentlyanalyzed,andoneshouldneveraccept version processhasmanyopportunitiesforerror,especiallywith derstanding oftheprocessyouareoperating. the datawithrespectanddevelopanexpertisethoroughun- life, nothingisfree.Thepropersolutiontomakesureyoutreat and willyouknowwhenhavecrossedit?Likemostthingsin When isthetransitionwherebyinvestigationexpertrequired, trivial forroutineeventsbutwillnotbecomplexevents. investigation partinthemiddle!Themaybe right—you can’talwaysjumpfromthedatatotrainingwith using animationsfortraining,youstillneedtomakesurethatitis demand thesametoolsandrequireexpertise.Ifyouare path theinvestigationlabshavealreadygonedownandeventually of FDMprograms,airlineswillalmostcertainlygodownthesame rest oftheindustry. evolved fromyearsofaccidentinvestigationareappliedtothe dards ordatarecording,extractionandprocessingthathave flight data,itwillbecomeincreasinglyimportantthatsimilarstan- to compromisethisgoldenrule. any expertisetogenerateanimations,wearesettingourselvesup combined withphilosophiesthatpurportyoudonotneed tems andthemannerinwhichsomesystemsprocessdata, With flightanimationbecomingamoreandpopularpart As airlinesmakemoreanddecisionsbasedonroutine ISASI 2003

N Proceedings • 73 ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS 74 • ISA SI 2003 P r oceedings SESSION IV DC-7 collidedover theGrandCanyonresulting inthelossof128 June 30,1956,aTWASuper Constellation andaUnitedAirlines air trafficcontrollersarerequired todotheirjobs.Forexample,on trol areas,andradiofixesasrequired reportingpoints. tem wasestablishedwithcontrolled airports,airwaytrafficcon- regulatory codesandairtraffic rules,andtheCivilAirwaysSys- more centers.TheCivilAeronauticsActof1938established new sumed theoperationofthreecentersandestablished five year. Finally,onJuly6,1936,theUnitedStatesgovernment as- Chicago, andathirdcenterinClevelandJunethefollowing trol CenterinNewark,N.J.;asecondcenterwasestablished in ber of1935,theairlinesestablishedfirstAirwayTrafficCon- gated viaradiobeacons,ranges,andbonfires.InDecem- grass stripwithtwowandsandawheelbarrow,airplanes navi- has comealongwaysinceArchieLeaguestoodattheend ofa of almost80yearsgovernmentdirection.Airtrafficcontrol the airtrafficcontrolcommunityalsocelebratesananniversary the Wrightbrothers’inauguralflight,itisinterestingtonotethat late, youweren’tnauseatedbythefood.” get hijacked,youdidn’tcrash,throwup,weren’t be disastrous,soyoudefineagoodflightbynegatives:didn’t to sayaboutmostairplanejourneys.Anythingremarkablemust was bestcapturedbyPaulTheroux,whosaid:“Thereisnotmuch and accidentinvestigation.Theinterestingthingaboutaviation provide expertreal-timeknowledgetoaidinaircraftincident maintains acadreofspeciallytrainedairtrafficcontrollersthat through itsownAirSafetyInvestigatorsProgram.ThisProgram tion safetyandiscommittedtoaircraftaccidentinvestigation provide thesafestairtrafficcontrolsysteminworld. strive toimproveandenhanceaviationsafety,weproudly mus testagainstwhichallourdecisionsarebased.Wecontinually the flyingpublic,aswell.Ourmotto,SafetyAboveAll,islit- the interestsofourmembership,butalsosafety safety professionals.NATCAisveryproudtorepresentnotonly members thatincludeengineers,architects,andotheraviation Puerto Rico,andGuam,alongwith2,500otherbargainingunit than 15,000airtrafficcontrollersthroughouttheUnitedStates, around thenation.Today,NATCAhasgrowntorepresentmore sure thesafetyandlongevityofairtrafficcontrollerpositions Throughout history,aviation accidentshavechangedtheway As theaviationcommunitycelebrates100thanniversaryof First andforemost,NATCAiscommittedtopromotingavia- By JohnCarr,President,NationalAirTrafficControllersAssociation,KeynoteSpeaker Controllers Associationwascharteredtoen- tion safetyprofessionals. fore thisverydistinguishedgatheringofavia- privilege torepresentNATCAandspeakbe- Controllers Association.Itisanhonorand am thepresidentofNationalAirTraffic Good afternoon.MynameisJohnCarr,andI Growth ofATCSystemand Founded in1987,theNationalAirTraffic Controllers Union by government employees.Overthenextfew weeks,thiscritically with thepublicinterestsomuch soastomandateitsperformance affects thelivesofeveryone.Air trafficcontrolisintrinsicallylinked inherently governmentalfunction, whichdirectlyandsignificantly of theairtrafficworkforce.In ourview,airtrafficcontrolisan controllers intheUnitedStates todayispreventingtheprivatization importantly, privatization.The mosturgentissuefacingairtraffic ment, staffingshortages,aviationsecurity,laborrelations,and, most States airtrafficcontrolsystemincludemodernizationofequip- ing theairtrafficcontrolsystem. cost aidindeterminingthestateofaviationindustryinclud- fic controlsystem.Safetyrecords,dependability,convenience, and cidents andaccidentsoftenhighlightcriticalissuesintheair traf- see theconsequencesofanoverloadedsystemdaily.Aviationin- Tampa fordinnertonight. to youfolks.WhenI’mdone,I’llgetonaplaneandbebackin Tampa, atebreakfastwithmyfamily,andgotonaplanetospeak frastructure ofthe21stcentury.Iflewtoworktoday.wokeupin tion toaviationsafetyhasmadeairtravelthetransportationin- not compromisingorreducingthemarginofsafety.Yourdedica- prove theairtrafficsystembyincreasingcapacitywhile challenge facingtheaviationcommunityistocontinueim- with airtrafficserviceproviders—thecontrollers.One responsibility tomeetthisincreasingdemandforairtravellies all experiencedthedelaysontaxiwaysorramps.Muchof tures ofendlessrowsairplanesqueuedforrunways.Wehave system tonearcapacityinrecentyears.Wehaveallseenthepic- minimum acceptableperformancestandard. ery controllerknowsthereisnoroomforerror.Perfectionthe while workinginstressful,high-energyenvironmentswhereev- the safetyofmorethanonemillionaviationpassengersperday the marginofsafety.Thisnation’sairtrafficcontrollersensure the abilitytobeflexibleyetuncompromisingwithoutreducing controller’s decision-makingprocessrequiresquickthinkingand airport utilizationbyallclassesofairtraffic.Antraffic at thesametimefacilitatemaximumefficiencyinairspaceand critical occupation.Theypreventcollisionsbetweenaircraft,and ernment enactingchangesinairtrafficprocedures. on theapproachcharts.Bothaccidentsresultedinfederalgov- and identifiedinadequaciesinthedepictionofaltituderestrictions on unpublishedrouteswithoutclearlydefinedminimumaltitudes cident identifiedalackofclarityinATCprocedureswhenflying crashed intohighterrainonapproachtoDullesAirport.Thisac- the airtrafficcontrolsystem.InDecemberof1974,aTWAB-727 creased governmentregulationofairroutesandmodernization lives. Althoughtragic,thisaccidenthighlightedtheneedforin- In NATCA’sopinion,theutmostconcernsfacingUnited Air trafficcontrollersare,bycharacter,safetyminded,andthey The increaseddemandforairtravelhasbroughttheentire Air trafficcontrollersserveinaunique,complex,andsafety- ISASI 2003 Proceedings • 75 ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS with yourcoffee.Onpage13ofthe investigation orworktopreventfutureincidents? sight andpublicview.Withoutthetapes,howdoyouconductan air occursatacontracttower,thetapesarehiddenfromFAAover- to thepublicviaFreedomofInformationAct,yetwhenamid- why tapesandrecordsofmidaircollisionsatFAAfacilitiesareopen oversight oftheContractTowerProgram,FAAcannotanswer involving contracttowers. controllers weredeniedpartystatustoinvestigationsofincidents scenario wouldholdtrue.Infact,therehavebeeninstanceswhere no sense.IntheUnitedStates,ourobviousfearisthatsame pate intheinvestigativeprocess.Ofcourse,thismakesabsolutely management officialsratherthanfront-linecontrollerspartici- In Canada,inspiteofthewishesbyCanadiansafetyboard, fic controlsystems,andhowtheyworkwithinvestigativebodies. are problemsthatmuddythewaters. But whenyoudealwithprivatizedairtrafficcontrolsystems,there ternal influenceandblamefocusonaccidentprevention. investigations ofairtrafficincidentsremainindependentex- put profitsoversafetyandthatisunacceptable. control system.Simplyput,privatizationofourindustrystandsto stand behindtheirvotesandsupportthesafetyofourairtraffic stance ontheissueofprivatizationandhaveaskedlawmakersto NATCA hasbeenworkingthismonthtobuildsupportforour important subjectwillagainbeaddressedintheU.S.Congress. 76 ing, airtrafficcontrollersallovertheUnitedStateslanded more der toshutdowntheNationalAirspaceSystemat9:45that morn- exceptional dedicationwhileperformingtheirjobsflawlessly. lers, truechampions,neverlosttheircomposureandmaintained dous actsinUnitedStateshistory,thenation’sairtrafficcontrol- flew. However,duringthissingleevent,oneofthemosthorren- truly theworstdayin100yearssinceWrightbrothersfirst over thisnationthatmorning.Thatfatefuldaymaystandaloneas events ofSept.11,2001,andtheroleairtrafficcontrollersall look atthepaper. If weneededanyevidencethatprofitrunscontrarytosafety,just 1990s ledtoanabdicationofresponsibilityforoverseeingsafety. today saysapushtowardprivatizationthatbeganinthemid- Had PotentialConflict.”Inthe 27, 2003],theheadlinereads“SafetyversusProfit—Contractors I’m sureyousawacopyofthemajornewspapersthismorning Here’s somethingelsetoconsider:Whileclaimingmaintain We havebeenwatchingCanada,andotherprivatizedairtraf- Well, Idon’thavetotellanyofyouthatitisessential How doesthataffectyou,youmightask? When TransportationSecretaryNormanMinetaissuedtheor- I wouldliketonowtakeafewmomentsdiscussthetragic • ISA SI 2003 P r oceedings Wall StreetJournal Washington Post , thebigstory today[Aug. draw! • boy, butnotforonewhostillis. • proach platesinthecar.Theairspeedyoudon’thave. Runway behindyou.Fuelinthetruck.Halfasecondago.Ap- • old, boldpilots! • back. • of yourtakeoffs. • • • journey intothenext100yearsofaviationhistory. continue tofightforsafetyaboveallelseaswebegintheexciting The safetyofourskiesisnotforsale,toanyone,andwewill als whoputsafetyaboveallelseandholditastheirsacredtrust. today thatbusinessisgood.Irepresentthousandsofprofession- fessionalism. plished afeatneverimaginedwithskill,determinationandpro- not onlywitnessedthetragedyofSept.11,2001,butalsoaccom- traffic controlsystemintheworldwasnevercompromised. ever; however,theefficiencyofsafestandmosteffectiveair spirit ofairtravelchangedinacataclysmicandabruptwayfor- flying publichascometodepend.Onthatinfamousday,the example ofthevaluedteamworkuponwhichFAAand and thecommitmentprofessionalismdisplayedwasatrue incredibly dynamicandexhaustingconditionsduringthattime, fusing airspacerestrictions.Everyoneworkedintenselyunder daily changesinproceduresandrapidlychangingoftencon- initial weeksaftertheattack,controllerscontendedwithalmost tional AirspaceSystemhasincreasedsignificantly.Duringthe never beforeaccomplishedinthehistoryofaviation. loss oflife,furtherdemonstratinganoutstandingachievement one. Theirextraordinaryactionsmostlikelypreventedadditional fore practiced,trainedorimagined,testedtheresolveofevery- errors. Thisunprecedentedchallenge,anundertakingneverbe- hours, controllerssafelyguidedanother4,000airplaneswithno effectively reroutingoneaircrafteverysecond.Overthenextfour rected everyaircrafttolandatthenearestairportimmediately, than 700airplaneswithinfourminutes.Airtrafficpersonneldi- I’d liketoleaveyouwithacoupleofclosingthoughts. I liketosaythatsafetyismybusiness,andI’mheretellyou On thatfatefulday,thesemenandwomenofpublicservice As aresultoftheeventsthatday,complexityNa- And finally Flying istheperfectvocationforamanwhowantstofeellik Things thatdoyounogoodinaviation:Altitudeaboveyou. Ther Y T K R ou cannotpr ry tok eep lookingar emember N e ar eep thenumberofyourlandingsequalto e oldpilots,andther , gravityneverloses!Thebestyoucanhopeforisa , you’r opel yourselffor ound; ther e alwaysastudentinanairplane. e’s alwayssomethingyou’vemissed. e ar war e boldpilots,butther d bypattingyourselfonthe e ar e no e a In reality,wearebuilding ontheworkofthosewhocame before. who believethatcrashworthiness isasomewhatnewapproach. derstand howwegottowhere weare.Therearesomepeople gathering thedata,evidence, andinformationnecessarytoun- respects. Westartfromtoday andworkbackwardsthroughtime, Accident investigationislike thedisciplineofhistoryinmany In thebeginning exist forsurvivaltobeaccomplished. throughout theimpactandbeyond. impact sequence,ANDtheenvironmentremainslivable tolerance, ANDoccupiablespaceismaintainedthrough the impact forcesthatreachtheoccupantsremainwithinhuman and astartingpoint.Asurvivableaccidentisoneinwhich protect theoccupantsofaircraftinaccidentswecannotstop. Only thencanwelearnandaccomplishthethingsnecessary to crashworthiness, wemustassumethatWILLhaveaccidents. accident investigationisfocusedonstoppingtheaccident.Todo fundamental paradigmshift.Muchofwhatwedoinsafetyand outcome. Forsomefolks(includinginvestigators),thisrequiresa in crashworthinesswilleverpreventanaccident,onlychangethe by whichweMITIGATEtheeffectsofaccidents.Nothingdo to MITIGATION.“Crashworthiness”isthetechnologyandmeans and findyour“safety”switch.”NowflipitfromPREVENTION to improvesurvivabilityinaviationaccidents. T tional SocietyofAirSafetyInvestigators. print media.HeispresidentoftheArizonaChapterInterna- articles onaviationsafetyandisfrequentlyquotedinelectronic investigation formorethan25years.Hehasauthored50 A properdefinitionofaccidentsurvivabilitygivesusbackground To startwith,Ineedeveryonetoreachbehindyourrightear Through CrashworthinessEvaluation provide theopportunitytochangedesignorprocedure worthiness investigation,andhowsuchinvestigations his paperwillexaminethehistoryandbeginningsofcrash- Enhanced OccupantProtection Crashworthiness Investigation: been involvedinaviation,safety,andaccident pilot andretiredU.S.CoastGuardofficerhas investigation andcrashsurvivability.Heisanactive graduate andundergraduatecoursesinaccident Aeronautical UniversityinPrescott,Ariz.Heteaches Aerospace SafetyEducationatEmbry-Riddle Science andistheAssociateDirectorofCenterfor Bill Waldock And AdvancesinDesign— A ViewfromtheWreckage holdstherankofProfessorSafety AllthreesituationsMUST By WilliamD.Waldock the thereby exposeobjectsinsideittodamage.” should notcollapseunderexpectedconditionsofforceand 1. “Thepackageshouldnotopenupandspillitscontents worthy design. ciples,” firstpublishedin1950,providethebasisforanycrash- founded thebeliefsystemthatweusetoday.His“packagingprin- himself. ThroughhisworkatCornell,AvCIR,andbeyond,he craft accidentinvestigators,hiscertificatesignedbyJerryLederer He wasagraduateofthe“first”formaltrainingcourseforair- him ontheroadtostudyofINJURIOUSMECHANISMS. which theoccupantswerekilledorseriouslyinjured.Thisput several crashesinwhichtheairplanewasrelativelyintact,but he wasassignedtoinvestigateplanecrashes.Heintriguedby months inthehospital.Afterthat,sincehecouldn’tflyanymore, was killed.Heseriouslyinjuredinthecrashandspent6 he wasinvolvedinamid-aircollisionwhicheveryonebuthim 1916, whiletrainingtobeaRoyalCanadianFlyingCorpspilot, folks inaviationsafety,hestartedwithaplanecrash—hisown.In DeHaven isliterallythe“father”ofcrashworthiness.Likesomany VESTIGATION, twopeoplestandout,fortheystarteditall.Hugh their verylivestothosefolks.InthefieldofcrashworthinessIN- ard Chandler,andmanyothers.Manysurvivorsofcrashesowe Harry Robertson,DocTurnbow,VicRothe,ChuckMiller,Rich- I couldspendhourstellingyouaboutfolkslikeDr.JerrySnyder, “father” ofcrashworthiness. Hugh DeHaven,literallythe They are INTERIOR PACKAGING. packaging engineerscall outer structurebywhat and immobilizedinsidethe package shouldbeheld 3. “Articlescontainedinthe container.” tect objectsinsidetheinner pact forcesandtherebypro- cushion anddistributeim- the outercontainersoasto sorbing energyappliedto force byyieldingandab- terials; theyshouldresist made offrailorbrittlema- container mustnotbe tures whichshieldtheinner 2. “Thepackagingstruc- ISASI 2003 Proceedings • 77 ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS tion, perhapsas afriend,perhapsasmento r. him, perhapsasaspurpushing theneedforsurvivalinvestiga- sure manyofthefolkswho are membersofISASIremember have knownandworkedwith Howardduringhislastyears.I’m an activepilotvirtuallytillthe endofhislife.Iwasprivilegedto investigator’s school”through AvCIRinPhoenix.Heremained technical adviceduringcreation ofthefirst“crashsurvival crashes. ThoughhewasalreadyattheFAA,Howardprovided reports, mostlyfocusedonhowtoimprovesurvivalin plane medical Branch.Duringhiscareer,hewrotemanyarticlesand continued throughhistenureatAvCIRandlatertheFAA Aero- Hasbrook becametheDirector. DeHaven “retired”fromtheAviationCIRprogram,andHoward east AirlinesConvair240accidentatLaGuardia.In 1954, vestigation onanairlineplanecrashinAugust1952ofaNorth- of survivors.Hasbrookconductedthefirst“crashworthiness” in- injury patternsdeterminedbyautopsyormedicalexamination ance tocorrelatethephysicalevidencegatheredatscenewith are oftenlackinginaccidentreportstoday.Italsoprovidedguid- be gathered:angles,velocities,anddistances—measurementsthat then. ItidentifiedthethreebasicpiecesofevidencethatMUST worthiness investigator.Mostofitisaspertinenttodaywas ity aspectsandidentifiedthetypesofdatanecessarytocrash- fied procedurestouseatanaccidentscenepreservesurvivabil- do acrashworthinessevaluationofanaircraftaccident.Itspeci- gators touseingatheringthetypesofinformationnecessary lished thefirstguide(writtenbyHasbrook)foraccidentinvesti- fatality. In1951,theCornellUniversityMedicalCollegepub- and constructionofaircrafttoreducethelikelihoodinjuryor gation into“survivabilityfactors”wouldallowchangesindesign Injury Researchcenter,Hasbrookbecameconvincedthatinvesti- dusting in1946.AsAdministratorofFieldResearchattheCrash DeHaven, Hasbrooksurvivedanear-fatalplanecrashwhilecrop- and canunderstandhowtodesignsystemsminimizeinjury. then wehaveanexcellentperspectiveonimpactcrashworthiness, “interior packaging”istherestraintsystemandtiedownchain, parts ofthecontainedobjects.” must transmittheforcesappliedtocontainerstrongest 4. “Themeansforholdinganobjectinsideashippingcontainer from impactagainsttheinsideofpackageitself.” overall design,foritpreventsmovementandresultantdamage This interiorpackagingisanextremelyimportantpartofthe 78 AG-1 concept. Hasbrook’s lifelongdevotiontocrashsurvivalinvestigation DeHaven’s assistantatCornellwasA.HowardHasbrook.Like If weassumethatthe“container”isanaircraftfuselage,and • ISA SI 2003 P r oceedings to theFAAin 1959 andbeganinvestigating crashestounder- (CARI), whichlaterbecameCAMI. TheybroughtHasbrookover John SwearingenrantheCivil AeromedicalResearchInstitute application. Doc continuestodazzlestudents withkinematicevaluationand International CenterforSafety EducationrunbySIMULA),and John Carrollstartedthefirstcrashinvestigator’sSchool(now the Desjardins, and,ofcourse,HowardHasbrook.DocTurnbowand Doc Turnbow,HarryRobertson,JohnCarroll,JoeHaley,Stan the programoveryears,includingJerryLederer,C.O.Miller , fire. Manyofthepioneersaviationsafetywereaffiliated with related issues,seatandrestraintsystems,fuelcontainment, and crashes. Experimentsgathereddataonkinematicandimpact- completed, includingthefamousDC-7andL1649Connie joint effortwithNASA,theU.S.Army,andFAA,43tests were aircraft togatherdataonwhatactuallyhappensinacrash. Ina program wasnowtodevelopandcarryouttestcrashesusingreal Crash InjuryResearch)inPhoenix,Ariz.Themainfocusofthe tion tookoveradministrationofwhatwasnowAvCIR(Aviation ated withtheFlightSafetyFoundationandFounda- to building“standard”airplanes. Beechcraft abandoneditscrashworthinesseffortsandwentback trapped inaburningairplanebytheshoulderharnesses.By1960, Aircraft shouldbemadetofly,notcrash.Theymighteven owners didn’twanttopayfortheextrasystemsintheiraircraft. of theirtime.Crashworthinesswasamarketingflop.GAaircraft ecutive, andAgriculturalAircraft.”ThefolksatBeechwereahead in 1953,“DevelopmentofCrash-SurvivalDesignPersonal,Ex- CIR. ThedetailsarespelledoutinareportauthoredbyDeHaven paign to“sell”safety,basedonDeHaven’sandHasbrook’sworkat like momandapplepie.Beechcraftbeganapublicrelationscam- after 1978).Toacrashworthinessinvestigator,theseideasseem quired intherestofGAfleetuntilthoseaircraftmanufactured even incorporatedshoulderharnessesinsomemodels(notre- and yokedesignedtoreduceheadupperbodytrauma.They with beltsattachedtotheseats,andabreak-awayinstrumentpanel ergy inacrash,seatsthatwereattachedwithboltstothespartrusses reinforced keelandcockpitarea,awingdesignedtoattenuateen- vintage Bonanzasandtwin-Bonanzashadalongnosesection, turer touse“crashworthiness”indesignofitsaircraft.The1950s than thoseinvolvedinaccidentswithothertypesofaircraft. these crashesexperienceamuchlowerinjuryandfatalityrate who haveinvestigatedAGaccidentswouldagreethatpilotsin nology (refinedabit)toprotectthepilotincrash.Mostofyou Pawnee. Crop-dusterstodaystillincorporatemuchofthistech- island ofsafety”forthepilot.TheAG-1evolvedintoPiper matically usinginertiareels.Itwasdesignedtoprovidea“40-g tied himintotheseatwithafive-pointharnessthatlockedauto- the cockpit.Ithadatri-axialsteelrollcagearoundpilotand thereby absorbingenergyandattenuatingtheg-loadsthatreached structure beingdesignedtoprogressivelydeforminacrash, above andbehindmuchofthemassinairplane,withthat and Hasbrook.Theconceptinvolvedpositioningthecockpit a crop-dusterincorporatingrecommendationsmadebyDeHaven In 1950,FredWeickdesignedtheAG-1atTexasA&M.Thiswas Applying thelessonslearned On thefederalsideofhouse, Dr.StanMohlerandr. In 1959,theCornell-GuggenheimFoundationbecameaffili- In thegeneralaviationworld,Beechcraftwasfirstmanufac- British AirTours737accident. with acigarette. tric inorigin.Tothisday,somefolksstillbelievethatitstarted someone throwingalitcigaretteintothetrashbin,butwaselec- for thenextfewyears.Ironically,firehadnotoriginatedfrom up. Thefocusremainedonthevariousplasticsinairlinecabins significantly theycontributedtothesmokeandtoxicgasbuild- issue wastherapidinvolvementofcabinfurnishingsandhow that smokedetectorsbeinstalledinairlinelavatories.Another detected. Asadirectresultofthisaccident,theFAAmandated the earlystagesoffireinlavatoryduringwhichitwasnot people onboarddied.Amongthesurvivalissuesexaminedwere building upthroughtheentiredescent.Twenty-threeof46 at Cincinnatiafter18minutes,withthesmokeandtoxicgasses an inflightfirein1983.Theaircraftmadeemergencylanding ability ofmodernairlineaircraft.Threeexamplesarecaseinpoint: learned fromtheseinvestigationsandhasimprovedthesurviv- airline crashes.TheFAAhasbeengraduallyapplyingthelessons ten verygoodatinvestigatingcrashworthinessissuesinmajor essential partofany“major”investigation.TheNTSBhasgot- ticularly overthelast25yearsorso,“survivalinvestigation”isan crashworthiness perspectivein1952,we’vecomealongway.Par- Since Hasbrookfirstinvestigatedanairlineaccidentfromthe Airline crashworthinessinvestigation on thelessonstobelearnedfrominvestigationofactualcrashes. not aboutthemanyfinetestprogramsthough.Ourfocushereis research impactandfiresurvivabilityissues.Thisdiscussionis years. TheFAATechnicalCenterinAtlanticCitycontinuesto eral facilitieshaveprovidedcriticalcrashworthinessdataoverthe and guidancerelatingtosurvivabilitycrashworthiness.Sev- years theprogramsandresearchhavecontinuedtoprovidedata changed totheCivilAeroMedicalInstitute(CAMI),andover source ofsurvivabilityinformation.In1966,thenamewas and theProtectionSurvivalLabatOklahomaCity. joined CARIandbecamechiefofthePhysicalAnthropologyLab stand howpeoplewereinjuredorkilled.In1960,Dr.JerrySnyder The firstinvolvesAirCanadaFlight797,whichexperienced Dr. Snyder’smanyarticlesandreportsareanotherexcellent A majorchangeherewastherequirementthatslidesbecovered by theropeordropping8feetsoaftertheywereoutside. had toexitthroughthecopilot’ssidewindow.Somewereinjured take, theslideburst.Severalpassengersandaflightattendant never touchedtheforwardslide,butduetoradiantheatup- were paintedorange-yellowforvisibilityifusedasliferafts.Fire fire. Thetworightforwardexitsdeployedproperly,buttheslides fuselage. Mostoftheemergencyexitswereunusabledueto concentrated mostlyontheleftsideofaircraftandcenter pooling fuelrantowardthefuselageandignited.Thefirewas ing severalholesinthatfueltank.Whentheaircraftstopped, they threwfragmentsintotheundersideofleftwing,open- the runwayandleftmaingearcollapsed.Whentiresfailed, and thecrewrejectedtakeoff.Theaircraftoverranendof 1978. TheaircraftblewtiresontheleftmainlandinggearatV1 Snozzle deviceinusetoday. troduce foamsintoaburningairlinercabin;culminatinginthe figurations. TheCFRcommunityexperimentedwithwaystoin- to experimentwithsmokehoodsandchangesinseatingcon- water spraysystems.Dr.HelenMuirandothersinBritainbegan began tofocusonfiresuppressionaswell,includinguseofcabin major changestomaterialsusedinairlinecabins.Experiments Canada accidentandeffortsoftheFAATechCenter,resultedin caught quiteabitofattentionand,incombinationwiththeAir efforts andactionsofthecrewweremagnificent.Thisaccident tion, 55peoplelosttheirlives,eventhoughthecrashfirerescue toxic gasses.Combinedwithmajorcompromisestotheevacua- involved veryquicklyandgeneratedatremendousamountof sidewall skininabout20seconds.Thecabinfurnishingsbecame does, resultinginaburnthroughandpenetrationoftheaircraft cumstances, thefireburnedmuchhotterthanJetAnormally resulting inamassivefuelleak.Duetosomewhatunusualcir- num inspectionplateandshattereda6-inchholeinthewing, shroud andintotheundersideofwing.Ithitacastalumi- left engineofthe737,throughasectioncan 1985. Duringthetakeoffrun,acombustorcanexplodedon The lastexampleistheContinentalDC-10accidentatLAXin The secondaccidenthappenedinManchester,England, in relativelyfewcrashes. cupants duringacrash,yetfire isinvolved by-products posethegreatest threattooc- pact fire.Intheairlineworld, fireandits duce thelikelihoodofimpact orpost-im- of fuelreleaseinacrashandtherefore re- derstand howtominimizethepossibility research projectisunderwaytobetterun- of fuelspillinaracecarcrash.Acurrent used inIndiecarstoreducethelikelihood craft aroundtheworldtoday.Theyareeven fire survivabilityareinuseonmilitaryair- in theworld.Hisdesignsforimproving haps thepremierefiresafetyinvestigator crashworthiness. HarryRobertsonisper- an evacuation. slide toremainfunctionallongerduring radiant heatsusceptibilityandallowthe with areflectiveouterlayer,toreducethe Over theyears, manyimprovements These examplesfocusmainlyonfire ISASI 2003 Proceedings • 79 ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS rienced, ascompared tothepassenger.Further,brittle na- tion totheleftanda50percent increaseintheg-loadsheexpe- through theleadingedge.This causedabouta15-deglateraliza- inches. Italsoprojectedupward andbenthisinboardseatrailup reduction incollapsiblespace (stoppingdistance)ofabout6 ward andpositionedunderthe pilot’sseatarea.Thiscauseda differing inseverity.Duringimpact,thenosegear failedrear- but atalesserseverityandwithoutdisplacement. passenger sufferedfacturesofthethoracicandlumbarvertebrae, tebrae andispermanentlyparalyzedfromthewaistdown. The pilot experiencedmultipledisplacedfracturesofhislumbar ver- sign, andflippedtotheleft,withlandinggearallfailed. The and hittherunway.Itslidoutininfieldarea,ataxiway near themidpointofrunway.Theaircraftthennosedover flaps appeareddown(40deg)attheinitialcontact,thenretracted about awingspanabovethesurface.Witnessesindicatedthat the The aircraftbouncedoffandflewhorizontallydowntherunway base tofinalturnandtouchingdownatahighverticalvelocity. first touch-and-goattempt,thepilotwasrushed,rounding flown in63daysandhadproblemswithlandings.Onhis friend (passenger)whataviationwasallabout.Thepilothadnot private pilot.Theflightwasintendedtoshowthepilot’sgirl- was aCessna172N,builtin1977,beingflownby120-hour points, aswellexaminetheissueswithcrashworthiness. to preventinjuriesinthefuture.Twocaseexamplesillustrate impossible, todeterminewhathurtspeople,normakechanges have causedtheinjuries.Withoutgooddata,itisdifficult,ifnot to nodatarelativethespecificinjuriesthemselvesorwhatmay serious injuriesand/orfatalitiesoccur.Inmostcases,thereislittle ences about2,000accidentsperyear.In25percentofthese, tion) operations.”Lookingata10-yearaverage,GAstillexperi- erations thanareavailableforPart135and91(generalavia- Therefore, moresurvivabilitydataareavailableforPart121op- ducted inconnectionwithaccidentsinvolvingPart121carriers. the majorityofBoard’ssurvivalfactorsinvestigationsarecon- only aircarrieroperationsperformedunder…Part121because tion crashsurvivability.Itstates,“TheSafetyBoardexamined out oneoftherecurringproblemswithanalysisgeneralavia- Air CarrierOperations,1983Through2000,”theNTSBpoints In thereport,“SurvivabilityofAccidentsInvolvingPart121U.S. General aviationcrashworthiness future. stand muchhighg-loadsinacrashthantheseats.Afocusfor and others,weknowthataproperlyrestrainedhumancanwith- weak pointandmoveswhereitfails.Plus,thankstoCol.Stapp Installing a16-gseatin9-gtrackandfloorjustchangesthe have awaystogo.Oneproblemremainsinvolvingoldaircraft. and betesteddynamically.We’vegottenbetter,thoughwestill floor tracks,changesweremadetorequire16g’shorizontally, were injuredorkilledbecauseseatsfailedpulledoutofthe Over theyears,aftermanyinvestigationsinwhichpassengers laterally. Theseseatswereonlyrequiredtobetestedstatically. passenger seatwithstand9g’shorizontally,4.5vertically,and1.5 for certificationweregrosslyinadequate,requiringonlythata systems areamajorfocusinairlinecrashes.Theoldstandards have beenmadeinimpactsurvivabilityaswell.Theseat/restraint 80 Examination ofthewreckagerevealedwhyinjurieswere The firstoftheseaccidentshappenedin1989.aircraft • ISA SI 2003 P r oceedings ing inashallow pull-offangle,whichmayactually bepositioned femurs. Ashortpersonmoves theseatallwayforward,result- pull-off angle.Inacrash,this candislocatethehipsorbreak moves theseatallwayback, resultinginaverticallapbelt (5’9” tallwithaveragearmleg torsoproportions).Atallperson other thantoaccommodatea 50thpercentilemalemesomorph change tothebeltpull-offangle iftheseatpositionedanywhere were madetothefloor,ratherthanseat.Thisresults ina that increasedinjuryseverity.Further,theseatbeltattachments were madeofrigid,brittlematerialsandtendedtofailin ways The olderGAseatshadlittletonocrashworthinessbuiltin. They Seat/restraint issues new aircraftweremuchless. than theolderaircraft.Yetinjuriestooccupantsof more severeimpact,withamuchcompleximpactsituation the previouscaseexample,newS-modelexperienceda much occupants survived.Contrastingthedamagetoaircraftwith normally. Eventhoughthiswasasevereimpactsequence,both of thewreckageshowedthatrestraintsystemsperformed structure. Theinstructorhadminorinjuriesonly.Examination wing rootintrudingintothecabinareaandhisheadstriking tal. Thestudenthadsufferedseriousinjuriesasaresultoftheleft the unconsciousstudentandbothweremedevacedtohospi- head frame.Theaircraftcametorestinverted.IPextracted at theroot,andfuselagebrokeapartaftcabinbulk- aircraft cartwheeledtotheleftoffrunway.Thewingfailed left. Theinstructor’sattempttocorrectwasunsuccessfulandthe wing cameupsuddenlyandviolently,rollingthemovertothe encountered aseverelateralwindgustormicroburst.Theright They wereattemptingatouch-and-golandingwhentheaircraft quirements specifying22-gseatsanddynamictesting. 172S, whichwasdesignedaccordingtothemodifiedPart23re- crashworthiness investigationcanbe.ItinvolvedanewCessna most olderGAaircraft. vestigation alsorevealedtheinherentlackofcrashworthiness derstand WHYtheinjurypatternswerewhattheywere.Thein- tances, angles,andvelocities)itwouldhavebeendifficulttoun- experienced bythetwooccupants.Withoutbasicdata(dis- both occupantsexperienced.Theresultwastheinjurypatterns ture oftheseatframeundoubtedlyincreasedg-spikethat Left, S-tubeseat.Right,pre-1988generalaviation The aircrafthadtwooccupants,aninstructorandastudent. The secondaccidentisagoodillustrationofhowimportant stand WHYthepeoplegothurtorkilled. the autopsyreportas“airplanecrash”justdon’tallowustounder- late tothevictims.Theolddaysoffindingcausedeathlistedon logic informationastoinjuriesandinjurymechanismstheyre- the surroundingterrain.Wealsoneedgoodmedicalandpatho- formations totheaircraft(photosworkwonders)astheyrelate and belowtheoccupants.Wealsoneedgooddescriptionsofde- ments, aswellverticalandhorizontalcrushdistancesinfrontof need distances.Theverticalandhorizontalgroundscarmeasure- and hownon-lineardecelerationsmighthavehappened.Wealso the surface,andagooddescriptionofentireimpactsequence at impact,etc.,velocities,impactairspeed,decelerationonceon fects. Forimpactanalysis,weneedangles…impactangles,attitude points, fireoriginandpropagation,ignitionsources,ef- tion. Forfirecrashworthiness,weneedevidenceoffuelrelease gator istherightdatanecessarytoDOacrashworthinessevalua- survivability. can beappliedtochangeaircraftdesignsandsystemsimprove plying thelessonslearnedthroughcrashworthinessinvestigation crash, withminimalinjury.Itisonefurtherexampleofhowap- America afewyearsago.Alltheoccupantssurvivedanextreme This particularseatwasinstalledinaC206thatcrashedSouth incorporated intheMissionAviationfellowshipaircraftforyears. accomplished intheCessna182andC206models,hasbeen 70s foruseinsomegeneralaviationaircraft.Theretro-fitcanbe military aircraft.TheS-tubeseatwasoriginallydesignedinthe factures severalenergy-attenuatingseatsforuseinavarietyof to attenuateenergyinacrash.TheSIMULACorporationmanu- ness investigation. ture, aswellotherchanges,alllearnedthroughcrashworthi- seat frameitself.Thenewer22-gGAseatsincorporatethisfea- 45 deg—achievableonlyifthebeltanchorageisattachedto spleen orintestines.Theidealpull-offangleacrossthepelvisis over thelowerabdomen.Inacrash,thispersonruptures What thecrashworthinessengineerneedsfromfieldinvesti- There areseveralseatsystemsthatintentionallydesigned 15. Swearingen,John,“GeneralAviationStructuresDirectlyResponsiblefor 14. Snyder,Richard, 13. Snyder,Richard,“CrashworthinessInvestigationofGeneralAviation 12. NTSB,“SurvivabilityofAccidentsInvolvingPart121U.S.AirCarrier 11. NTSB,“SAFETYREPORT-TheStatusofGeneralAviationAircraftCrash- 10. Miller,C.O.,“AviationCrashInjuryPrevention:Perspective1981,” 9. Hasbrook,H., 8. Hasbrook,H.,“CrashInjuryStudy:NortheastAirlinesCV240Accidentat 7. Hasbrook,H.,“TheHistoricaldevelopmentoftheCrash-ImpactEngi- 6. DeHaven,H.,“DevelopmentofCrashSurvivalDesigninPersonal,Ex- 5. DeHaven,H.,“AccidentSurvival—AirplaneandPassengerAutomobile”, 4. DeHaven,H.,“TheRelationshipofInjuriestoStructureinSurvivable 3. DeHaven,H.,“MeasuresforIncreasingSafetyofFlyingPersonnelin 2. Chandler,Richard,“Seats,Restraints,andCrashInjuryProtection,” 1. AircraftCrashSurvivalDesignGuide,Vol.1–5 References for thedurationofthisprogram. the past!Youcannowflipyoursafetyswitchbackto“prevention” changes necessary.WeCANchangethefuture,bylearningfrom investigation canprovidethebestdataavailabletomakethose cine, FederalAviationAdministration,Jan.1971. Trauma inCrashDecelerations,”SpecialReport,OfficeofAviationMedi- & PilotsAssn.,May1981. Accidents,” SAEPub.750537,SocietyofAutomotiveEngineers,1975. Operations, 1983-2000,”NTSBSR-01-01,2001. worthiness,” NTSBSR-80-2,1980. Journal Injury Research Crash InjuryResearch,CornellUniversityMedicalCollege,August1952. La GuardiaAirportonJan.14,1952,”InformativeAccidentRelease#14, neering PointofView,” sity MedicalCollege,May1953. ecutive, andAgriculturalAircraft,”CrashInjuryResearch,CornellUniver- Jan. 1950,aspartoftheSAESymposiumon“PackagingPassenger.” presented attheannualmeetingofSocietyAutomotiveEngineers, Medicine, Report#440,July9,1945. Aircraft Accidents,,NationalResearchCouncil,CommitteeonAviation November 14,1941. tional ResearchCouncil,CommitteeonAviationMedicine,Report#34, Crashes: AMemorandumtotheCommitteeonAviationMedicine,”Na- ference onCabinSafety 1989. Command, USAAVSCOMTR89-D22A-E,SIMULA,Inc.Phoenix,AZ, If wearetrulygoingtoimprovesurvivability,crashworthiness , Summer1981. , CIR-H-2,CornellUniversityMedicalCollege,Jan.1951. Handbook forAircraftAccidentInvestigatorsCooperatinginCrash General AviationAircraftCrashworthiness , FlightSafetyFoundation,Dec.1984. Clinical Orthopaedics N , 8:268-274,1956. , U.S.ArmyAviationSystems ISASI 2003 , AircraftOwners Proceedings LPBA • Con- 81 ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS T tory, atleastfor pilots. tem toxicologyisalsocollected foranalysisbytheFAAlabora- manner ofdeathsuchassuicide, homicide,ornatural.Postmor- an accidentunlesscircumstances andfindingssuggestanother stab wounds,orexplosives.The mannerofdeathiscertifiedas injuries andtheabsenceofsuspicious featuressuchasbullets, For thispurpose,theexamination needonlydocumentlethal cause ofdeathisusuallybluntforceinjury,firey,orboth. certify thecauseandmannerofdeath.Inaircraftmishaps, the examiners orcoronersistoscientificallyidentifytheremains and tem examinationforeachoccupant. local regulationsandpoliciesdeterminetheextentofpostmor- can orderpostmortemexaminationsforcivilianpilots,state or gists, employedbyelectedcoroners.Whilefederalinvestigators tions areperformedbycivilianmedicalexaminers,orpatholo- investigation board.Incivilianmishaps,postmortemexamina- trained militarypathologistswhodirectlysupportthemishap tary aircraftmishapfatalitieshavebeenperformedbyspecially diction. U.S.military—pilots,crew,andpassengers. vilian—pilots byfederallaw,crewandpassengerslocaljuris- aircraft mishap. tem examinationofsome,orall,occupantsfatallyinjuredinan United StatesofAmerica,federalregulationsmandatepostmor- result inanoccupant. order ofmagnitudelessthanitwouldtaketoproducethesame than anyaircraft:aircraftstructurewillfragmentunderforcesan investigators. Infact,thehumanbodyisstructurallymorerobust cupant injuriescanprovideimportantscientificdataformishap General. aviation pathologyandforensictotheUSAFSurgeon hundreds ofaircraftaccidentinvestigationsandservedasaconsultantin Pathology, heparticipatedasanaviationpathologyconsultantinmany pathology. Duringalmost20yearsattheArmedForcesInstituteof Ohio atToledo.Heiscertifiedinanatomic,clinical,andforensic sity ofCalifornia 82 The legalpurposeofpostmortemexaminationbymedical For morethantwodecades,postmortemexaminationsofmili- Postmortem Examinationofaircraftmishapfatalities:U.S.ci- When aircraftmishapinvestigationsareconductedinthe • ISA SI 2003 ing analysisofthewreckage,medicalexaminationoc- damaging eventsastheaircraft.Analogoustoengineer- he victimsofafatalaviationmishapexperiencethesame Through InjuryPatternAnalysis Enhanced OccupantProtection P at r chemical andbiomedicalengineeringfromtheUniver and Medicine.Dr.GormleyreceivedhisPh.D.in monwealth University;VirginiaInstituteofScience Medicine, MedicalCollegeofVirginia,VirginiaCom iner, CommonwealthofVirginia;DepartmentLegal MC CFS(retired);OfficeoftheChiefMedicalExam oceedings William T.Gormley DavisandhisM.D.fromtheMedicalCollegeof , M.D.,Ph.D.;colonel,USAF By WilliamT.Gormley - - - the pilotwas65 yearsoldandthepassenger was45yearsold,it and therewassevereatherosclerosis inthecoronaryarteries.Since of mostinternalorgans.Aheart wasrecoveredfromthecrashsite died instantly,andtheirbodies werefragmentedwithevisceration trees nearanairportonadark andfoggynight.Bothoccupants sion, andtheheartdiseasewas notafactorinthismishap. There wasnoevidenceofpilotincapacitationpriortothe colli- to thecommercialairportandinadvertentlyflewintoairliner. road map,unawarethathehadwanderedintoanapproach path indicated thatthepilotwasmildlylost,navigatingvisuallyusing a the privateaircraft,crashscenedocumentation,andradar data interviews, wreckageanalysis,autopsydatafromalloccupants of many peoplelivingquitewellwithsimilardisease.Investigative attack. Whileatherosclerosiscancauseaheartattack,there are most 100deaths,werecausedbytheprivatepilothavinga heart teries andlocalauthoritiesannouncedthatthemishap, al- vate pilotdemonstratedsevereatherosclerosisinthecoronaryar- aircraft crashed,andtherewerenosurvivors.Autopsyofthepri- aircraft intoacommercialairlinerapproachinganairport.Both specific anatomicdiseasetodeterminetheroleinamishap. aircraft mishap.Theflighthistorymustbecorrelatedwiththe pilots. Thisdiseasemayornothaveasignificantroleinthe commonly identifiednaturaldiseaseinmiddleagedorolder pacitation. Atheroscleroticcardiovasculardiseaseisbyfarthemost to bepossiblemishapfactorsbecausetheycancauserapidinca- Natural diseasesoftheheart,brainandlungaremostlikely ing traumaticinjuries,willidentifypre-existingnaturaldisease. A completepostmortemexamination,inadditiontodocument- Document naturaldiseaseandevaluateroleinmishap cerns ofaccidentpreventionandoccupantsafety. accurate, effective,andcompletemedicalinputformanycon- amination results,and6)otherpertinentdata.Thiswillpromote wreckage analysis,4)occupantmedicaldata,5)postmortemex- team sharingofdataon1)flighthistory,2)crashsitedata,3) disciplinary effort,withthepathologistandmishapinvestigation postmortem medicalexaminationsshouldfunctionasaninter- cupants aretrulydead.Formaximuminvestigationalimpact, ministrative exercisetoproducedocumentsprovingthattheoc- ment thatcertifiesthecauseandmannerofdeath. the postmortemexaminationwillproducelittlemorethanadocu- ally getlittleornoresponsefrominvestigators. stances ofdeathverballyandinwriting.Unfortunately,weusu- investigation, weroutinelyrequestinformationaboutthecircum- topsies onallfatalitiesinaircraftmishaps.Aspartofourdeath In anothercase,asmallaircraft withtwooccupantscrashedinto In onemishap,amiddle-agedpilotflewsingle-engineprivate The postmortemexaminationshouldbemorethananad- With limitedinformationaboutthecircumstancesofdeath, In theCommonwealthofVirginia,weperformcompleteau- fractures ofthe metacarpals,andfracturesof thewristsandlower palmar lacerations,fracture-dislocation ofthethumbbase,serial characteristic injuries.Classic injuriesofthehandsmayinclude may bemechanicallytransmitted tothehandsandfeet,causing sticks, throttles,rudderpedals, anti-torquepedals)crashforces pilot orcopilotareinfirmcontact withaircraftcontrols(yokes, aircraft atthetimeofcrash. Whenthehandsandfeetofa nation ofpilotsandcopilotsinvolveswhowascontrolling the injury andwasnotdyinginthehelicoptercrash. the softtissues.Thus,patientwasaliveattimeof neck vessels andpressuretopropelbloodthroughthetorn into intramuscular hemorrhage,etc.)requiresbothdamageto blood loss ofbloodpressureandcirculation.Toformabruise(contusion, crash. Why?Withmultiplebluntforceinjuries,thereisvery rapid entific datathattheneckfractureoccurredbeforehelicopter of bleedingintheneckmusclesaroundfractureprovided sci- accident thatmayhaveinvolvedneckinjury.Thedocumentation this correlation. crashing helicopter.Engineeringanalysisofwreckageverified for thepilot’srestraintsystem,andseparatingpilotfrom have passedthroughthecockpit,damagingtie-down-chain sistent witharotorstrike.Thisinjurysuggeststhattheblademay the rightarmwascausedbyasharp,chop-likeinjurymostcon- rated fromthehelicopterpriortoimpact.Theamputationof fire incineratedmostofthewreckage. and burnedabout50yardsbeyondthetree-strike.Thepost-crash the helicopterhitatreewithmainrotorblade,thencrashed some informationaboutthemishap.Thecrashsceneshowsthat manner ofdeathwascertifiedasaccidental. All deathswerecertifiedasduetobluntforceinjuries,andthe ture ofhisneck,withhemorrhageinthedeepcervicalmuscles. moglobin wasnotelevatedinanyvictim.Thepatienthadafrac- There wasnosootintheairwaysofanyvictim,andcarboxyhe- fractures andlacerationsoflungs,liver,heart,aorta,brain. and multiplelethalbluntforceinjuries,includingskullrib tation oftherightarm.Theotherbodieshadextensiveburns burned buthadmultiplelethalinjuriestothetrunk,andampu- died andtheremainsweresentforautopsy.Thepilotwasnot occupants, includingpilot,copilot,twomedicsandapatient.All trative examples. nate, proposedmishapsequencesasshowninthefollowingillus- analysis. Suchtechniquesoftenhelpidentifyandvalidate,orelimi- help reconstructthemishapsequencethroughinjurypattern Correlation ofmedicalwithotheraccidentinvestigationdatacan Mishap sequencereconstruction eased heartwasfromthepassenger,notpilot. with scientificcertaintyDNAanalysisdemonstratedthatthedis- unusual. Therewasnowaytodeterminethetruemishapfactors events occurredatthesametime,suchacoincidencewouldbe nate airportisalsonotuncommon.Whileitpossiblethatboth sis, crashingwhilemakinganapproachinsteadoffindingalter- tacks arenotuncommonin65-year-oldpeoplewithatherosclero- was assumedthattheheartofpilot.Whileat- Another questionthatmaybeansweredwithdatafromexami- The patientwasbeingtransportedtoahospitalfollowingan Since therewerenothermalinjuries,thepilotmusthavesepa- What morecanwedo?Wecorrelatethemedicaldatawith A medicalevacuationhelicoptercrashedandburnedwithfive hap investigation: incorporation ofsignificantmedical dataintoanyaircraftmis- The followingstepswilloptimize collection,documentation,and Optimal medicalconsultation lethal protectiveequipmentinjuries. sequences. Suchobservations areextremelyimportanttoavoid ated withrotationofhelmetsaroundneckstrapsduring crash gans. Theyhavealsodemonstratedlethalneckinjuriesassoci- points ofrestraintandsubsequentlethaldamagetointernal or- documented submariningfailureofsystemswithlessthan five ment theimpactofrestraintsystems.Patterninjurieshave also use andfunction.Bruisesonthesurfaceofbodymay docu- systems. Examplesincludepatterninjuriestodeterminerestraint in evaluatingthesafeguardsprovidedbyoccupantprotection Postmortem injurypatterncorrelationisanimportantelement restraints, energyabsorption Evaluation ofsafetysystems— in otherwisesurvivablehelicoptercrashesisanobviousexample. cess ofcrashworthyfuelsystemdesigninpreventingfiredeaths to preventorremediatethroughengineeringdesign.Thesuc- Usually, post-crashfireanddrowningarethehazards then mishapsurvivaldependsonthepost-crashenvironment. also themostvaluableforfuturesurvivabilitydesign. engineers, andothermishapinvestigators.Thesecorrelationsare tions naturallyrequiresignificantconsultationbetweenphysicians, can becorrelatedwithstructuresandmechanisms.Suchcorrela- tions occur,theoccupantmaysustainpatternedinjuries,which forces maybeexchangedwithlethalresults.Whensuchinterac- crash sequenceforaplacetobe,tremendousequalandopposite man bodymustcompetewithenvironmentalstructuresduringa vivability dependsonmaintenanceofoccupiablespace.Ifahu- Body Fragmentation Transection ofAorta Laceration ofAorta Compression ofSpine Injury stood. Generalmedicalcrashforceindicatorsareshownbelow: general magnitudeifthecrashsequenceanddynamicsareunder- injuries. Thesethreeindependentestimatesshouldbeofthesame scene data,aircraftpost-crashstructuralintegrity,andoccupant Crash forcescanbeestimatedbyengineeringphysicsbasedon must belessthan50gtoavoidlethalinjuries(lacerationofaorta). able post-crashenvironment. able crashforces,maintenanceofoccupiablespace,andasurviv- Throughout themishapsequence,survivaldependsupontoler- decrease deathsandinjuriesinthoseaircraftmishapsthatoccur. vivability analysiscanhelpprovideimproveddesigncriteriato Incorporation ofdatafrompostmortemexaminationintosur- Survivability analysis ankles, andlowerlegsmaybecharacteristic. arms. Onthefeet,plantarlacerationsandfracturesof If crashforcesaretolerableandoccupiablespaceinmaintained, If theoverallcrashforcesarewithinsurvivablelimits,thensur- Overall crashforcesasexperiencedbytheoccupantsingeneral 350+ Gxyz 100 Gxyz 50 Gxyz 20-25 Gz Forces ISASI 2003 No No Borderline Yes Survivability Proceedings • 83 ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS • 84 • mal butbetterthannothing.) (Contact pathologistaftertheautopsywithquestions—sub-opti- and questions. • tory, crashsite,andspecificconcerns. • and medicalconsultants. not havethetimeorinteresttofullyparticipateininvestiga- arranged. Localmedicalexaminersorcoroner’spathologistsmay and othersdependingonmishapspecifics. Human Factors,Investigator-in-Charge,Structures,Engineering, with othermishapinvestigationdata. Sometimes, suchactiveandinteractiveconsultationcannotbe Participants shouldincludepathologistandrepresentativesof Contact thepathologistbefor Ar Attend thepostmorteme Ar • ISA range abriefvisittothecrashsite(wr range meetingtor SI 2003 P r oceedings eview medicalfindingsandcor xam toshar e autopsyto e information,concerns, eckage) bypathologist xplain mishaphis- r elation dent investigationprocess. tive incorporationofinjurypatternanalysisintheaircraftacci- financial reimbursementforadditionalexpensesarekeytoeffec- investigation systems.Mutualrespect,friendlypersuasion,and may severelytaxthebudgetandresourcesofmanyciviliandeath pilots. • unclothed. • • • • tion pathologistsifnecessary: examination canprovideabasisforlaterconsultationwithavia- tion. Whenthisoccurs,thefollowingelementsfromautopsy Such documentation,especiallythephotographsandX-rays, T T Complete autopsyr Diagnosis listincludingallsignificantinjuriesanddisease. Photographs ofalle otal bodyX oxicology r eport. -rays withspecialattentiontohandsandfeetof eport. xternal bodysurfaces,clothedand N lies tocommence grieving. death, fulfillingthehumanitarian roleofhelpingsurvivingfami- tified remains,thedeathcertificate providestangibleevidenceof benefits, andspousalremarriage. Alongwiththereturnofiden- purposes ofsettlingestatematters, collectionoflifeinsurance tion (ID)effort.Certification ofdeathisessentialforthelegal surviving familiesofthevictims whoareservedbytheidentifica- Urgently needingtoknowthefateoftheirlovedones,it isthe Humanitarian mission medical examiner(ME),canprovidemutualsupport. air safetypersonnelandlocalauthorities,suchasthecoroner or ducted independently,theparallelinvestigationscarried out by effort, encompassingawiderangeofdisciplines.Although con- gation offatalaviationmishapshasevolvedintoamulti-agency ries totheoccupants,andcorrelationofthesedata.Theinvesti- counts, examinationofthedamagetoaircraftand inju- ton CemeteryonSeptember25th. one ofthewires.HewasburiedwithfullmilitaryhonorsatArling- sustained afatalskullfractureashestruckwoodensupportor the eyeareaandlip.Lieut.Selfridgediedthatevening,having tion. Wrightsurvived,withfracturedribs,femur,andinjuriesto their positionswerenoted,andrecoveredformedicalatten- seats atimpactbywiresbracescrossinginfrontoftheirbodies, The machinepitchedandfell75feettotheground.Heldintheir wire, breakingthebladeandcausingsubsequentlossofcontrol. ant ThomasSelfridgeashispassenger.Thepropellerstruckaguy for theU.S.ArmyEvaluationBoard,withBoardmemberLieuten- O in theNationalAircraftAccidentInvestigationSchool. City, Okla.,shelecturesonmedicalfactorsinaircraftaccidentinvestigation phy. AsassociatestaffattheTransportationSafetyInstituteinOklahoma and deathinvestigatorsinforensicdentalevidencephotogra- and EmergingTechnologies.Shehastrainedlawenforcementpersonnel Forces InstituteofPathology’sAnnualCourseinForensicIdentification course intheSchoolofDentistry.SheisafacultymemberatArmed prosthodontics laboratoryanddirectsagraduateheadneckanatomy College ofHealthSciences,aswellinstructsinthepreclinicalfixed Investigation oftheaccidentinvolvedcollectionwitnessac- By MaryCimrmancic,D.D.S.,AssociateStaff,TransportationSafetyInstitute,OklahomaCity,Okla., Wright wasatthecontrolsofFlyerbeingtested craft intheUnitedStatesoccurredatFt.Myer,Va.Orville n Sept.17,1908,thefirstfatalmishapofapoweredair- Protection: VictimIdentification Marquette University,sheteachesgrossanatomyinthe Brookfield, Wisc.AnAdjunctAssistantProfes she isinprivatepracticewithGreenbrookDen Medical CenterinMilwaukee,Wisc.Ageneraldentist, general dentistryattheZablockiVeteransAdministration Marquette Universityandcompletedaresidencyin Dr. MaryCimrmancic Forensic AspectsofOccupant Adjunct AssistantProfessor,MarquetteUniversityMilwaukee,Wisc. receivedherD.D.S.from - - sor at tistry, in • • The identificationproceduresarecarriedoutinthreephases team offorensicpersonneltorecoverandexaminetheremains. death, anddeterminingvictimidentity,theMEdirectsadiverse fulfill therequirementsforestablishingcauseandmannerof deaths, theMEconductsamedicolegaldeathinvestigation.To Responsible foraccountingfatalitiesandcertifyingtheir Death investigation identified, theflight crewcanbeexaminedfor pre-existingdisease individuals intothesequenceof eventssurroundingthecrash.Once the fatalitiesneedtobeidentified. Thiscanfacilitateplacementof medical evidencetobestcontribute totheoverallinvestigation, vestigation: preventionofdeath andinjury.Inordertoallowthe examination oftheoccupants canservethegoalsofsafetyin- to themishapinvestigationitself. Onabroaderscale,however, The postmortemexaminationprocessmayatfirstseemunrelated Safety investigation federal resourceswithlocalandstateauthorities. tation DisasterAssistancewiththeresponsibilityforcoordinating cilitated bylegislationthatassignstheNTSBOfficeofTranspor- sistance, andmortuarysupport.Accesstotheseresources isfa- thorities tosupplementemergencymedicalresponse,family as- States canbedeployedtofunctionunderlocaljurisdictional au- sponse capabilities,federallevelresponseteamsintheUnited tions basedonDNAprofilescantakeweeksormonths. fatalities, ordelayswithobtainingantemortemrecords.Identifica- culties withrecovery,extremefragmentation,sheernumbersof time frameforthisprocesscanbesignificantlyextendedbydiffi- analyses, theMEassignsidentitytoindividualremains.The Finally, withevidencecompiledfromthevariousexaminationsand individual featureswithreferencesmostlikelytoidentifythem. mortem andantemortemfiles,permittingexaminerstocompare mated databasesspeedthetediousprocessofsortingthroughpost- ing askilledrecordslibrariantomanagetheinformation.Auto- tem findings,anextensivecollectionofdataisproduced,requir- antemortem medical/dentalrecordsandthecollectionofpostmor- anthropologic examination,andembalming.Withthearrivalof ceed tofingerprinting,dentalexaminationandX-rays, collection oftoxicologyandDNAspecimens.Remainsalsopro- atically movedtostationsforstandardautopsyexaminationsand documented andremovedforsecurity,thebodiesaresystem- are assignedcasenumbers,personaleffectsandotheritems lishing identity. • providers. When amajoraviationmishapoverwhelmslocalandstatere- At themortuary,asetflowofproceduresiscarriedout:remains P Comparison ofpostmortemandantemortemdataforestab- Antemortem datacollectionfr ostmortem datacollectionfr om thesiteandr om familyandhealthcar ISASI 2003 emains. Proceedings • 85 e ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS the facialfeatures ofabodyresemblingthe generaldescription which body,theywerefound.Furthermore, whenaskedtoview carefully documented,alsonoting whereatthescene,andwith and transportedtothemortuary facility.There,theitemsare ever, theseitemsaretobekept withthebodyuntilitisrecovered near abodyatthesitemaynot necessarilybelongtothem.How- be misleading. However, thesemethodsneedtobeusedcautiously,astheycan evidence indicatingthepresenceofindividualatmishap. other personaleffectsareuseful,particularlywhenthereisstrong extent ofdamageincurred.Facialfeatures,clothing,jewelry,and in part,whatmaterialsarerecoveredfromthescene,and the clothing wererelativelyintact. firmed hisparticipationintheflight.Hisphysicalfeatures and his identity:peopleknewwhohewas,andwitnessaccounts con- was notinquestion.Strongcircumstantialevidencesupported In the1908WrightFlyercrash,identityofLieut.Selfridge Identification tation oftheflightcrewandresultantcrash. note anddischargedthefirearminaircraft,causingincapaci- collected toidentifythegunman,demonstratethathewrote destruction oftheaircraftandoccupants,sufficientevidencewas the writingwasthatofgunmanhimself.Despiteextreme ness bagwasalsofound.Handwritinganalysisconcludedthat tion oftheinstrumentpanel.Evenmoreremarkably,airsick- rifling patternthatmatchedabulletfoundimbeddedinsec- to fitthegunframe.Atestbulletfiredfrombarrelrevealeda airline. Thebarrelofthegunwasfoundseparately,and print recordedontheformeremployee’sfingerprintcardat trigger arearevealedafrictionridgepatternmatchingthatof a gunframewithsixexpendedrounds.Skinrecoveredfromthe tion withlawenforcementauthorities. and itsoccupants.Theinvestigationwascarriedoutincoopera- was widelyscattered,withseverefragmentationoftheaircraft pacted onagranitehillside,killingall43occupants.Thescene self. Theaircraftwentintoasteep,high-velocitydiveandim- flight deck,shootingthecaptainandfirstofficer,thenhim- fired thegunathim.Thegunmanthenproceededtoinvade sor onanairsicknessbag,passedittohim,andsubsequently supervisor. Inflight,hewroteathreateningnotetothesupervi- smuggled arevolveraboardcommuterflighttakenbyhisformer from hisposition,adisgruntledformerairlineticketagent flight andconfirmedhisactions:Onthedayhewasdismissed dence haddemonstratedthepresenceofsuspectaboard related withpreflighteventsandinflightrecordings,physicalevi- clude asuspectandcrimescene.Inthefollowingcase,whencor- the criminalactivity.Physicalevidenceisessentialtolinkorex- was present,whattheiractionswere,andwhoresponsiblefor Criminal investigatorssearchforevidencethatdemonstrateswho Criminal investigation ference withtheoperationofaircraft. Also ofinterestisevidencecriminalactivityandpassengerinter- tected, thesedatacanbecorrelatedwiththeeventsofmishap. and incapacitation,aswellevidenceofcontrolinjuries.Ifde- 86 Items wornorcarriedaboard byanindividual,orfoundon The methodsbywhichfatalitiesareidentifieddependupon, Although raincomplicatedthesearch,fieldpersonnellocated • ISA SI 2003 P r oceedings sue destruction duetofire,fragmentation,and decomposition. tality incidents,particularlywhere therehasbeensignificanttis- prints havecomprisedthemajority ofidentificationsinmassfa- tooth affixedwithsilverwire, replacing amissingcuspid. what waslikelytohavebeen asculptedpieceofhippopotamus fixed bridgethathehadplaced thepreviousyear,consistingof by hisfriendandpracticingdentist,PaulRevere.Herecognized a stripped ofhisuniform,theidentityremainswasconfirmed body wasrecoveredseveralmonthslaterbyfriends.Having been a shallowgravebyBritishsoldiersalongsideanothercasualty, his wound totheheadinBattleofBreed’sHill1776.Buried in ranks withhismen,MajorGeneralWarrensustainedafatal bullet Paul RevereandphysicianJosephWarren.Whilefighting in the history involvedtwopatriotsoftheRevolutionaryWar:silversmith One oftheearliestrecordedforensicidentificationsinAmerican Background whom theremainsdonotrepresent. provisional IDwhendifferencesintheevidencedemonstrates combined withotherevidencetoidentity.Anexclusionrefutesa an identificationbyonemeansoranothermaybesubsequently no unexplainabledifferences.Insufficientevidencetosupport essentially thesameasthoseinantemortemreference,with resent. Featuresexhibitedbytheremainsaredeterminedtobe mortem datathatdemonstratewhomtheremainsactuallyrep- veloped fromcomparisonofantemortemreferenceswithpost- provisional); insufficientevidence;exclusion.ApositiveIDisde- the followingterms:positive;possible/presumptive(tentative/ print databases,forexample,canprovidethesematerials. photographs, todentists,physicians,andhospitals,finger- proceed. Avarietyofsources,rangingfrompersonalitemsand provisional ID,thechallengingsearchforantemortemrecordscan from survivingfamilymembersbyaffairspersonnel,and vide moresubstantialevidence. eton, aswellDNAandartifactsofsurgery,disease,pro- unique anatomicalfeaturesoftheteeth,fingerprints,andskel- porting information,dependingupontheiruniqueness.The tattoos, piercings,andotherbodymodificationscanprovidesup- the shapeofears,moles,andscarscanbeuseful.Tosomeextent, distinctive. Forexample,thecontourandpositionofteeth, ence, providedthatthefeaturesbeingcomparedaresufficiently ence exemplars. compared withthosepresentinanalogousantemortemrefer- are collectedfromremains,andthedocumentedfeatures X-ray andphotographicimages,fingerprints,DNAprofiles mented insomeformduringtheindividual’slife.Postmortem upon uniquephysicalfeaturesofthebodythathavebeendocu- positively identifiedonthisbasis. sent. Whenthisevidenceissufficientlystrong,abodymaybe identification (ID),establishingwhomtheremainsmightrepre- visual features,canformthebasisforaprovisionalortentative they hadlost.Inspiteofthesedrawbacks,personaleffectsand erroneously concludethatthoseremainsareoftheperson of theirlovedone,adistraughtfamilymemberorfriendmay Used aloneandincombination, dentalfeaturesandfinger- The strengthofconfidenceforidentificationisreflectedby Directed byinformationfromtheflightmanifest,datagathered For example,aphotographportraitmayserveasrefer- Scientific methodscanconfirmorrefuteaprovisionalID,based tion. Advances intechnology,suchastheIntegrated Automated bases, obtainlatentprints,and makecomparisonsforidentifica- from decedents.Theysearch forrecordedprintsinvariousdata- ian roleinobtainingfingerprints, palmprints,andfootprints Analysis BranchoftheFBILaboratory haveservedahumanitar- cation. agentsandlatentfingerprint specialistsfromtheForensic requests oflocalauthoritiesto assistwithdisastervictimidentifi- and antemortempatterns. cations areaproductofcomparisonbetweenthepostmortem lifted fromitemsfrequentlyhandledbytheindividual.Identifi- sources. Intheabsenceofrecordedprints,latentprintscan be with theindividual’semployer,fingerprintdatabases,orother tails andcomparethemwithrecordedantemortemprints on file finger padsfromtheeffectsoffire. gers tocurlinward,formingfists.Toanextent,thisprotectsthe major musclegroupscausesthedecedentsarms,hands,andfin- gone fragmentationanddecomposition.Infire,contractionof recorded fromremainsthathavebeenexposedtofireorunder- tion ofidentificationsinaviationdisasters.Thesepatternscanbe tion ridgedetailsformingthebasisforwhatcanbeafairpropor- The fingers,palms,andsolesexhibitdistinctivepatternsoffric- Fingerprints Scientific methods being identifiedbymorethanonemethod. ability ofantemortemreferences.Thiscanresultinanindividual the remains identification. Inall,identificationsarebaseduponthenatureof protocols tofacilitaterecoveryofremains,makingpossibletheir of thesescenarioshavenecessitatedadvancesintechnologyand field, swamp,openocean,andurbanenvironments.Thedemands ing files maytakeweeksormonthstoprocess. weeks (sometimeshours)ofrecovery.IDsbaseduponDNApro- ing classicmethodscanbecompletedtypicallywithindaysor and burningoftheaircraftitsoccupants.Identificationsus- as massfatalityincidents(MFIs),involvingseverefragmentation involving explosion,fire,andhigh-velocityimpactshavepresented evidence thatisavailableforexamination.Airtransportdisasters accordingly; therefore,thenatureofmishapdetermines its causationandpresentation,theconditionofremainswillvary able referencematerials.Witheverymishapbeingdistinctivein the materialsrecoveredfromscene,andsecondarily,avail- these remainsrepresent. derived, greatlyincreasingthepotentialforestablishingwhom materials nowprovideasourcefromwhichDNAprofilescanbe to whereDNAanalysisofeveryfragmentisnotunheardof.These and recognizedbyamemorial,thestandardhasnowbeenraised past, thoseremainsmayhavebeenburiedinacommongrave those remainsnotidentifiablebyothermeans.Whereinyears ing classicmethods,DNAhassupplementedthem,extendingto emerged asaprimarymeansofidentification.Insteadreplac- diographic examinationsandanthropologicanalyses. Additional classicscientificmethodsofidentificationincludera- For 60years,theFBIDisasterSquadhasrespondedto Fingerprint expertscanprintpostmortemfrictionridge de- A numberofcrashsiteshavepresentedexceedinglychalleng- Methods employedtoidentifyremainsprimarilydependupon With thearduousdisasterscenariosofrecentyears,DNAhas

recovery situations:mountainside,tropicalrainforest,muddy

recovered asaresultofthemishap,and

the avail- many structures aredeveloping:handsand wrists,knees,feet, tant factorsinidentification. flight manifest.Fingerprintsand otherfeatureswerealsoimpor- associating mostofthehands withspecificjuvenileslistedonthe atlas ofhand-wristdevelopment. Ageestimateswerevaluablein derived fromanormalreference populationrepresentedinan hand, andwrist.Ageestimateswerecalculatedfromstatistics hand, basedonthedevelopmentofspecificbonesfingers, X-rayed. Anthropologistsestimatedapproximateagesfor each materials. Anumberofhandswererecoveredandsubsequently access, associatedenvironmentalhazards,andsunkenfragmented craft intoswamppresentedachallengingrecoverysite:difficult childhood, andadolescence.Forexample,thecrashofan air- of developmenttheskeletonanddentitionduringinfanc y, sional ID. sons listedontheflightmanifest,formingbasisforaprovi- The profilecanbecomparedwithphysicaldescriptionsofper- profile, anapproximatephysicaldescriptionofindividual. gender ofanindividual.Thecompileddataformsabiological estimates ofstature,agerange,muscularity,anddeterminethe construct andanalyzetheindividualbonesskulltomake mechanical traumaorheatfracturesproducedinafire. anthropologic interpretationastomechanismsofinjury,beit sented. Somepatternsofdamagetoboneareamenable fundamental toestablishinghowmanyindividualsarerepre- allows themtolocate,sort,andreassociatefragmentedremains, with thedistinctiveanatomicalfeaturesofhumanskeleton skeletal materialsfromthoseofanimalorigin.Theirfamiliarity rect excavationofdifficultrecoverysitesanddistinguishhuman tensive fragmentation,commingling,andburning.Theycandi- the identificationprocess,particularlyinscenariosinvolvingex- anthropologists iskeytosearch-and-recoveryeffortsaswell While weareonthesubjectofbones,expertiseforensic Anthropology mented forsubsequentidentificationprocedures. bones, displacedteeth,andjewelrycanbelocateddocu- items potentiallyhazardoustomorguepersonnel.Fractured their injurypatterns,butalsopersonaleffects,aircraftparts,and that canbetracedtothemanufacturer,andrecipient. nient isthatthelastthreeitemstypicallyhaveserialnumbers defibrillators canprovideidentifyingdata.Whatismoreconve- ration ofsurgicalplates,wires,implants,pacemakers,and identifying features.Whenviewedonradiographs,theconfigu- formation, evidenceofdisease,andsurgicalartifactscanserveas ties. Normalanatomicalvariationsoftheskeleton,abnormalbone view oftheskull,chest,abdomen,spinalcolumn,andextremi- tem full-bodyradiographs.Usefulmedicalradiographsinclude Medical X-rayimagesareusefulforcomparisonwithpostmor- Radiographic identification driver’s licenseapplicationsinCaliforniaandTexas. bases. Additionalsourcesforfingerprintrecordsnowinclude pedited thesearchthroughcivilandcriminalfingerprintdata- Fingerprint IdentificationSystem(IAFIS),havesignificantlyex- Estimates ofageatdeatharegenerally baseduponareaswhere Age attimeofdeathcanbeestimated,baseduponthestages Depending uponmaterialsavailable,anthropologistscanre- X-ray scanningofbodybagscanrevealnotonlyremainsand ISASI 2003 Proceedings • 87 ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS tinctive. Methods foraccessingDNAfromteeth havebeendevel- inherently unique.Whenaltered, teethbecomeevenmoredis- tours oftheinternalandexternal aspectsoftheteethmakethem protection fromdamagingenvironmental conditions.Thecon- and rootstructure.Teethareexcellentvesselsfor DNA, offering pulp andporousdentincomprising thegreaterpartofcrown for genderdetermination. females thanitisinmales,whichmakesaconvenientmarker The lengthofthegenethatcodesforthisproteinisshorter in amelogenin, aproteininvolvedintheproductionoftoothenamel. nuclei throughoutthebody,containagenethatcodes for ing thesubjectinquestion. probabilities indicatingthelikelihoodofremainsrepresent- same geneticprofile.Populationstatisticsprovidetheanalyst with individual inquestion,orthereisanotherwith the two conclusionscanbedrawn:Eithertheremainsrepresent file derivedfromthesubjectmatchesthatofaknownreference, reference profilesareexaminedforconsistencies.Whenthepro- close relative,canalsobeused.Thepatternsofthesubjectand reference, suchasabloodsampleorcheekswabobtainedfrom from atoothbrush,clothing,orotherpersonalitems.Anindirect men, oritcanbeprovidedbyahairsample,biologicalresidues dent sometimeduringlife,suchasabloodsample,biopsyspeci- known sources.Adirectreferencesampleistakenfromthedece- those derivedfromdirectorindirectreferencesamples profile generatedfromapostmortemsampleiscomparedwith the molecularreferencepointsbeingexamined. file thatreflectsthesequenceorlengthofcomponents ume. SuitablequantitiesofDNAareprocessedtogenerateapro- and fluids.SmallyieldsofDNAcanbecopiedtoamplifyitsvol- saliva, muscletissue,aswellavarietyofcell-containingtissues DNA fromavarietyofbiologicsamples:bone,teeth,wholeblood, per identification. Environmental insultscandamagetheDNAmoleculeandham- utilized toreassociateotherfragmentsfromthesameindividual. already identifiedbyothermeans,suchasteethorhands,canbe otherwise beidentified.DNAprofilesgeneratedfromremains offers ameansofestablishingidentityforremainsthatcouldnot a distinctivegraphicpattern,orDNAprofile.Thistechnology another. Thesemolecularvariationsareprocessedandviewedas human geneticmaterialthatcandistinguishoneindividualfrom DNA analysisisbaseduponexaminationofspecificsegments DNA or teeth. DNA analysisfromalready-identifiedstructures,suchashands fragments withthesameDNAprofile,samplesmaybetakenfor eton, orasasourceofDNA.Fpurposesreassociatingother still servetheidentificationeffortaspartofareconstructedskel- estimates. adults, degenerativechangesareusedformakingbroadagerange exhibit numerousareasofgrowthAsprocessesceasein age rangescanbederivedfromyoungerindividualswhostill and anklesjawswithdevelopingpermanentteeth.Narrower 88 DNA ispresentwithintheteeth, containedwithincellsofthe Interestingly, humanXandYchromosomes,presentin cell As withclassicformsofidentification,thepatternaDNA Molecular biologistshavedevelopedmethodsofextracting Portions ofbonelackingsofttissuesordistinctiveanatomycan • ISA SI 2003 P r oceedings cate thattheadjacent toothwasrestoredwith agoldalloy,while croanalysis oftracematerials. Tracesofelementalgoldcanindi- visual examinationofatooth surfaceandprovideelementalmi- persive X-rayfluorescence(EDS) canallowhighmagnification cent teeth.Scanningelectron microscopy(SEM)andenergy-dis- mation aboutmetallicrestorations thatmaybepresentonadja- plication anddamagesustainedinafire. hibit specificphysicalpropertiesthatinfluencetheirclinical ap- ins; andall-porcelainrestorations.Eachofthesematerials ex- alloys; castsemi-preciousorbasemetaltooth-colored res- metal alloyscomprisedofsilver,tin,copper,andmercury; gold for present/missing/filledteetharevast. as eithermetallicortoothcolored,thepotentialcombinations faces pertooththatcanberestored,andcategorizerestorations tions ofmissingandpresentteethareconsiderable.Addfive sur- other. With32possiblepermanentteeth,thepotentialcombina- tive, alterationscanfurtherdifferentiateoneindividualfroman- storative materials,andthermaldamagetonameafew. are alsocarriedout,dealingwithdentalappliances,specificre- issues indentalidentification,additionalanalysesandinquiries sent oneofthoseindividuals.Amongdealingwithmanyother those filestodeterminewhethertheremainsinquestionrepre- case description.DentiststhencomparetheX-raysanddatafrom directs thedentisttomostlikelyantemortemfilesthatfit for example,thecomputerquicklysortsavailabledataand queries thedatabaseregardingaspecificpostmortemcasefile, tem dataareenteredintoacomputerdatabase.Whendentist most recentvisittothedentist.Thesepostmortemandantemor- depicting theexistingdentalconditionsofanindividualat rays, andothermaterials,acompositechartisconstructed, types ofdataarecompiledfromantemortemdentalrecords,X- odontic appliances,andotherfeaturesaredocumented.Thesame and absenceofteeth,theirrestorationsreplacements,orth- data thatwouldbecompiledinaclinicalsetting.Thepresence graphs areconductedinastandardmannertogatherthesame the remainsinquestionrepresent. consistencies ordifferencesthatwouldconfirmrefutewhom and arecomparedwithanalogouspostmortemX-raystoreveal images. AntemortemX-rayimagesdocumenttheselandmarks ations areuniquethemselves,andalsoexaminedonX-ray outlines ofrestorations,rootcanaltreatments,andotheralter- nals arevisibleondentalradiographs.Theinternalandexternal placed singletooth,toformthebasisforidentification. enough toallowajawfragmentwithcoupleofteeth,ordis- and impact.Theirinternalexternalcontourscanbeunique acteristics throughtheeffectsoffire,decomposition,immersion, are durable,whichallowsthemtoretaintheirfundamentalchar- terns oflossaredistinguishingcharacteristicsreadilyseen.Teeth times. Variationsinshape,position,color,alterations,andpat- Teeth havebeenusedforidentificationsinceancientRoman Dental identification lent identifiers. oped topreservetheanatomiccontoursthatmaketeethexcel- Single displacedteeththatare notrestoredcanprovideinfor- Individual teethcanberestoredwithavarietyofmaterials: While untreatedteethandjawstructuresarealreadydistinc- Postmortem dentalexaminations,photographs,andradio- The outlineofthecrown,root(s),pulpchamber,androotca- be greatlyreduced,seeminglyinsignificantfragmentsbecome remains yettoberecovered.Asmaterialsinthesescenarios can In difficultfieldsituations,on-scenepersonnelmayencounter Considerations nel atthesceneandmortuaryencouragestheirbesteffort. challenging situation,thehumanitarianmissionservedbyperson- are fewermaterialswithwhichtomakeidentifications.Despitethe reduced, includingteethandbone.Recoveryisdifficult,there are highlyfragmentedaswellburned,theycanbesignificantly portable X-raysbeforestabilizingthemfortransport.Whenremains sonnel, whomaydocumentthematthesitewithphotographsand handled. Thesematerialsrequirecarefulhandlingbyforensicper- fire bysurroundingjawmuscles,cheeks,andtongue. become ashed.Posteriorteethareaffordedsomeprotectionfrom and teethcanbelost,eventuallytheremainingbone The outerlayerofbonethatsurroundstheteethcanbedestroyed develop cracks,andenamelcanseparatefromtheunderlyingtooth. ing durationandtemperature,teethcanchangecolor,dryout, mal damagetoteeth,bone,anddentalrestorations.Withincreas- useful inidentification. appearance onX-ray,andtherestorationstheysupport,canbe tured ofpuretitaniumspecificdesignanddimensions.Their replacement tooth,oralargerprosthesis.Manyaremanufac- ses can,tovariousextents,resisttheeffectsoffireandimpact. nations ofacrylics,resins,andmetalalloys,removableprosthe- patient forwhomtheyaremade.Constructedofvariouscombi- removable denturesandpartialsdiscretelybearthenameof person forwhomtheyaremade.Requiredbylawinmanystates, were exposedtointhemishap. tures, givinganindicationastothetemperatureoffirethey can begintodistortasafirenearstheirspecificfusingtempera- from theoralcavity.Fabricatedunderhightemperatures,they unique contoursusefulforidentification,evenwhendisplaced porcelain. Aswithnaturaldentition,thesestructurespossess the surfaceforesthetics,whilesomeareconstructedentirelyof bridges arecastinastrongmetalalloy,withporcelainfusedto supported byimplants.Manypermanentlycementedfixed the sameinformationforthosespecificteeth. computer databasetorevealwhichantemortemrecordscontain silver amalgamrestoration.Thisdatacanbesortedthroughthe the detectionofsilver,mercury,ortincanindicateanadjacent Burned teethandbonecanbeveryfragilecrumbleif As withaircraftcomponents,firecanproducecharacteristicther- Implants functionasartificialtoothroots,supportingasingle Like Cinderella’sglassslipper,removableappliancesfitthe Replacements formissingteethcanberemovable,fixed,or 24. CarrRF,BarsleyRE,DavenportJr.,WD,PostmortemExaminationof 23. MyersSL,WilliamsJM,HodgesJS, 22. SmithBC,APreliminaryReport:ProximalFacetAnalysisandtheRecov- 21. AmericanBoardofForensicOdontology,Inc.,ABFOGuidelinesand 20. LudesB.etal, 19. BrannonRB,MorlangWM, 18. AveryJK, 17. GaytmennR,SweetD, 16. SmithBC, 15. RudinN,InmanK, 14. WarrenMW,etal, 13. FBIDisasterSquadwww.fbi.gov/hq/lab/disaster/disaster.htm. 12. USDOJFBIHandbookofForensicServiceswww.fbi.gov/hq/lab/hand- 11. CimrmancicMA,GormleyWT,CinaSJ,AviationPathology,InHand- 10. FroedeRC,etal, 9. LuntzLL,P, 8. TransportationDisasterAssistance,www.ntsb.gov/Family/family.htm. 7. PublicLaw106-181,TitleIVSec.401-404–FamilyAssistance,April5,2000. 6. PublicLaw105-148,ForeignAirCarrierFamilySupportAct,December 5. PublicLaw104-264,TitleVII,AviationDisasterFamilyAssistanceActof 4. FederalFamilyAssistancePlanforAviationDisasters 3. DisasterMortuaryOperationalResponseTeam,www.dmort.org. 2. OfficeofEmergencypreparedness,NationalDisasterMedicalSystem, 1. LahmFP, References to servethesurvivingfamilies. In assistingthiseffort,airsafetyinvestigatorscanultimatelyhelp ing them,andnotifyingtheMEoftheirpresenceposition. and protectingthesematerialsatthescene,perhapsdocument- facilitate therecoveryandidentificationprocessbyrecognizing increasingly necessaryforidentification.Fieldinvestigatorscan Jan 1986,Vol.31,No.1,pp.307-311. Incinerated TeethwiththeScanningElectronMicroscope,JForensic Sci, Implications forHistologicProcessing Vol. 35,No.4,July1990,pp.873-880. ery ofTraceRestorativeMaterialsfromUnrestoredTeeth,JForensicSci, eds., AmericanSocietyofForensicOdontology,1995. Standards, InManualofForensicOdontology,3rd,BowersCM,BellGL, 1147-1152. Mount Ste-Odile,France J ForensicSci,2001;46(3):722-725. 2nd of HumanTeeth Vol. 45,No.2,April2001,pp.229-235. Press, BocaRaton,2002. Sci, 1999;45:467-470. book/examds.htm. thologists, NorthfieldIL,2003. book ofForensicPathology,2nd,FroedeRC,ed.CollegeAmericanPa- Northfield IL,2003. Forensic Pathology,2nd,FroedeRC,ed.CollegeofAmericanPathologists, Dentistry 16, 1997. 1996, October9,1996. Safety Board,August1,2000. www.ndms.dhhs.gov/NDMS/ndms.html. spection andSafetyUSAF. 1909; DirectorateofAerospaceSafety,DeputyInspectorGeneralforIn- Lt. ThomasE.Selfridge17September1908 , SteelePFed.,MosbyInc.,andSt.Louis,2000. , Lipincott,Philadelphia,1973. First UnitedStatesMilitaryAircraftAccident;Mr.OrvilleWrightand Essentials ofOralHistologyandEmbryology:AClinicalApproach, Introduction toDNAAnalysis , JForensicSci,May2003,Vol.48,No.3. Medico-Legal InvestigationsoftheAirbusA320Crashupon Investigation ofMultipleFatalityIncidents Use ofRadiographicAtlasesinaMassFatality Handbook forDentalIdentification;TechniquesinForensic , JForensicSci,Vol.39,No.5,September1994,pp. An IntroductiontoForensicDNAAnalysis,2 Quantification ofForensicDNAfromVariousRegions Tenerife Revisited:TheCriticalRoleofDentistry , JForensicSci,1999;44(4):805-809. N Effects ofExtremeHeatonTeethwith , DentalClinicsofNorthAmerica, , SignalCorps,USA,February9, ISASI 2003 , NationalTransportation , InHandbookof Proceedings , JForensic nd , CRC • 89 , ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS P information derived fromaccidentinvestigations toderivethe tion ismuchmoreeffectivethan treatmentaftertheaccident. vent accidentsofaircraftand other formsofvehicles,forpreven- than therapyafteroneisalready sick.Theirjob,then,istopre- colleagues inthatpreventionof illnessandinjuryisthegoalrather over thepast50years.They differfrommostoftheirmedical residency trainingandbecomingcertifiedintheUnited States cialties, withabout1,000physicianscompletingthe3years of the UnitedStates,aerospacemedicineissmallestofall spe- cine. Oneofthe24recognizedmedicalspecialties in areas ofpreventivemedicineanditssubspecialty,aerospace medi- ning offlightbeenthecausemostaviationaccidents. craft physician isresponsibleforthemostimportantpartof air- and consistofaweek-longcoursetaughtinOklahomaCity. This perform flightexaminations,thetrainingrequirementsarebasic AME isdesignatedbytheFederalAviationAdministrationto flight surgeon,oraviationmedicalexaminer(AME).Whilethe neer, alsoknownastheaerospacemedicinephysician(AMP), dent investigationteamisthehumansystemsmaintenanceengi- systems, andoperationaldata,aneffectivememberofacci- the multidisciplinaryworkinggroupsfocuseduponstructures, united intheirpursuitoftheadvancementaviationsafety.Amid the progressofaviationsafety. earliest daysofflight,physicianshaveplayedanintegralrolein quent actioninthelitigationandregulatoryarenas.Since hap areessentialforpurposesofestablishingcauseandsubse- and improvingmishapsurvivability.Thirdly,thefactsofmis- death andinjury,withtheobjectiveofmodifyingthosefactors Kansas City,Mo. currently isinprivatepracticeatMidwestOccupationalMedicine 1992. HewasthemedicaldirectorofTransWorldAirlinesand Tex. Afternumerousassignments,heretiredfromtheAirForcein medicine attheUSAFSchoolofAerospaceMedicineinSanAntonio, 90 While notallaccidentscanbe prevented,theAMPcanuse The AMPisamedicalspecialistwhofunctionswithinin the Experts inadiversearrayofdisciplines,ISASImembersare • ISA : thepilot.Malfunctionofpilothasfromverybegin- SI 2003 purpose forconductinganinquiryistoderivecausesof goal ofaircraftaccidentinvestigation.Thesecondary revention offuturemishapshaslongbeentheprimary Aircraft AccidentInvestigation— P r Texas andcompletedaresidencyinaerospace and mastersinpublichealthfromtheUniversityof He receivedhisM.D.fromtheUniversityofKansas Vietnam andNORADbeforegoingtomedicalschool. Academy andservedonactivedutywithtoursin engineering fromtheUnitedStatesAirForce oceedings Dr. AllenParmet The RoleofAerospaceand Preventive Medicine receivedhisB.S.inchemical By AllenJ.Parmet,M.D.,M.P.H.,F.A.C..M. 1980. Justbehind himwasanAmericanphysician withateamof trian ReinholdMessnerclimbed Mt.Everestwithoutoxygenin m), whichwouldultimatelybe provenacenturylaterwhenAus- pressures below45mmHg(equivalent toairat33,000ft/10,000 periments thatdetermined humanscouldnotliveatoxygen poxia. Followingthis,theFrenchphysician PaulBertbeganex- but theywereunconsciousabove 8,833m(27,382ft)duetohy- aeronauts GlaisherandCoxwellascendedto9,480m(29,388 ft), lems ofcoldandhypoxiabecameapparent.In1862theEnglish and high-altitudeballooningcameintobeing.Soon,the prob- altitudes weremadewithsafermethodsofhandlinghydrogen, say thenthatphysiciansinventedtheaviationaccident. combination, anddeRozierdiedinthefierycrash.One might balloon. Hothydrogen,however,provedtobeaverydangerous invention ofahydrogenballoonwiththeMontgolfier’shot air aloft inahotairballoon.DeRozierdecidedtocombineCharles’ of muchstrongerprevailingwindsandneededtospendlesstime Jeffries, onJanuary7ofthatyear,butJeffrieshadtheadvantage Channel. HewasprecededbytheAmericanphysicianJohn a balloonfromFrancetoEngland,westboundacrosstheEnglish Montgolfier brothersballoonin1783,soughttobethefirstfly physician whohadbeenonboardtheveryfirstflightofa true accidentoccurredin1785whenPilâtredeRozier,aFrench been numerousaviationaccidentsanddeaths.Probablythefirst tors. EvenbeforetheWrightbrothersflewin1903,therehad participant inaircraftaccidentinvestigations. specialist ofthehumanaspectisnotaproportionatelyroutine proportion ofaviationmishapcausationishumanrelated,the gation closesandwhenlitigationstarts.Althoughaconsiderable through invitationofthesafetyinvestigatorsorafterinvesti- accidents, whentheAMPiscalledtoparticipate,itusually gations haveaflightsurgeonasmemberofeveryboard.Incivil Working Groups.However,U.S.militaryaviationmishapinvesti- provide traumadatatotheSurvivalFactors/Crashworthiness participation ofthelocalmedicalexaminerorcoroner,whomay Group inU.S.civilaviationmishapinvestigations,asidefromthe pert witnessinhelpingderiveforensicandlegalconclusions. find itswayintothecourtroomasAMPmaybecomeanex- and improvecrash/rescueoperations.Theinformationmayalso training tohelppreventfutureaccidents,designsaferaircraft, tors, andprovidingforsurvivalrescueinthepost-crashphase. tion ofoccupants,reductioncrash-relatedenvironmentalfac- dition ofengineeringchangesforbothactiveandpassiveprotec- causes ofinjuryandhelpmodifythosefactors.Thisleadstoad- Over thenexthundredyears,ascentstohigherand The primarycausesofaccidentshavealwaysbeenhumanfac- Customarily, thereisnoseparateMedicalFactorsWorking The lessonslearnedfromaccidentsarealsotranslatedinto become thefirst federalairsurgeon.Dr.Bauerwould begin the In 1926Dr.LouisBauerwouldbereassigned fromtheArmyto would screenoutnearly30percent ofallU.S.flyingapplicants. flight surgeonfortheArmyflying andestablishedstandardsthat In theUnitedStates,MajorTheodore Lysterbecamethehead military doctor,knownasasurgeon, wasassignedtoflightunits. pilot justsatthere.Medicalregulations weresooninplaceanda Aviation wasdeemedtobemuchlikeanofficejob;afterall, the ered unsuitableforsuchmilitaryarmsastheinfantryorcavalr y. ing ulcers(RoyBrown),andpsychosis(FrankLuke)wereconsid- Guynemer), blindness(WilliamThaw,EdwardMannock), bleed- and epilepsy(ManfredvonRichtoffen),tuberculosis(Georges ditions suchasasthma(GermanOswaldBoelcke),skullfracture Pilots wereenteringtrainingandoftendyingthere.Medical con- problems accountedfor8percentandcombatamere2percent. accidents, spatialdisorientationanother30percent.Mechanical the mixture. would todaycallhumanfactors.Whathaschanged,however,is mere 2weeks.Ofthedeaths,90percentwereduetowhatwe Service in1915foundthatthelifeexpectancyofapilotwas dards—particularly afterearlystudiesofBritain’sRoyalFlying military activities,andmostmilitariesdevelopedmedicalstan- sign changesinhelmetsafety. not beuntilthe1940sthatDr.JohnPaulStappwouldleadde- to weartheirWestPointfootballhelmetswhileflying.Itwould leagues, suchasLt.Henry(Hap)Arnold,werelaterencouraged the aircraftonimpactanddiedofaskullfracture.Hisarmycol- ducted theautopsyandfoundthatSelfridgewasthrownoutof passenger, ArmyLt.ThomasSelfridge.Ansurgeoncon- onstration flight,seriouslyinjuringOrvilleWrightandkillinghis 1908, atFt.Myer,Va.Thestarboardpropellerfailedonadem- cause ofaccidents. induced lossofconsciousness(G-LOC)becameanadditional beginning todogfightintherealmabove5gsandacceleration- capable of5-gpulloutsinthe1930s.Soonfighteraircraftwere dent untiltheU.S.Navyinventeddivebombingandaplane Finally theproblemofaccelerationforceswouldnotbecomeevi- gases bothinthesystemsandeventofcrashesfires. with lowbarometricpressureataltitudesover50,000feet,toxic ins cameintouseinthe1940s.Otherphysiologicissuesremain until oxygensupplysystemswereperfectedandpressurizedcab- cause ofaviationaccidents.Hypoxiawouldremainachallenge would cometoahaltuntilthe1930s. disaster wouldrockAmerica.Asaresult,high-altitudeattempts a nationaltragedythatshookFranceasmuchtheChallenger were unconsciousandonlyTissandiersurvived.Theaccidentwas the trioascended.Whiletheyprobablydidexceedtheirgoal,all Bert warnedthemthatthiswasinadequate,butonApril15,1875, 72 percentoxygenandsimpletubesthatwereheldinthemouth. primitive oxygensupplysystem,whichconsistedofthreebags they attemptedtosetanewaltituderecordof10,000musing nauts, Crocé-Spinelli,SivelandTissandier.Inthesummerof1874, Step inordertoreplicateBert’swork. Sherpas carryingabicycleergometertothebaseofHillary In 1915,medicalconditionswerethecauseof60percent of Within afewyears,aviationwasanimportantpartofworld The firstfatalaccidentofanairplaneoccurredonSept.17, Physiologic issueshadbecomeestablishedasonefactorinthe Bert wasalsothemedicaladvisortoateamofFrenchaero- placently, task saturation, fatigue,andcomplacency allhaveroles ability, rate,experienceandtransfer. Attention,boredom,com- judgment, butmanyfactors come intoplayincludinglearning and drugsarethemainissues. Itisdifficulttoevaluateapilot’s highways andmotorcycle footballhelmets. plied tomanyotherareasof safetyincludingautomobilesand discovered thatwhatevercangowrong,eventuallywill. seating arrangements.Stapp’schiefengineer,EdMurphy,also shock absorbers,passiverestraints,crashhelmets,airbags, and work. Stapp’steamwoulddevelopthestandardsforejection seats, and allmodernenergyabsorptionlimitsarederivedfrom his veloped thelimitsofhumantolerancetoimpactaccelerations (now knownasEdwardsAirForceBase).Hisresearchteam de- ranch inCalifornia,usingMurocdrylakebedforsledtesting John PaulStappbeganhisimpactaccelerationworkontheCorum ommended freefalltoloweraltitudes.AfterWorldWarII,Dr. ered thehighopeningshockforces,cold,andhypoxiarec- bailouts wereresearchedbyDr.RandallLovelace,whodiscov- support andescape,aswellcrashworthiness.High-altitude strument simulators,laterbroughtintoitscommonformbyLear. use theirinstruments.The“OckerBox”wastheforerunnerofin- the groundandtoinstillinthemconfidencetheyneeded and bankindicatortotrainpilotsintheeffectsofvertigowhileon developed acombinationofrotatingchairequippedwithturn but itmetwithresistancefrompilotswhodistrustedit.Dr.Ocker solution tothisproblemwastheSperryturn-and-bankindicator, began toflyintoclouds,badweather,andatnight.Thefirstreal Hostile actionssuchasterrorismarestill2percent. collisions 6percent,andmechanicalproblemsonly2percent. 38 percent,drugsandalcoholare6-9midairorground spatial disorientation36percent,controlledflightintoterrain craft accidents.Todaymedicalaccountsfor2percentofaccidents, However, humanfactorsstillcomprise80-90percentofallair- mized medicalconditionsastheprimarycauseofaccidents. personnel involvedinwreckagesearchandrecovery. military flightsurgeonswhodeployedandcaredforthe cals usedforpropulsionandpower.Theauthorhadtrainedthe plicated bythehazardstopeopleongroundfromtoxicchemi- such acatastrophe.ThelossofspaceshuttleColumbiawascom- well ascounselingandhelpingcompanyemployeesaffectedby the NTSB,supportfordisasterresponseandfamilyassistance,as data tothelocalmedicalexaminer,crewinformation loss ofTWA800,theauthor’srolefelltoprovidingidentification sive decompression,notterroristbombs.AstheAMPduring directed byanAMP,whonotedthatthedeathswereduetoexplo- trast, theinvestigationof1950scrashesCometIAwere evidence. Bodieswerebarelyidentifiedbytheirclothing.Incon- Spectators drovearoundthesite,takingsouvenirsanddestroying Rockne. Therewasnonationalsystemforinvestigatingaccidents. in ChaseCounty,Kans.,thatkilledNotreDamecoachKnute high profileaircraftaccident,the1931crashofaFokkerTrimotor cal standardsforcivilpilotswerealsoestablishedatthistime. training programofAMEsthatisthestandardtoday.Themedi- Accidents stilloccurduetopsychological factors.Judgment The discoveriesinaviationsafetyweretoeventuallybe ap- AMP physicianscontinuedtheirworkimprovingvision,life Spatial disorientationaccidentsbegantooccurassoonpilots The impositionofmedicalstandardsonpilotseffectivelymini- The presenceofAMEswouldnotdomuchtoaffectthefirst- ISASI 2003 Proceedings • 91 ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS alty issues.Leavingoutthepreventivemedicinephysicianmeans the medicaltriageteamsandshouldbeinvolvedinmass-casu- the nextaccidentfromoccurringandreducinginjury. the AMPwillhelpinrecommendingremedialactionstoprevent cape, andPost-crashfactors.Asaresultofthisaccidentanalysis, analyzed usingCREEP:Container,Restraints,Environment, Es- bon monoxidepoisoning,orhypoxia.Crash-relatedinjuryis tion maybeduetomedicalcausessuchascardiacdisease,car- tated priorto,during,orafterthecrash.Pre-crashincapacita- portant todetermineifthepilotandpassengerswereincapaci- turn thatinformationbacktopreventthenextaccident.Itisim- it beassumedthatthemedicalexaminerwouldautomatically mining whowasatthecontrolsofamultipilotaircraft.Norshould to thespecificneedsofaccidentinvestigator,suchasdeter- cal examiner.Medicalexaminersmaynotnecessarilybeattuned causes ofdeath/Injuryalongwiththeforensicpathologist/medi- tion andhelpdeterminethecauseofaccident,aswell of thosewithillicitdrugproblems. duty 90percentofpilotswithalcoholproblemsand50 nated bytheFAAandALPAin1960s,successfullyreturnsto who areaddicted.Treatmentworks.TheHIMSprogram,origi- does servetodeterusebycasualillicitdrugusers,butnotthose a societalproblemaswellanaviationsafetyissue.Drugtesting a growingproblem.Thedetectionanddeterrenceoftheiruseis prescription andnon-prescriptiondrugsaswellillicitis and militaryaviation.Howeveringeneralaviation,theuseof nificantly toaviationaccidents.Theiruseisrareincommercial scene ofanaccident. The unsafeactisadecisionmadebythepilottoarrivefirstat as medicalproblems,weatherortraffic,andfinallyunsafeacts. or mechanicalproblems.Activeissuesincludepreconditionssuch sure toperformschedulesandignorecrewrest,fuelreserves, zational andsupervisorypreconditions.Thesemayincludepres- fied intolatentandactiveissues.Latentissuesconsistoforgani- is inevitablytheendresultofachainerrors.Theseareclassi- Factors AnalysisandClassificationStytemsin2000.”Anaccident after Drs.ShappellandWeigman,intheirarticle“TheHuman air trafficcontrolallinteract. issues ofaircraftsystems,transitions,maintenance,weather,and ordination andcockpitresourcemanagement.Organizational there aresupervisoryandmanagementissuesaswellcrewco- ily, andcommunityconflicts.Withintheoperatingenvironment psychosocial factorsofjobsatisfaction,careerexpectations,fam- sory pressuresandcumulativeworkloadinteractwithoutside to play.Personalitystatesofself-discipline,motivation,supervi- 92 The aerospacemedicinephysiciancanbearesourcetoassist The AMPshouldparticipateintheaircraftaccidentinvestiga- Drugs andalcoholarecommoninsocietycontributesig- Analysis ofthesefactorsisknownasthe“SwissCheeseModel” • ISA SI 2003 P r oceedings Whinnery, JE. Sheffield PJ,HeimbachRD Parmet AJ,GillinghamK, McMeekin R. Gibbons HL,SnyderRG, DeHart RL, References fessionals, athttp://www.asma.org. ciation andcanbecontacted,alongwithotheraeromedicalpro- them. MostAMPsaremembersoftheAerospaceMedicalAsso- will continueaslongthereareairplanesandpeoplewhofly cluding pilots,passengers,rescuersandinvestigators.Thisrole future accidentsandillnessineveryoneinvolvedaviation,in- check Travelers’Healthforthelatestareaassessment. ease ControlandPreventionwebsiteathttp://www.cdc.gov vaccinations arealsoofuse.AlwaysconsulttheCentersforDis- what areasareatriskanduseinsectrepellants.Medications fever, dengue,Lyme,andplague.Thebestprotectionistoknow and fleascarryillnessessuchasmalaria,WestNilevirus,yellow nesses suchastuberculosisandSARSarerealrisksinsomeareas. may beanissueduetothesurroundingsocialconditions.Ill- ease frompersontoisunlikelyattheaccidentsitebut polio canbepreventedbyvaccination.Airbornespreadofdis- of contamination.SuchdiseasesashepatitisA,typhoidfever,and C orHIV,theAIDSvirus.AvaccineexistsonlyforhepatitisB. tigators toavoidcontaminationbysuchdiseasesashepatitisBand cal hazard.Personalprotectiveequipmentmustbewornbyinves- spillage ofbloodandhumanbodyproductsrepresentsabiologi- can spreadtotheinvestigatorsthroughfourmethods. alties. Finally,thereistheproblemofinfectiousdiseases.Disease to preventtheinvestigatingteamfrombecomingadditionalcasu- ronment existsandadditionalprotectivemeasuresmustbetaken times ahazardouschemical,high-altitude,orunderwaterenvi- arrangements, andthermalprotectionneedtobeaddressed.Some- human healthhazard. gators. Theenvironmentoftheaccidentsitemayrepresenta ventive measuresneededforbothrescuersandaccidentinvesti- feeding isabasicpublichealthfunction.Finally,therearepre- jured duringtheresponsephaseandrecovery.Theircare that thedisasterrespondersmaythemselvesbecomeillorin- space Medicine,3rdEd,DeHart&Davis,2002. space Medicine,3rdEd,DeHart&Davis,2002. Aerospace Medicine,3rdEd,DeHart&Davis,2002. Medicine, 3rdEd,DeHart&Davis,2002. Medicine, 3rdEd,DeHart&Davis,2002. Ed, DeHart&Davis,2002. The aerospacemedicinephysicianhastheroleofpreventing Last istheproblemofvector-bornediseases.Mosquitoes,ticks, Food andwatersuppliesmustbesecuredtoavoidthespread First isbloodandbodyfluids.Anyareawheretherehasbeen Environmental issuessuchascleanwaterandfood,sleeping The Atmosphere Aircraft AccidentInvestigation Biodynamics: SustainedAcceleration Aircraft Accidents . RespiratoryPhysiology Spatial OrientationinFlight , inFundamentalsofAerospaceMedicine,3rd N , inFundamentalsofAerospace , inFundamentalsofAerospace , inFundamentalsofAero- , inFundamentalsofAero- , inFundamentalsof well asguidance materialandprocedures containedinDoc. dards andRecommendedPractices containedinAnnex13,as would enabletheStatetoimplement andadheretotheStan- and qualifiedexpertsthe availability ofotherresourceswhich investigations) includingtheavailability ofappropriatelytrained latory frameworkanditsorganizational structure(foraccident incident investigationsthrough anevaluationoftheState’sregu- evaluate aState’sabilitytoconducteffectiveaircraftaccident and 1.2 TheconceptofanAnnex13auditwouldbetoassess and ment ofnationalaccidentinvestigationandpreventionprograms. States thatproceduresandguidancetoassistwiththeestablish- has beeninexistenceformorethan30years,thusproviding ance materialintheformofanaccidentinvestigationmanual prevention ofaccidentsandincidents.Inaddition,ICAO guid- taken byaStatewhichdirectlyrelatetotheinvestigation and for auditing.Itsprovisionsclearlydefineactionsrequired tobe Incident Investigation program (USOAP)toincludeAnnex13— 1.1 AnexpansionoftheICAOuniversalsafetyoversightaudit 1. General 007, aBoeing747onAug.31,1983. ber 1992-June1993)onthedestructionofKoreanAirLinesFlight Team leaderfortheICAOteamcompletinginvestigation(Decem- civil aircraftbyCubanmilitaryonFeb.24,1996.1993— ICAO teaminvestigatingandreportingontheshootingdownoftwo New Delhi,India,onNov.12,1996.1996—Teamleaderforthe Arabian AirlinesBoeing747andtheKazakhstanIL-76near Arabia, fortheinvestigationofmid-aircollisionbetweenSaudi cooperation assignmentasaccidentinvestigator,assignedtoSaudi Air Boeing747inGuamonAug.6,1997.1996—Atechnical Republic ofKorea,fortheinvestigationaccidenttoKorean cooperation assignmentasaccidentinvestigator,assignedtothe special assignmentsincludethefollowing:1997—Atechnical in Montreal.In1996hewasappointedChiefofthesection.Frostell’s In 1980,hejoinedtheAccidentInvestigationandPreventionSection the BoardofAviationinFinlandandinvestigatedsome300accidents. Include Annex13—AircraftAccident Safety OversightAuditProgramto Expansion oftheICAOUniversal to 1980,hewasChiefofAccidentInvestigationwith Studies, UniversityofToronto,Canada.From1967 applied sciencedegreefromtheInstituteofAerospace University inHelsinki,Finland,andamasterof in aeronauticalengineeringfromtheTechnical flight trainingintheAirForceFinland,adegree Caj Frostell’s By CajFrostell(MO3596),Chief,AccidentInvestigationandPrevention,ICAO isfeasibleasAnnex13inaformatsuitable And IncidentInvestigation educationalbackgroundincludesbasic Aircraft Accidentand 6920— investigation, the developmentsofthefinal report includingthe i.e., thedefinitions,applicability, thenotificationprocess, the implementationofSARPs coversallaspectsofAnnex13, 2.2 ImplementationoftheSARPs inAnnex13.Itisessentialthat • • • • • • • • tices (SARPs)inAnnex13 • following aspectsofaircraftaccidentinvestigation: 2.1 ItisenvisagedthattheAnnex13auditswillfocuson the 2. ContentofAnnex13audits 2004. Annex 13auditsisenvisagedtocommenceinthefirsthalfof schedule areplannedforlate2003.Theconductoftheactual ment ofrequiredpersonnel,andthedevelopmentanaudit development ofthememorandumunderstanding,recruit- Manual havebeeninitiated.Administrativeactionssuchasthe tors, andanamendmentoftheICAOSafetyOversightAudit audit questionnaire,protocols,trainingcoursesforaudi- 1.4 Preparatoryactivitiessuchasthedevelopmentofpre- verification duringtheactualaudit. provisions. AllinformationprovidedbyStateswillbesubjectto may existbetweentheirnationalregulationsandtheAnnex13 their nationallegislationandtoidentifyanydifferenceswhich the Annex13StandardsandRecommendedPractices(SARPs)in assist Statesinascertainingthestatusoftheirimplementation for auditschedulingandplanningpurposes.Inaddition,itwill sential auditingtool,designedtosolicittheinformationrequired related guidancematerial.Thepre-auditquestionnaireisanes- dit questionnaires,auditprotocols,auditors’trainingcourses,and developing relevantauditingdocumentation,includingpre-au- 1.3 ThepreparatoryworkfortheAnnex13auditsisfocussedon and Planning of AircraftAccidentandIncidentInvestigation Accident pr R Equipment P F P Legislation The agencyr Implementation oftheStandar ersonnel unding olicy and/orpr eview oftheinvestigationsome r Manual ofAircraftAccidentInvestigation ,andDoc.9422— evention measur esponsible forair ocedur es manual(s) es Accident PreventionManual craft accidentinvestigation ds andR ecent accidents ISASI 2003 , , Doc.9756— ecommended P Part I:

Proceedings Organization Manual . rac- • 93 ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS restricted access toallevidenceandwitnesses. Thenon-disclo- accident investigators(andthe investigationagency)tohaveun- the legislation,aswell rightsandresponsibilitiesofthe jective andthescopeofinvestigations wouldbecoveredby cent oftheStandardsare).It wouldappearevidentthattheob- are suitableforlegislativematerial (perhapsonly20to30per- 2.7 Legislation.Itisapparent thatnotallStandardsinAnnex13 administrative channels. investigation arenotdilutedduringpassagethroughregular ernment sothatthefindingsandsafetyrecommendations ofthe that suchacommissionreportdirecttoministeriallevelof gov- seconded fromthecivilaviationadministration.Itisessential jor accidenttobeinvestigated,themembersofwhichare often point aseparateaccidentinvestigationcommissionforeach ma- nent accidentinvestigationauthority.TheseStatesgenerallyap- 2.6 InmanyStates,itmaynotbepracticaltoestablishaperma- Parliament oraministeriallevelofgovernment. the accidentinvestigationauthorityusuallyreportstoCongress, separate fromthecivilaviationadministration.IntheseStates, or byestablishinganaccidentinvestigationorganizationthatis accident investigationauthorityasanindependentstatutorybody dards. ManyStateshaveachievedthisobjectivebysettinguptheir policymaking orfailingsinthesettingandpolicingofsafetystan- not feelconstrainedtoconsiderandaddressapparentflawed oversight oftheaviationsystem.Accidentinvestigatorsshould sidered criticaloftheregulatoryorganizationsthatprovidesafety deficiencies andthedevelopmentoffindingsthatcouldbecon- cause anyinvestigationmayresultintheidentificationofsafety authority overitsconduct.Independenceisseenasessentialbe- in theconductofinvestigationandmusthaveunrestricted 2.5 Theaccidentinvestigationagencymusthaveindependence tive 94/56/EC). tion onInternationalCivilAviation(andtheEUCouncilDirec- • lic thataccidentsandincidentsarethoroughlyinvestigated; • future; dents andincidentssothataccidentsmaybeavoidedin • accident investigationareto 2.4 Theessentialfunctionsoftheagencyresponsibleforaircraft to undertaketheinvestigations. dent investigationauthority,commission,board,orotherbody accidents anddesignateanappropriateagencysuchasacci- islation mustestablishaprocessfortheinvestigationofaircraft legislation onaircraftaccidentinvestigationisinplace.Suchleg- the accident.Thisobligationcanonlybemetwhenappropriate accident occurstoinstituteaninquiryintothecircumstancesof Civil Aviation,itisincumbentontheStateinwhichanaircraft conformity withArticle26oftheConventiononInternational 2.3 Theagencyresponsibleforaircraftaccidentinvestigation vention measurescalledforinChapter8ofAnnex13. Incident DataReporting(ADREP)System,andtheaccidentpre- safety recommendations,thereportingtoICAOAccident/ 94 fulfil theStates’obligationsunderAnne maintain theconfidenceofaviationindustryandpub- identify aviationsafetydeficienciesbytheinvestigationofacci- • ISA SI 2003 P r oceedings x 13totheConven- . In appropriate investigation procedures. vestigations inordertoensure compliancewithAnnex13and interview theinvestigator-in-charge foroneormoreofthesein- dents inadvanceoftheaudit. Theauditteammayalsowishto audit teamwouldreviewthe finalreportsonsomerecentacci- 6.7 ofAnnex13,theFinalReports shallbeavailableandthe years intheStatetobeaudited. Inaccordancewithparagraph tain alistoftheaccidentstoaircraftover2,250kginlast 3-5 ICAO Accident/IncidentDataReporting(ADREP)System toob- preparation foranaudit,theICAOauditteamwoulduse the 2.12 Reviewoftheinvestigationsomerecentaccidents. In the bloodborne pathogens. sonal protectiveequipmentagainstbiologicalhazards,such as tion andhepatitisimmunization,aswellthenecessary per- wreckage areadvisedtohaveavalidanti-tetanusseruminocula- quirements, andtravelfacilities.Investigatorswhoworkamong should alsobegiventosuchdetailsasinoculations,passportre- ceed withoutdelaytotheaccidentsite.Advanceconsideration essential personalitemspackedandreadysothattheycanpro- dent investigatorsshouldhavetheirinvestigationfieldkitsand tent ofinvestigationfieldkit(s)wouldbepartanaudit.Acci- and therecordingofobservations.Theavailabilitycon- ting ofimpactpointsandwreckagepatterns,partsidentification, cient equipmenttoenableexaminationofthewreckage,plot- 2.11 Equipment.Theinvestigationfieldkitshouldcontainsuffi- etc.). tion (airtrafficservices,meteorology,airports,humanfactors, ments forthecoverageofotheressentialareasinaninvestiga- The accidentinvestigationagencywouldneedtomakearrange- aeronautical engineer/aircraftmaintenanceengineerbackground. the corestaffwouldhaveaprofessionalpilotbackgroundand tional investigatorsifrequiredonasecondmentbasis.Normally, investigation, aswellaprocedureinplacetoacquireaddi- pected tohaveacorestaff,competentandtrainedinaccident 2.10 Personnel.Theaccidentinvestigationagencywouldbeex- pected tobecoveredbyanannualbudgetinsmallerStates. event, andsuchamajorinvestigationwouldnormallynotbeex- supplementary fundingasrequired.Anairlineaccidentisarare ments foraccidentinvestigation,provisionshouldbemade Since itisimpossibletoaccuratelyforecastannualbudgetrequire- accidents andincidentsthatfallwithinitsareaofresponsibility. access tosufficientfundsenableitproperlyinvestigatethose 2.9 Funding.Theaccidentinvestigationagencyshouldhaveready manual(s) issuedbytheinvestigationauthority. ered byregulationsorbeincludedinpolicyandprocedures that arenotsuitableforinclusioninlegislationshouldbecov- 2.8 PolicyandProceduresManual(s).ThoseSARPsinAnnex13 used whennecessaryandwiththeutmostdiscretion. tain relevantdocuments.Thesepowersshould,however,onlybe evidence, therighttotestanythingseized,andob- powers, includingauthorityoveranaccidentsite,possessionof ties, accidentinvestigatorsshouldbegrantedsuitablestatutory constitute legislativematerial.Toeffectivelydischargetheirdu- sure ofcertainrecords(Paragraph5.12Annex13)wouldalso • accident investigation; craft andtheirsystemshaveaddednewchallengestothetaskof • dents; tive 94/59/EC)toinvestigateseriousincidents,aswellacci- • workload considerably.Theseinclude come partoftheinvestigationprocessandhaveincreased constant. Inthelast10years,manynewconsiderationshavebe- workload isincreasingalthoughthenumberofinvestigationsstays ing withtheincreasingcomplexityofaviation.Asaresult, and tasksfacedbyaccidentinvestigatorsarecontinuallychang- the numberofaccidents,itshouldbenotedthatchallenges 3.1 Althoughtheremaynotbeanynoticeableupwardtrendin 3. Newchallenges can bemadeavailabletootherStatesonrequest. that eitherthroughEUorICAO,theECCAIRSdatabasesystem fully compatiblewiththeICAOADREPsystem.Itisenvisaged tre forAviationIncidentReportingSystem(ECCAIRS),whichis and incidentdatabasesystem,theEuropeanCo-ordinationCen- to notethattheEuropeanUnion(EU)hasestablishedanaccident sources, andanaccident/incidentdatasystem.Itmaybeofinterest punitive innatureandprovidesprotectionoftheinformation report system,avoluntaryincidentreportingsystemthatisnon- accident preventionmeasuresandcallsforamandatoryincident 2.13 Accidentpreventionmeasures.Chapter8ofAnnex13covers increased incomplexity;forexample,investigatorsarerequired International anddomesticpr Advances intechnologyandthecomple The internationalr equir ements (Anne ocedur x 13andtheEUDir es andpr xity ofmodernair- ocesses have ec- get datesforcompletionoffinalreportsinvestigations. investigation authoritiesandarelikelytohaveanimpactontar- 3.3 Allthesefactorsindicateagrowingworkloadfortheaccident preparing accidentreports. nology efficientlyandeffectivelyinconductinginvestigations the necessaryskillsandcompetenciestouseinformationtech- • authorities. requires anincreaseinliaisonwiththelocalemergencyplanning • assistance programsarerapidlyincreasingworldwide. source implications.Also,theawarenessandexpectationsoffamily tention. Itisnotataskthatcanberushedand,assuch,hasre- • tion topursuecompensationclaims. from legalrepresentativesofvictims’familiesseekinginforma- inquiries. Thereisalsothequestionofmanagementinquiries coroner’s courtsandinqueststoappearinofpublic nal) toinvestigationwork.Investigatorsarerequiredattend • to beaddressedbyinvestigators. 3.2 Therearealsoanumberofnon-investigativeareasthatneed important elementintheconductofinvestigations. of investigators.Healthandsafetyconsiderationsarenowan • cation; and to consultinterestedpartiesondraftreportsbeforetheirpubli- Ther Emer The taskoffamilyliaisonr Ther Health andsafetyr e isanincr e isanincr gency planningisalsoanar easing legaldimension(bothcivilandcrimi- easing r equir equir ements haveincr equir ement forinvestigatorstohave es car ea thathasdevelopedand eful andappr ISASI 2003 eased theworkload Proceedings opriate at- N • 95 ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS 96 • ISA SI 2003 P r oceedings SESSION V Figure 1 won someawardsforitsCFITwork,andtheratestartedto on CFIT.TheFoundationsentoutthousandsofthetrainingaids, of theCFITchecklist,avideo,andtwovolumesinformation products. TheseincludedtheCFITtrainingaid,whichconsisted were released.Inadditiontorecommendations,therealso aviation experts,recommendationsconcerningCFITprevention makes itevidentthatthechallengeofCFITandALAstillexists. to baseinterventionson.Figure2,showingcurrentnumbers, confirm that.However,therewasnostudy,nothingquantitative lems intheseareas;therewasalotofqualitativeinformationto most commontypeofaccident.Everyoneknewtherewereprob- ing causeoffatalitiesincommercialaviation,andALAwasthe 1 showswhythiseffortwasinitiated.In1992CFITthelead- bird strikesareallimportantsafetyissues,butthechartinFigure portant? Yes,runwayincursions,uncontainedenginefailures,and incursions oruncontainedenginefailures?Aren’tbirdstrikesim- Many mightask“WhyCFITandALA?”Whynotstudyrunway E was thedirectorofSchoolAviationSafetyinMonterey,Calif. his 30-yearcareer.PriortojoiningtheFlightSafetyFoundation,he having commandedanattacksquadronandacarrierairwingduring In 1996,after3yearsofworkbymorethan150international terrain (CFIT)andapproachlandingaccidents(ALA). on aprojecttoreducetheriskofcontrolledflightinto leven yearsagotheFlightSafetyFoundationembarked The CFITandALARChallenge: Attacking theKillersinAviation influences onsafety.HeisaretiredNavycaptain, investigation, safetyeducation,andorganizational factors, safetyprogramorganization,accident terrain, approachandlandingaccidents,human work inaviationsafetyincludescontrolledflightinto 27 yearsofexperienceintheaviationsafetyfield.His Jim Burin By JimBurin(M04448),DirectorofTechnicalPrograms,FlightSafetyFoundation has35yearsofaviationexperienceand an averageoffour commercialjetCFITaccidents ayearforthe ness, andsomenewexciting technologies,therehavebeen for theyear.Despiteafocuson training,effortstoincreaseaware- 1999, 14wereCFIT,accountingfor80percentof theirfatalities in 1999.However,ofthe28commercial turbopropaccidentsin from Figure3,therewasonly onecommercialjetCFITaccident dents happenedduringapproach andlanding.Asyoucansee the last10years,96percentofcommercialjetCFIT acci- any phaseofflight.Infact,Iamsureitisnosurprisethat over of flightwhileCFITisatypeaccidentthatcanhappenduring prevention effort.Asareminder,approachandlandingis phase CFIT accidentsin1998,arealshockandsetbacktothe CFIT come down.Thencame1998.Thereweresevencommercial jet Figure 3 Figure 2 ISASI 2003 Proceedings • 97 ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS 98 facts. just afew.Itreplacedlotof qualitativeideaswithquantitative and bytheNetherlandsTransportationSafetyBoard, to name entire sectionintheKAL/Guam accidentreport),byTSBCanada, report hasbeenusedasareference byNTSB(whoreprintedan page reporthavebeendownloaded fromtheFSFwebsite.The book onCFITandALA.More than48,000copiesofthis278- in Aviation”wasreleased.Thisreporthasbecomethereference In Decemberof1998,theALARTaskForcereporttitled“Killers ing overrun,runwayexcursion,andnon-stabilizedapproaches. study wereCFIT(includinglandingshort),lossofcontrol, land- weight over12,500pounds.ThetopfiveALAtypesfrom this included alljetandturbopropaircraftwithamaximumtakeoff landing accidentsthatoccurredbetween1980and1996. These work. TheALARstudywasbasedon287fatalapproach and Task Forcewasalreadyunderway.ItincludedmuchoftheCFIT program, theApproachandLandingAccidentReduction(ALAR) ability, andthustheriskofaCFITaccident. severity ofaCFITaccident,butTAWSgreatlyreducestheprob- risk equalsprobabilitytimesseverity.Wecan’tdomuchaboutthe mits thatonewillhappenatsomepoint.However,remember one—even DonBateman,theinventorofGPWSandTAWS,ad- craft equippedwithTAWS.Thatdoesn’tmeantherewon’tbe that’s thenumberofCFITaccidentsthathavehappenedtoair- of theregionalaircraft(10-30seats)haveTAWS.And0—well, with terrainawarenesswarningsystems(TAWS).Only4percent five percentofthecommercialjetsinworldareequipped some numberstothinkabout65percent,4and0.Sixty- one ofeverythreeturbopropaccidentswereCFITs.Hereare in additiontofourof15commercialjetaccidentsbeingCFITs, ternational widebodies—isimmunetoCFITaccidents.In2002, that nosegmentofaviation—fromturbopropcommuterstoin- four CFITaccidentsinthefirstmonthof2003.Thedatatellus from July27totheendof2002.Thebadnewsisthattherewere dent amonthfor8months!!Thegoodnewsistherewerenone jet CFITaccidentsin8months.Thatisalmostoneacci- From Nov.25,2001,toJuly27,2002,thereweresixcommercial about average,butthetableinFigure4showsitisfarfromit. cial jetCFITaccidentsin2001andfour2002.Thismayseem last 10years.AsshowninFigure3,therewereonlytwocommer- Figure 4 As weevaluatedthelackofsuccessourCFITprevention • ISA SI 2003 P r oceedings regional implementation plantheCAAGproposed wasthecre- implementation effortonaglobal basis.Oneofthekeysto The firstgoaltheCAAGset wastoconductaregionalALAR ence tomoreeffectivelyimplement theALARrecommendations. CAAG wastoutilizethelessons learnedfromtheCFITexperi- was knownastheCFITand ALAActionGroup(CAAG).The way toimplementtheinterventions oftheALAReffort.Thisgroup In 2000anewgroupwasformedandtaskedwithfinding a cal datasourcesavailableontheInternetasofJanuary2001. • they couldhavebeenavoided. sentation ofseveralCFITaccidentsandstrategiesbywhich • and landingaccidentsstrategiesforavoidingthem. You: • learned. tation ofapproachandlandingaccidentdatalessons • struction-protection criteria. cident data,lessonslearned,andareviewofapproachob- • equipment. Speaker’snotesincluded. tion ofequipmentandmethodsforoptimumuseexisting • operating environments.Speaker’snotesincluded. controller communicationandunderstandingofeachother’s • lots. Speaker’snotesincluded. cedures, andrecommendationsforaircraftoperatorspi- • Figure 5 • terrain awarenessandwarningsystem(TAWS). ing fromagroundproximitywarningsystem(GPWS)or • ing, standardoperatingprocedures,andequipment. to helpchiefpilots,lineanddispatchersevaluatetrain- • ing controlledflightintoterrain(CFIT). proach andlandingaccidents(ALAs),includingthoseinvolv- of 34documentsonavarietytopicstohelppreventap- • ing pilotawarenessoffactorsthatcontributetoCFIT. for assessingriskinspecificflightoperationsandincreas- • lustrate lessonslearnedaboutapproachandlandingaccidents. • training procedures. elements forcompany-standardoperatingproceduresand • approach. awareness ofhazards,includeselementsastabilized to thenormalapproachbriefingforincreasingflightcrew • ing onapproachandlandingaccidentsCFIT. which theyarebased.SelectedFSFpublicationsrelatedread- the FSFALARTaskForce,andaccident/incidentdataon into terrainaccidents”conclusionsandrecommendationsof Presents factsaboutapproachandlandingcontrolledflight Air T Flight SafetyDiges S Flight SafetyDiges Links toA CFIT A An Appr Appr P Equipment forAir Flight OperationsandT CFIT Alert: CFIT Checklist: Business &Commer Appr Appr tandard OperatingProceduresTemplate: ilot GuidetoP A 19-minutevideopresentationofthecausesapproach oach andlandingA oach andlandingRiskA oach andlandingRiskReductionGuide: raffic Contr war oach andLandingA viation StatisticsontheInternet: eness andP Procedure forimmediateresponsetoawarn- r Guidelines inseveraldifferentlanguages eventing CFIT ol: craft andAirT t: “KillersinAviation:FSFTaskForce”: t: “ALARBriefingNotes”: A presentationaboutimprovingpilot- cial A raining: r evention: ccident DataOverview: viation war ccident: ItCouldHappento : A presentationofdata,pro- A presentationofCFITac- raffic Contr eness T Posters: A 32-minutevideopre- ool: Aviation statisti- Four postersil- Recommended ol: A supplement A collection A presenta- Guidelines A presen- elements ofthe ToolKitisthebriefing notes.Thereare34 of to helppreventCFITandALA accidents.Oneoftheprimary uct. TheToolKitcontains19differentelements, eachdesigned CFIT efforttoproducethisself-contained andready-to-useprod- aviation experts.TheCAAG utilizedlessonslearnedfromthe sions, andrecommendations of9yearsworkbyalmost300 Kit. ThisCDconsolidatesthe data,products,findings,conclu- with afocused,userfriendlyproduct. implement theALARinterventionsgloballyonaregional basis regulators, andairports.ThesegoalswouldenabletheCAAG to dress commercial,cargo,andcorporateoperatorsaswell ATC, tions oftheALAReffort.TheCAAGwantedtoolkit to ad- friendly toolkitthatwouldassistinimplementingtheinterven- their region.ThesecondgoaloftheCAAGwastocreate a user who togoto,andwhathadbedonemaketheplanwork for the ALARinterventionsfortheirregion.TheRTLswould know the region.WewantedRTLstorunimplementationof aviation community,andhadcontactscreditabilitywithin the predominantlanguageofregion,wasactiveinregion’s to beanindividualororganizationthatwasanativespeakerof ation ofthepositionRegionalTeamLeader(RTL).Thiswas Figure 7 Figure 6 The culminationoftheCFIT/ALAReffortisALART ool of thesekillers inaviation. recommendations onaglobal leveltohopefullyreducetherisk our effortstodisseminate,educate, andcommunicatetheALAR of go-arounds,etc.Thisclearly showsthatwemustcontinue weather, non-radarenvironment, unstabilizedapproaches,lack scope oftheALARToolKit—non-precisionapproaches, causes ofanytheseaccidents werenew,oroutsidethe you haveseen,CFITcamebackwithavengeance.None of the have notimproved.In2001theALArateincreased.2002, as one outtherelistening?OnelookatFigure7showsthat things However, hereisaquestionforallsafetyprofessionals—Is any- these impressiveendorsements,thingsshouldlookpromising. shops complete,morethan28,500toolkitsincirculation, and the ToolKitandencourageditsuse.Sowith11ALARwork- programs.…” IATAandtheU.S.CASTteamhavealsoendorsed accident Tool Kithasbeenassessedascontainingextremelyvaluable cursion. TheICAO33rdAssemblyreportedthat“theALAR that canbeusedforothersafetyinitiativessuchasrunwayin- ALAR messageout,butalsoinestablishingasafetynetwork shops. Thesehavebeenverysuccessfulinnotonlygettingthe most efficiently. program toensurethatimplementationintheregionisdone strength oftheregionalapproach.Eachregioncantailorits approach andplanningforimplementation,butthatisthe course, weknowandexpecteachregionwillbedifferentinits annual requirementtorenewyourpilotlicenseinMexico.Of addition, a4-hourALARcoursebasedontheToolKitisan America havereceivedALARtrainingfromtheToolKit.In More than15,000pilotsand300ATCcontrollersinLatin tion SafetyTeam(PAAST),whichhasdoneanimpressivejob. America. TheRegionalTeamLeaderisthePanAmericanAvia- The firstregionaleffortwasstartedinDecember2000Latin in assistingtheRTLsimplementingALARinterventions. IATA, IFALPA,ALPA,andIFATCAhaveallprovidedsupport Kit oranyotherrequirementstheymayhave.Inaddition,ICAO, gion. TheCAAGsupportstheRTLswithworkshopsonTool corporate operators,ATC,regulators,andairports)inthere- Tool Kitgetstoeveryaviationorganization(commercial,cargo, plan. ThejoboftheRTListoensureinformationin the RegionalTeamLeaderandregionalimplementation plete listingoftheToolKitelements. tion” publicationcontainedinthetoolkit.Figure5givesacom- ALAR posters,anSOPtemplate,andtheentire“KillersinAvia- PowerPoint briefingswithspeakersnotesincluded.Therearealso and avideo).AllthebriefingscontainedinToolKitare information (severalbriefings,theALARriskassessmenttool, (the CFITchecklist,abrief,andvideo)ALA the briefingnotes,ToolKitalsocontainsCFITinformation 2,500 pagesofreferencematerialintheToolKit.Inadditionto ence yougettheentirereferencedocument.Therearemorethan references arenotonlylistedbyname,butselectingarefer- support it,adiscussionsection,summary,andreferences.The ment duringapproach.Eachbriefingnotehasstatisticaldatato tractions, beingpreparedforago-around,andenergymanage- briefing notetopicsincludeSOPs,managinginterruptions/dis- these 3-7pagedocuments,eachonaspecifictopic.Somesample Figure 6showsthecompletedandfutureALARregionalwork- Of coursethekeytogettingfullbenefitfromToolKitis

prevention materialwhichwillgreatlyassistaccident N ISASI 2003 Proceedings • 99 ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS • Stage 1 detailed below: The researchprojectwasdividedintoseveraldistinctstages as Research principles conclusions astheresearchdrawstoaclose. contained withinthispapermaybesupplementedbyfurther this paperdetailstheprogressmadetodate.Theconclusions privacy. additional informationjustifythepotentialinvasionofflightcrew mation isprovidedandwhetherthebenefitsassociatedwith by FDRsandCVRsdeterminewhat,ifany,additionalinfor- provided byflightdeckimagerecordingagainstthedata invasion oftheirprivacy. that flightdeckimagerecordingwouldconstituteasignificant that flightdeckimagerecordersmayprovideusefulinformation. to them.Asaresultofthisaccident,investigatorshavepostulated able toprovethisfromtheflightrecorderdatacurrentlyavailable series ofhuman-factors-relatedincidents,buttheyhavebeenun- tigation agencieshavesuspectedthatthecausesstemmedfroma There havebeenanumberofaccidentswheretheaccidentinves- Background/introduction (CAST) andtheJAAFlightRecorderStudyGroup. She iscurrentlySecretaryoftheCertificationAuthoritiesSoftwareTeam including theBoeing737and767AirbusA330/340A380. systems specialist.SheisnowamemberofmultinationalJAAteams 100 —whether there wereanyhealthandsafety issuesrelatedtothe dent typestoavoidskewingthe resultsofthetrial. —whether thescenarioscovered abroadenoughrangeofacci- • Stage 2 • protect flightcrewprivacy. —whether itispossibletousecameralocationsandangles to —appropriate cameralocationsforthetrial. representative todetermine This researchprojecthasnotyetbeencompleted;however, The purposeofthisresearchprojectistocomparethedata In contrastwiththis,flightcrewshaveexpressedtheconcern Static simulatorworkwithcameramanufactur Scenario evaluationmeeting to determine Draft aseriesofr • ISA SI 2003 Flightdeck ImageRecordingon P aircraft certificationprojectsasaCAAandJAA Surveyor withtheCAAfor7years,workingoncivil fields. ShehasbeenanAvionicSystemsDesign flight controlcomputersinboththemilitaryandcivil organization, sheworkedonarangeofsafety-critical prior tojoiningtheUKCAA.Whilewiththis Pippa Moore r oceedings epr esentative accidentscenarios. worked forGEC-MarconiAvionics Commercial Aircraft er andBALP By PippaMoore,CAA,UK A • Stage 4 age dataforanalysisbyindependentinvestigationagencies. • Stage 3 seen difficulties. —whether theproposedscenariosposedanyadditionalunfore- planned scenarios. —whether theproposedcameralayoutwasappropriatefor the simulatorequipmentavailable. —whether iswaspossibletoreplicatetheaccidentscenarioswith planned scenarios. • like thisarenotpossible. the aircraftbecomingdetached andsoscenariosincludingevents • flightdeck imagerecordercannot beestablishedduringthistrial. tions and,therefore,theeffects ofchangesinambientlightona • mencing thisresearch: The followinglimitationsofthistrialwereidentifiedpriorto com- Trial limitationsknownatthestart —establish thelimitationsoftrial’soutput. nical andpoliticalterms. —establish thelimitationsofproposedsysteminboth tech- cident investigation. protecting pilotprivacywithoutcompromisingthebenefit toac- —establish whethercameralocationcanbeusedasameans of from theimagerecordingsystem. —determine exactlyhowmuchadditionalinformationisgained information wasgainedfromeachandhowitgained. —present theanalysesofflightrecordersandestablishwhat • Stage 5 supporting data(e.g.,radarplots,etc.)tosupporttheiranalysis. Note: Bothinvestigationagencieswereprovidedwiththeusual a finalreport. report andBEAanalyzedtheFDRCVRdatasubmitted —Part 3:BFUanalyzedtheimagedataandsubmittedafinal were swapped. —Part 2:Onreceiptofthetwointerimreports,datapacks and submittedaninterimreport. an interimreportandBEAanalysedtheimagerecorderdata —Part 1:BFUanalyzedtheFDRandCVRdatasubmitted parts asfollows: R Fly theagr The simulator pr The simulatorpr Simulators ar W eplay oftheflightr ash upmeetingto eed scenariostopr e notcapableofpr ovided cannotaccuratelyr ovided isnot equipped withaCVR ecor ders. Thisoccur ovide voiceparameterandim- oviding genuinelightcondi- r ed inthr epr esent partsof ee distinct . This following progresshasbeenmade todate. This researchprojecthasnot yetbeencompleted;however,the Progress todate of the“flighttrials.” sible toavoidrecordinganexplicitlyidentifyingimage. crew weretolookdirectlytowardthecamera,itwouldbeimpos- flight deck.Itwas,however,acceptedbyallpartiesthatifthe avoid identifyingimagesofthecrewastheymovedaround the crew. era, withoutprovidingexplicitlyidentifyingimagesofthe flight that whichwouldbeproducedbyasingle,centrallylocated cam- viding aviewofthegeneralflightdeckareathatwasequivalent to cise couldthen“stitch”thetwosetsofimagedatatogether, pro- deck aspossible. position) whileprovidingasgreatageneralviewoftheflight member theyweredirectedtowards(whileinthenormalseated their viewingangleexcludedtheheadandshouldersofcrew corner-located camerasinsteadofonecentrallylocatedcamera. area viewoftheflightdeckcouldequallybegeneratedusingtwo acceptability asasolution.Thisresearchproposedthatgeneral explicitly identifyingimagesoftheflightcrew,thuslimitingits general viewoftheflightdeck,ithasalsobeenshowntoprovide flight deck.Althoughthismethodhasbeenshowntoprovidea using asinglecamera,probablylocatedtowardtherearof sumed thatageneralflightdeckareaviewwouldberecorded normal flyingpositions. the headandshouldersofflightcrewwhenseatedintheir where practicable,bedevelopedtoavoidrecordingimagesof unnecessary invasionsofprivacyimagerecordingsystemsshould, trying todistorttheexplicitlyidentifyingimagesofcrew. accidentally distortingimagesoftheflightdeckinstrumentswhile ology couldnotbeused,asitwastechnicallyfeasibletoavoid the problem. be useful,itwasconcludedthatnotacompletesolutionto tortion. have beendiscussed,includingdataencryptionandimagedis- general flightdeckareaviews.Varioussolutionstothisproblem around flightdeckimagerecording,particularlywithrespectto Intrusion ofpilotprivacyhasbeenamajorissueinthedebates Protection offlightcrewprivacyusingcameralocation of theFDRdatawouldnotbeentirelyrealistic. subject tothelimitationsofsimulator.Thisimpliedthatsome FDR datainthattheyweresampledatahigherrateand • come ofthetrial. anticipated thatthiswouldhaveasignificanteffectontheout- may resultinanon-standardequipmentinstallation.Itwasnot meant thataCVRwouldhavetobetemporarilyinstalled,which The accidentscenarios havebeendraftedand flownandboth It wasagreedthatthisconcept would betestedduringthecourse This dual-cameralocationwasbelievedtomakeiteasier to Once theimageshadbeenrecorded,apost-processingexer- It wasfurtherproposedthatthecamerasbelocatedsuch Previous discussionsonflightdeckimagerecordinghaveas- EUROCAE WorkingGroup50agreedthatinordertoprotect Discussions onimagedistortionconcludedthatthismethod- Although itwasagreedthatsomeformofdataencryptionmay The FDRdatawer e notcompletelyr epr esentative ofnormal • • • • • terrorist). • provide imagesofthefollowingtypesincident: ments werecarriedouttodeterminewhetherthecameras would missed bythiscombinationofcameraviewssoaseriesexperi- cerned thattheremaybesomeaccidentscenarioscould be and therequirementsofEUROCAEED-112. dressed boththeconcernsraisedbyvariouspilotassociations crew whiletheywereseatedintheirnormalpositions.Thisad- cameras excludedimagesoftheheadandshouldersflight each other.Itwasfurthershownthatthiscombinationoffour area viewoftheflightdeckoncetheirimageswereviewednextto adequate coverageoftheflightdeckinstrumentsandageneral the staticsimulatortrialrevealedthatfourcamerasprovided ner-located cameras.However,extensiveexperimentingduring eral areaviewoftheflightdeckcouldbeobtainedusingtwocor- The initialproposalputforwardbythisresearchwasthatagen- Protection offlightcrewprivacyusingcameralocation initial conclusionshavebeendrawn.Thesearediscussedbelow. Although theresearchprojecthasnotyetbeencompleted,some Initial researchfindings about theinstallationofimagerecorders. would berequiredbeforeanyfinalconclusionscoulddrawn ing thisequipment.Itis,however,notedthatacost-benefitanalysis investigators andsodoesnotaddressthepotentialcostofinstall- recording systemcanprovideadditionalinformationtoaccident Note: Thisresearchislimitedtodeterminingwhetheranimage • image recordingdata. to avoidmistakenconclusionsbasedoninsufficientanalysisof • lar camerasetupinrealaircraft. • • conclusions canbereached: The followingtasksneedtobecompletedbeforeafinalsetof Work tobecompleted complete. further conclusionsmayneedtobeaddedoncetheresearchis the research.Althoughtheseareunlikelytochangesignificantly, it alsohighlightedtheneedtodiscusssomeofissuesfurther. however, althoughthatmeetingdidproducesomeusefulresults, BEA andBFUhaveevaluatedallthedata. • because thecameraanglewas suchthat sion), theimagerecorderprovided aclearview.Thiswaslargely Despite this,theteamworkinginstaticsimulatorwascon- This beingso,thispaperrepresentsthecurrentconclusionsof An initial“washup”meetingwasheldtodiscusstheresults; In allcasesofaggression(including inter-flightcrewaggres- Physical incapacitationofthe flightcr Inter-flight cr Mor Non-flightcr Flightcr Aggr Completion offinal,agr Analysis ofwhatspecificconsiderationsshouldbeaddr Investigation ofwhetheritwouldbepracticaltoinstallasimi- Analysis oftheissuesraisedininitialwash-upmeeting. anyone attacking theflightcr e thanonepersoninanyflightdeck seat. essive intrusionintheflightdeck(e.g.,fr ew memberschangingseats. ew membersbeinginvitedtoflytheair ew aggr ession. eed r esear ew fr om behindwould beseen. ch paper ew ISASI 2003 . . om apassenger/ Proceedings craft. essed • 101

ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS • workload. human factorsissues,includingsomethatresultfromhigh could provideinvaluabledatarelatingtoarangeofflightcrew It hasbeensuggestedthatflightdeckimagerecordingsystems Human factorsandflightcrewworkload vide informationonanyotherformofflightcrewincapacitation. the flightcrewasimagerecordersareunlikelytobeablepro- Note: Thisassessmentwaslimitedtophysicalincapacitationof tion ofamembertheflightcrew. utes) wouldprovideevidencetosuggestthephysicalincapacita- absence ofmovementforaprolongedperiod(e.g.,severalmin- are notphysicallyincapacitated.Theinferenceofthisisthatthe hands overtheinstruments,thusprovidingevidencethatthey normal courseofflying,theflightcrewareseentomovetheir mation aboutflightcrewincapacitation.However,duringthe in theirnormalseatedpositions,itdidnotprovidedirectinfor- were designedtoavoididentifyingimagesoftheflightcrewwhile seated positionandintotheviewofinstrumentpanel. flight crewbecausetheyhadfallenforwardfromtheirnormal ing backwardintheseat. • ing forwardontothestick. • quired theassessmentoftwodifferentissues: information aboutphysicalincapacitationoftheflightcrewre- would bevisible. should theseatbeoccupiedbymorethanonepersonthattoo ing seatsintheflightdeck,andcameraanglewassuchthat and theirvictimthatitcouldbepickedupbyanimagerecorder. ciently differentpatternofbehavioronbehalftheaggressor weapon wouldbeseenortheactionitselfresultinasuffi- • 102 • • flight crewbycabinorpassengers.Thisisduetotwofactors: to providemeaningfulinformationrelatingdistraction ofthe sive intrusiondiscussedabove,flightdeckcameraswillnotbe able factors have,sofar,beenmixed. image recorderstoprovideinformationaboutflightcrewhuman The resultsoftheanalysisrelatingtoabilityflightdeck Flight crewhumanfactors of soundmeans thattheywouldbeunableto determinethescale system couldshowwasthatthey hadbeendisturbed.Theabsence ness, andsupposingtheydid lookback,allanimagerecording trained tokeeplookingforward toavoidlosingsituationalaware- • tion wouldnotbeprovidedby animagerecordingsystem. image recordersystemsprovide imagesnotsound;thisinforma- interruptions areusuallyviasomeformofinterphonesystem. As The analysisofthisissuecanbesplitintotwoareas The secondcasewasnotsoclear-cut.Becausethecameraangles In thefirstinstance,camerasprovidedaclearimageof The assessmentofwhetheranimagerecordercouldpickup The imagerecorderalsoprovidedclearviewsofanyonechang- It istruetosaythat,withtheexceptionoftypesaggres- Flight cr A memberoftheflightcr A memberoftheflightcr if amemberoftheflightcr Flight cr Flightdeck doorsar Even ifthecabincr • ISA SI 2003 ew workload ew humanfactors P r oceedings ew didentertheflightdeck, cr e nowlock ew becomingincapacitatedandfall- ew becomingincapacitatedandfall- ew wer ed, whichmeansthatcabincr e towieldaweapon,eitherthe ews ar ew e highly trained that theyfrequentlyactinunison withouttheneed by accidentinvestigatorsisthat commercialflightcrewsareso and aircraftenvironmentalnoise. Oneofthemajorissuesfaced did notoccur. not done).Theycannotprovide informationaboutactionsthat can onlyrecordsystemstatus (i.e.,whatwasdoneand use ofimagerecordersthathasnotbeendiscussedbefore. about unsuccessfulattemptstoresolveproblems.Thisisapotential It wasfoundthatimagerecordingsystemscanprovideinformation Unsuccessful flightcrewactions tion oftheimagesprovided. image recordingsystemsinthisareawilldependontheresolu- to theflightcrewhasbeensmoothed.However,effectivity of may providesomeinformationastowhetherthedatadisplayed the numberofincident/accidentsattributedtopiloterror. flightdeck imagerecordingsystemscouldresultinareduction missing dataandfailflags.Thisimpliesthattheuseof images ofsufficientresolutiontoenableinvestigatorsseeboth has alsobeenshownthattheimagerecordersystemsprovide vide clearevidenceofthefailureaircraftelectronicdisplays.It different informationtothatwhichisrecordedontheFDR. cess of“smoothing”mayalsoresultintheflightcrewbeinggiven electronic displaysareprovidedwithsmootheddata.Thispro- worth notingthatFDRsystemsgetraw,unsmootheddatawhile and, consequently,maynotreacttheexpectedmanner.Itisalso never receivetheinformationthathasbeenrecordedonFDR, cal failure,etc.)orthedisplaysthemselvesfail,flightcrewwill should belostatapointaftertheFDR(e.g.,throughfire/electri- tronic displays,theFDRusuallygetsdatafirst.If flight crewdisplaydatafromthesamesourceasrelevantelec- responses inarangeofinvestigationreports. been postulatedaspossiblecauseforinappropriateflightcrew FDR maynotactuallybedisplayedtotheflightcrew.Thishas an electronicdisplay,informationthathasbeenrecordedon It hasbeensuggestedthat,whereaircraftdataaresuppliedvia Loss offlightcrewdisplays • • cording system.Theseare of informationthatareunlikelytobeprovidedbyanyotherre- shown thatimagerecorderscanprovidetwoveryimportantpieces vide usefulinformationaboutdistractionoftheflightcrew,ithas reason fortheirreactioncanonlybedeterminedfromaCVR. they turnedtofacethedisturbanceorconsultedachecklist), tion showingthecrew’sreactiontodisturbance(e.g.,whether Although imagerecordingsystemswouldprovidesomeinforma- tion relatingtoflightcrewdistractioncomesfromCVRsystems. rear galleyandthepassengersarepanicking”). cup ofcoffeesir”and“Captain,thereisanuncontainedfireinthe of theinterruption(e.g.,differencebetween“Wouldyoulikea CVRs canonlyprovideinformation onflightcrewdiscussions FDRs andCVRscanonlyprovidecertaintypesofdata. This researchhasalsoshownthatimagerecordingsystems This researchhasshownthatimagerecordersystemscanpro- The reasonforthisisthatalthoughFDRsystemsgettheir Although imagerecordersystemshavenotbeenshowntopro- This researchclearlyshowsthatthemostmeaningfulinforma- unsuccessful flightcr loss offlightcr ew displaysthatisnotdetectedbytheFDR ew actions. . • • lating toflightcrewworkloadneedsbeassessedinthreeways: The abilityofimagerecordersystemstoprovideinformationre- Flight crewworkload dents/incidents thatareattributedtopiloterror. This, again,couldresultinareductionofthenumberacci- providing informationaboutfailedattemptstosolveproblems. to taketheappropriateactionbutwereunabledoso. sored fornottakingappropriateactionwhen,infact,theytried to supplementthelackofFDRinformation. environmental noise,theCVRwillnotprovideanyinformation solve aproblemandtheattemptdoesnotresultinchange tually sayinganything. visual communication(e.g.,lookingorpointing)ratherthanac- sion. Iftheirworkloadisparticularlyhightheymaywellresortto to solveaproblemandfaildosowithoutanyaudiblediscus- crew didnotdiscussitatall. the imagerecordershowedthatthishadhappened,flight first officer’sdisplaysfailedduringoneofthescenarios.Although workload isveryhigh.Aclearexampleofthisoccurredwhenthe for discussion.Thiscanbeparticularlytruewhentheflightcrew • tal, problem-solvingactivities. tions performedbytheflightcrew,andcognitivetasksaremen- tive tasksandcognitivetasks.Manipulativearephysicalac- In generalterms,flightcrewshavetwotypesofwork,manipula- Cognitive versusmanipulativetasks • they arein,system malfunctions,etc). communication difficulties,simple concernaboutthesituation of thestress(i.e.,crewcould bestressedduetoworkload, der stressfromtheCVR,itisnot possibletodeterminethecause Although itispossibletodetermine thattheflightcrewareun- Flight crewresponsetoworkload/stress determine fromanyflightrecorderdata. than usual.Thisissomethingthatwouldbealmostimpossible to the crewarelessabletocombinecognitiveandmanipulative tasks and theremaybesituationswhere,duetofatigueordistraction, about cognitiveworkload. workload, itwouldbedifficulttomakeanypositivestatements or incident. • very difficult. complex cognitivetasks,maymakesimplemanipulativetasks • for aflightcrew. even asmallnumberofthesimplestcognitivetasksverydifficult nipulative tasksthatrequiregreatstrengthordexterity,maymake Image recordingsystemshavebeenshowntobecapableof This absenceofdatacouldresultintheflightcrewbeingcen- If thereisnoaudiblediscussionrelatingtoafailedattempt This researchshowedthatflightcrewscansometimesattempt In simplisticterms,thefollowingstatementsaretrue: It shouldalsobenotedthatthereare“gradients”ofworkload, Although imagedatacouldprovideevidenceofmanipulative Flight cr Cognitive versusmanipulativeworkloads. An e Accident investigatorassessmentoftheflightcr Either oneofthesecombinationscouldr An e x x cessive numberofcognitivetasks,orasetparticularly cessive numberofmanipulativetasks,orasetma- ew r esponse toworkload/str ess. esult inanaccident ew r eaction. any smokethat isvisibletotheflightcrew. recording willenableinvestigators todiscoverthepresenceof son forthisaction.Thisbeing so,theCAAbelievesthatimage will lookbackthroughtheimage recordingtodeterminetherea- smoke hoods,anditisreasonable topresumethatinvestigators crew seessmokeintheflight deck,theirreactionwillbetodon deck (particularlyifthesmoke isnotdense);however,ifaflight not beimmediatelyapparentthatthereissmokeinthe flight image recordercameraswouldalsobeabletoseesmoke.It may vide theabilitytodetectsmokeinflightdeck. It hasbeensuggestedthatflightdeckimagerecordingwould pro- Smoke intheflightdeck accurate andusefulinformationrelatingtoflightcrewworkload. CVR oranimagerecordingsystemcanbeguaranteedtoprovide vestigator perception.Giventhisfact,itisunlikelythateither a two othersetsofhumanfactors;flightcrewpersonalityandin- workload, andtheassociatedhumanfactorsissues,dependsupon data aretobeusefulanaccidentinvestigation. that carefulproceduresneedtobedevelopedifimagerecorder to detectotherindicators.ThishasledtheCAAconclusion vestigation mayresultinananalysisthatisnotofsufficientdepth crew arecoping,thehumanfactorsassociatedwithaccidentin- either CVRorimagedatahasledtotheperceptionthatflight form averydetailedanalysis.Ifpreliminaryinvestigationof ever, toobtainthisinformationinvestigatorswouldneedper- dications andbreathingratesequatingtophysicalstress);how- substantiating evidenceofworkloadorstress(e.g.,heartratein- reported asacontributoryfactor. sound orlookasiftheyareunderpressure,thisisunlikelytobe if aninvestigatordoesnotgettheperceptionthatflightcrew data. workload/stress dependsuponhowaninvestigatorperceivesthe image recordingsystemtoprovideinformationaboutflightcrew This researchalsofoundthattheabilityofeitheraCVRoran Accident investigatorassessmentofflightcrewreaction input forcesthusprovidingfurtherindicationsofphysicalexertion. Note: TypeIaorIVaflightdatarecorderinstallationswillrecord determine whethertheylookorsoundstressed. low), thepersonalityofindividualflightcrewmemberswill “Accident InvestigatorAssessmentofFlightCrewReaction”be- stress. provide anyinformationthatindicatestheeffectofworkloador sions, obviousphysicalexertion,etc.),animagerecordercannot affecting theirabilitytomanagethesituation(e.g.,facialexpres- crew makenovisualresponsethatsuggeststheirworkloadis dent uponthepersonalitiesofflightcrewconcerned. recording systemtoprovidethisinformationisactuallydepen- However, thisresearchhasshownthattheabilityofanimage ther informationaboutthepossiblecausesofflightcrewstress. This researchdeterminedthatiftheflightcrewcouldseesmoke, This impliesthattheaccurateassessmentofflightcrew It istruethatcarefulanalysisofCVRdatamayprovidefurther Even ifthereisevidencethatsuggestshighworkloadsorstress, Although excessivephysicalexertionshouldbeapparent(see Image recorderscanonlyrecordvisualdataandiftheflight It hasbeensuggestedthattheimagedatacouldprovidefur- ISASI 2003 Proceedings • 103

ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS • include that applytotheinvestigationofimagerecorderdata.These This researchhasidentifiedarangeofadditionalconsiderations Additional considerations completely representativeoftheflightcrew’sview. event ofsmoke,animagerecorder’sviewtheflightdeckisnot than theimagerecordingsystemcameras. and, assuch,haveabetterchanceofseeingwhatishappening to theinstrumentpanelthanimagerecordersystemcameras this additionalimpedanceto“vision.” scure theirvision.Imagerecordersystemcameraswillnothave the worstphysiologicaleffectsofsmoke,theywillpartlyob- oxygen masks.Althoughthemaskswillprotectthemfrom smoke intheflightdeck. visibility theflightcrewhasoftheirenvironmentwhenthereis image recordersystemtoprovidearepresentativeideaofthe fumes intheflightdeck. they couldbetheonlysourceofinformationaboutsmokeor donning smokehoodsand,intheabsenceofanfunctionalCVR, tems couldprovideadditionalinformationaboutthereasonfor the CAAconsidersthat,atveryleast,imagerecordingsys- this mayimplythattheuseofanimagerecorderissuperfluous, cation systems,whichcanreadilybedetectedbyaCVR.Although ning smokehoods,itwouldalsoresultinachangeofcommuni- drawing anyconclusions. and toreferotherFDRdatarelatingaltitude,etc.,before subject theimagedatatoanappropriatelevelofinvestigation provide definitiveevidenceofsmoke,andinvestigatorsneedto at highaltitude).Thismeansthatcrewactioncannotbetakento oxygen maskscouldequallyimplyadropincabinairpressure(if smell smokewithoutactuallybeingabletoseesmoke.Theuseof 104 • gators intheir appraisal offlightcrewreactions. However,accu- port ofanappropriatelytype-rated pilotwouldbebeneficial. prior toperformingtheinvestigation. Itisalsonotedthatthesup- age recordersshouldrequire thatacquisitionofthisinformation believes thatanyproceduresrelating totheinterpretationofim- cluding OEMtechnicalmanuals andtheInternet,CAA actions andflightdeckalerts. enable accidentinvestigatorstosuccessfullyinterpretflight crew the flightcrewpersonalities. detailed knowledgeoftheflightdecklayoutand of not berequiredtoanalyzeCVRorFDRdata.Theseinclude a Image dataanalysisrequiressomeitemsofinformationthat would Additional information analysis. These inversefactsleadtotheinevitableconclusionthatin Inversely, itisalsotruetostatethattheflightcrewwillbecloser If thereissmokeintheflightdeck,crewwilldon One furtherpointthatshouldbeaddressedistheabilityofan As thepresenceofsmokeshouldresultinflightcrewdon- A flightcrewmaywelldonoxygenmasksbecausetheycan There are,however,somecaveatstothis. Knowledge oftheflightcrew personalities wouldassistinvesti- This informationcanbeobtainedfromarangeofsources in- A detailedknowledgeoftheflightdecklayoutisessential to additional informationr the needforcar • ISA SI 2003 P r oceedings eful pr ocedur equir ed forimagedataanalysis. es r elated toimager ecor der GECAT Training Facilities AD Aerospace(forsupplying,installing, andprovidingtechnicalsupport BFU Germany BEA France UK AAIB BALPA The CAAwouldliketoextendits grateful thankstothefollowingorganiza- Acknowledgements and thenafullreportwillbeissued. son forflightcrewactionsorreactions). corder systemscanprovidelittleinformationabout(e.g.,the rea- by otherrecorders(e.g.,smokeintheflightdeck). displays andfailedflightcrewattemptstosolveproblems). available fromotherflightrecordersystems(e.g.,lossof itly identifyingimagesoftheflightcreworpostprocessing. eral areaviewsoftheflightdeckwithoutneedforeitherexplic- can bemadeatthispoint. final conclusionscanbedrawn,however,thefollowingstatements As theCAAhasyettocompleteitsresearchintothissubject,no Interim conclusions incorrectly identifiedasthehandlingpilotduringtakeoff. • panel blanked18secondsaftertakeoff. whereas otherinformationdemonstratedthattheinstrument that thefirstofficer’sinstrumentpanelwasblankbeforetakeoff, • tual errorsintheanalysisofimagedata.Examplesthisinclude Note: the soleflightrecorderforanyaircraft. data, itwillconcludethatimagerecordersshouldnotbeusedas likely that,duetotheoccasionallyambiguousnatureofimage avoid mistakenconclusions. ticular careshouldbetakenwiththeanalysisofimagedatato of investigationintoothercausalfactors. misinterpretation oftheinformationprovidedandabsence makes itmorepronetoprecipitatejudgement.Thismaylead data canbemisinterpreted,thecompellingnatureofimage The CAAhasconcludedthatalthoughalltypesofflightrecorder recorder analysis The needforcarefulproceduresimage to provideausefulbalancepurelytechnicalinformation. However, someknowledgeofthepersonalitiesinvolvedisdeemed tion ofimagerecorderdataintermstheaccuracyresults. sence ofthisinformationwouldsignificantlyaffecttheinvestiga- dient (asdiscussedearlier),theCAAdoesnotbelievethatab- need toaccountforthecognitiveworkloadandgra- and sincecompleteassessmentoftheflightcrewreactionwould rate informationonthissubjectwouldbeverydifficulttoobtain, for theimagerecorder systemandcameras) tions fortheirassistanceinthisresearch: The CAAexpectstofinishtheirresearchbytheendofyear There arestill,however,sometypesofeventsthatimage re- They canalsoprovidevaluableconfirmationoffactssuggested Image recordingsystemscanprovideusefuldatathatarenot It ispossibletoinstallimagerecordingsystemsthatprovidegen- The CAAhasnotcompleteditsresearchintothisissuebutitis As aresultofthisconcern,theCAAhasconcludedthatpar- In anotherofthedynamicsimulatortrials,firstofficerwas In oneofthedynamicsimulatortrialsaninvestigatorconcluded Thisconcernisdemonstratedbythepresenceofseveralfac- N meeting two,this hadbroadenedtotheconsideration ofimage dress therequirementsfor recordingofdatalinkdata.By report. could meetthisneedandset upWorkingGroup50(WG50)to (EUROCAE) recognizedin1995 thatmodernvideotechnology strument text. the pilotactuallysees,ratherthandatathatdefines in- ing updatalinkdata,whileforthefirsttimeshowingjust what tion togettingextradataonthecockpitenvironment,and pick- Accident Investigatorshavelookedtovideorecordingasa solu- Image recording pilot incorrectinformationandleadtodisaster. translation ofelectronicdatatovisualpresentationcangive the instruments thattheyhavereplaced,andanyanomalies in the however, theirdisplaysareonlygraphicalinterpretationsofthe pits arenowtheprimaryflightdisplaysonwhichpilotsrely; flightdeck crewareunder,almostredundant.Modernglasscock- cockpit environment,crewinteraction,andthepressuresthat (CVR), whichgivestheaccidentinvestigatordetailsabout tools oftheairaccidentinvestigator,cockpitvoicerecorder moval ofthevoicelinkalsomakesonemostimportant and lesssubjecttomisinterpretationerror.However,there- communication betweenaircraftandthegroundmorereliable trol radiomessagesarebeingreplacedwithtextuplinks,hasmade The recentmovestowarddatalinksystems,whereairtrafficcon- Introduction and afullmemberoftheInstitutionElectricalEngineers. section ofthefinalspecification,ED-112.HeisaCharteredEngineer 50, andwassecretaryofthesub-groupthatwroteImageRecording 2002, hewasanactivememberoftheEUROCAEWorkingGroup ments, whichisnowbeingofferedtotheworld’sairlines.From1996 first “digitalvideonetworkserver”compatiblewithaerospacerequire- his timeatthecompany,Hornehasoverseendevelopmentof in thedesignandmanufactureofvideosystemsforaerospace.During later splitoffastheindependentcompany,ADAerospace,specializing company DedicatedMicrocomputerstoformDMAerospace,which systems. In1995hejoinedthesuccessfulvideosecuritysystems His workhasalsoincludedimageintensifyingandthermalimaging guidance andfirecontroltopipeinspectioncommercialsecurity. In 1996,EUROCAEWG50was formedwiththeremittoad- The EuropeanOrganization forCivilAviationEquipment systems forawidevarietyofpurposesfrommissile 1983. Hehasworkedextensivelyinvideocamera student apprenticeshipwithMarconiAvionicsin electrical andelectronicengineering,includinga University ofBradfordwithanhonorsdegreein Director ofADAerospaceLtd.Hegraduatedfromthe Mike Horne Flightdeck ImageRecording By MikeHorne,ManagingDirector,ADAerospace,Ltd.,ManchesteUK On CommercialAircraft , Beng,Ceng,MIEE,isManaging Purpose instruments andcontrols. Coverage areas a) Generalflightdeckarea which resultsinverydifferenttechnicalsolutions. ED-112 detailsfiveseparaterequirementsforimagingsystems, ED-112 RequirementsofanImagingSystem the systemstobeusedonboardaircraft. Finally, inMarch2003thefinalMOPSwaspublished,defining ED-112 MOPS datalink recordingsystem. formance Specificationforthegroundrecordingportionof In 2002,ED-111waspublishedasaminimumoperationalper- ED-111 MOPS of thepilotcommunity. cident investigationrecording,bearinginmindthesensitivities tigators discussinganddebatingthe“fundamentalneeds”ofac- detailed specifications.Thisworkstartedwiththeaccidentinves- mance characteristicofthesystem.Workcouldthenstartonmore specification, orMASPS,whichdefinedthe“endtoend”perfor- published. Thiswasaminimumaircraftsystemperformance Five yearsafterthestartofGroupfirstdocumentwas ED-92 MASPS ing theirnamesrecordedasauthorsofthefinaldocument. around 85peopleattendingmorethanthreemeetingsandhav- Working Groupmorethan150peoplesatonthepanel,with and KievUniversity.Overthe22meetings7yearsof tours ofairtrafficcontrolfacilitiesintheUK,ARINCfacilities, own premisesoratnearbyhotels.Themeetingsalsocovered meeting washostedbyamemberofthegroup,eitherattheir tawa, Annapolis,Sarasota,Memphis,Kiev,andVenice.Each and inEurope,asfarwideLondon,Paris,Brussels,Ot- were heldthreetimesperyear,alternativelyinNorthAmerica Canada, Sweden,andtheNetherlandsamongothers.Meetings group camefromtheUnitedStates,UK,France,Germany, tors, airlines,andavionicsmanufacturers.Themembersofthe a balanceofregulators,airlinepilotunions,accidentinvestiga- The membersofEUROCAEWG50arevoluntaryandcomefrom EUROCAE WG50 tigation recorderrequirements. recording andtoageneralupdateofthecurrentaccidentinves- • etc.). • General cr Ambient conditionsontheflight deck(smok —To determinethefollowing: ew activitiessuch asuseofchecklists,charts, etc., —All flightcrewstationswork areasincluding ISASI 2003 e, fir Proceedings e, lighting, • 105

ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS 106 • • and healthwell-beingofcrew. • the displays. • Resolution partial display,automaticdisplaymodechanges,etc.). • on theflightdatarecorder. • Purpose installed). center pedestal,andvideodisplayspresentedtothecrew(where Coverage areas b) Instrumentsandcontrolpanels Color actions. Frame recordingrate and ambientconditions. Resolution tion (switch/throttle/flightcontrols,etc.). the air,whileeconomically attractive,willresult inearlyproblems corders designedfortheoffice environmentandusingthemin that environment.Takingstandardoff-the-shelfcameras andre- need tobedesignedandmanufactured specificallyforusein To withstandtheharshaerospace environment,allcomponents General requirementsofan airbornecamerasystem at theflightdeckinstrumentpanel. ment isforahigh-resolutioncameraspecification,directed solely view, coveringtheflightdeckcrew.Theinstrumentpanelrequire- Color Frame recordingrate can onlybemetbyimagerecording. The generalflightdeckarearequirementisforawide-angle Cockpit selectionswithincr Non-verbal communications(handsignals,pointing,etc.). Determine parametervalueswhoser Determine instrumentdisplaystatusandoperationalmode of Status ofinstrumentdisplaysanddisplaymodes(blankscr Information (includingcr • ISA —Required. —Required. SI 2003 —To determinethefollowing: —Sufficient todeterminestatusofinstrumentdisplays —Sufficient to P r —Forward instrumentpanel,overhead oceedings — —As shownintableabove. S ufficient todeterminesignificantcrew ew selections)note ew r each whileseatedatdutysta- ecor ding r xplicitly r equir ecor ements een, ded the engineinstruments andautopilotcontrols. sole andthemiddlesection of theinstrumentpanel,covering specification. a “recognizableviewofthepilots,” aswasrequiredbytheED-112 In thiswayaclearviewofthe PFDs couldbeseen,withoutgiving displays (PFDs)ofthepiloton theoppositesideofaircraft. to viewacrosstheinstrumentpanelandontoprimary flight and behindthepilots’heads. which wasscrewedontotheoverheadinstrumentpanel,just above available foratrialofthiskind. highly specializedpieceofequipment,andiscertainlynotreadily lens toreplacethemultiplestandardcameras.Thiswould bea would betouseonevery-highresolutioncamerawithastabilized alternative tothis,whichwouldprobablybeadoptedinpractice, cameras couldbeused,witheither6mmor8lenses.The a systemoffourstandardresolution(600pixelhorizontal)CCTV instrument panelrecordingrequirementsofED-112setoutabove, angles weretriedout.Itwasdecidedthat,inordertomeetthe 112 specification. recording system,tobeasfarwaspracticalinlinewiththeED- (GCAT) 737-300simulatorsectioninGatwick. a trialthatitwassettingupattheGECCapitalAviationTraining In 2002,theCAAapproachedADAerospaceforassistancewith CAA trial pending onthephasedifferencebetweensignals. which canleadtoawholerangeofdifferentbeatingeffectsde- ting canbeavariablescanratelikethoseusedinradardisplays, but outofphase,withthecamerascanrate.Evenmoreoff-put- caused bythescanrateoftubeorLCDdisplaybeingsimilar, When usedtoviewinstruments,thiscanlead“beatingeffects” of resolution. into thepicture,whichleadstoadeterioratingandloss effects withinacameracanleadto“thermalnoise”beinginjected temperatures canrisetowellover50degreesCelsius.Heating a constrainedavionicsbayinanaircraftparkedthedesert, fore inevitablytoelectronicshuttering. to apurelyelectronicsolution,usingnomovingparts,andthere- will besubjecttoagreatdealofstressandwear,whichreallyleads parked onanaproncandropfarbelowzero.Anymovingparts Celsius, andjustasimportantlytheinteriorofanaircraftleft light controlsystem. either amechanicallydrivenirisortowide-rangingelectronic dimmed passengercabin.Forthecameradesigner,thisleadsto clouds, forinstance,beingupto100,000timesbrighterthanina even withinasinglepicture,withthebrightestscenes,above highly specializedvideosolution. They needtotakeintoaccountthefollowing,whichleadsa and reliableusingsolid-stateelectronicshutteredlightcontrol. and manyfailures.Specifically,camerasneedtobesmall,light, The twocentralcameraswere angledtocoverthecentralcon- Cameras fromtheextremerightandleftofbarwereangled The fourcamerasweresuspendedfromanaluminum bar, A “setup”daywasheldinAugust2002,whenequipmentand AD Aerospacewastodesignandsupplythevideocamera Standard CCTVvideocameras“scan”atafieldrateof60Hz. With theintenseradiationofsunduringaflight,orwithin The exteriorofanaircraftinflightcanbedownto–60degrees Within anaircraftenvironment,thelightrangeisverywide, the trial.Anactualinstallationwould,ofcourse,behardmounted; a) Weunderestimatedthevibrationlevelsthatoccurredduring Following thetrial,severallessonswerelearned. Conclusions specification. frame sizeof30kB.ThisismoreorlessinlinewiththeED-112 per second,withatargetresolutionof640x256,and The recordingratewassettobearoundfourframesperchannel compression torecordthevideoandaudioonaharddiskdrive. an ADHoldings“TransVu”digitalrecorder,usingmodifiedJPEG all viewoftheflightdeckenvironment. behind andabovethepilotposition. lines horizontalresolution.Camerasmountedonabar,just All CamerasWATEC201PALstandard,color,approx.350TV 5 3.5mmlenscockpitgeneralview 4 8mmlenscopilotmaindisplays 3 6mmlensthrottlesandcenterpanel 2 6mmlensengineinstruments 1 8mmlenspilotmaindisplays The cameras,plusasmallmicrophone,wererecordedusing One furtherwide-anglecamerawasinstalled,togiveanover- lead toerroneousconclusions. voice, andimagerecording,thatanysinglerecordingcould ture ofanaccidentorincidentcanonlybebuiltupusingdata, the FDRanalyses. image system.ThiswasthesameconclusionaswithCVRand sented, therewereseveralfactorsthatonlypickedupbythe and furthertrialsshouldnotberuledout. investigator. Thistrialshouldbeseenaspartofongoingwork, aged byED-112wouldbeausefuladditionaltooltotheaccident space inlate2002showedthattheimagerecordingsystemenvis- The trialscarriedoutbytheUnitedKingdomCAAandADAero- Summary ing initself,shouldnotbesurprising. the earlydaysofCVRandFDRthisfinding,whileinterest- certainly truethatasimilaramountoftrainingwasnecessaryin indicators, switchpositions,andreadingpointersongauges.Itis flightdeck, whilequalifiedpilotshadnoprobleminidentifying cate, ordistinguish,thereadingsofinstrumentson investigators. Inseveralinstancestheinvestigatorscouldnotlo- edge ofthecockpitlayoutthatwasrequiredfromaccident d) Oneofthemostinterestingconclusionswasdepthknowl- able indaylight)tobelost. signal tonoiseratiointhevideopicturecauseddata(read- able. Inothercases,lackoflighting,andthesubsequentloss ments, whichinsomecasesmadetheinstrumentsquiteunread- caused severalproblems,including“blooming”oflightedinstru- low lightlevelsimposedbythesimulatoroperatingconditions c) Whilewesetupthecamerasingoodlightingconditions, the interpretationofresults. there, theredidnotseemtobeanyproblemscausedbythisin scanned LCDandtubedisplays.Whiletheeffectswerecertainly b) Weexpectedtoseedistracting“beatingeffects”whenviewing of theresultingpictures. arrangement, whichledtosomedifficultiesintheinterpretation however, therewasnoticeablemovementinthe“aluminumbar” Throughout thistrial,itwasapparentthatthecompletepic- While notgivingthefullsolutiontoaccidentscenariospre- N ISASI 2003 Proceedings • 107

ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS T dent investigation process.IntheUnited States, theNational of accidentreportsaffordsthe opportunitytoimprovetheacci- affords theopportunitytoimprove designandoperations,analysis recommendations. Inthesame sensethatanalysisofanaccident recurrences throughthedevelopment ofviableandeffectivesafety The purposeofanaircraftaccident investigationistoprevent Introduction gation process. area forpotentialimprovementintheNTSB’saccidentinvesti- dent investigationprocess,theseresultsindicatethatthis isan ommendations. Sincethereportisanintegralpartof acci- exhibit strongcorrelationbetweentheirfindings,causes,and rec- tions weredrawnregardingtheseresults. analyzed, andsummarized,conclusionsrecommenda- causes, andsafetyrecommendations.Thedatawerecompiled, the quantitiesofandcorrelationsbetweeneachreport’sfindings, 1990 to1997.Thebasicmethodologyconsistedofexamining Covers”) fromtwoaccidentseachyearduringtheperiod tions. ThisstudyexaminedtheNTSBaccidentreports(“Blue ity ofthereports,and,byinference,qualityinvestiga- the AirLinePilotsAssociation,International. not intendedtorepresentanyofficialpositionsoropinionsof Note: Thispaperreflectsthepersonalviewsofauthorandis Abstract Midwest 5481,Emery17,Swissair111andTWA800. participated inanumberofaccidentinvestigations,includingAir ment rating,andisalicensedA&Pmechanic.Inaddition,hehas Investigation School,holdsaprivatepilotcertificatewithaninstru- and manufacturingexperience.HeattendedtheUSCAircraftAccident performance andflighttest,buthealsohasnearly4yearsofdesign 108 By MichaelHuhn(MO3689),SeniorSafetyInvestigator,AirLinePilotsAssociation,International.Thispaper This studyshowsthattheNTSBaccidentreportsinfrequently • Of Finding-Cause-Recommendation was presentedbyChrisBaum,ManagerofALPA’sEngineeringandAccidentInvestigationDepartment. ISA craft accidentreportsinanefforttodeterminethequal- recent NationalTransportationSafetyBoard(NTSB)air- he intentofthisstudywastoevaluateaselectedgroup SI 2003 An AnalysisoftheRelationship P Fairchild. Hisprimarybackgroundisaircraft including DouglasAircraft,Rockwell,Northrop,and after 16yearsatvariousaerospacecompanies, aeronautical engineeringandcametoALPAin1994 Pilots Association,International.HehasaB.S.in the EngineeringandAirSafetystaffatLine Michael Huhn r oceedings From SelectedRecentNTSB Aircraft AccidentReports istheSeniorSafetyInvestigatorof qualitative methods wereusedinthisanalysis. Thispaperin- the subjectNTSBaccidentreports, andbothquantitative Sweginnis astheprimarymeans ofdeterminingthequality tion. Thisstudyutilizedthe methoddescribedbyWoodand only havebeentheproductof ahigh-qualityaccidentinvestiga- cation ofrobustinvestigative and analyticalprocesses,could dent report,onecanconclude thattheseresultsareastrongindi- the conclusionsbyanalysis,andsoon. sure thattherecommendationsaresupportedbyconclusions, port, theauthorssuggestreviewingreportbackwards to en- safety recommendations.Finally,asameansofauditingthere- conclusions shouldleaddirectlyto,aswellsubstantiate, the rectly to,aswellsubstantiate,theconclusions.Similarl y, the properly andinthisF.A.C.manner,theanalysisshouldleaddi- sions. WoodandSweginnisstatethatifthereportisconstructed logically andcontainstheappropriateinformationdiscus- F.A.C. conceptasameanstoensurethattheaccidentreportflows Quality Control,”andthissectionutilizesexpandsonthe ments intheaccidentreport.Sub-chapter5isentitled“Report cates alogical,building-blockapproachtodevelopingtheargu- fact, analysisandconclusion,respectively,themethodadvo- method oftreatingtopicsdiscussion.Theselettersstandfor bodied bythereport.Inthischapter,theydiscuss“F.A.C.” concept, andameansofdetermininghowwellithasbeenem- 38 ofWoodandSweginnis(1995)providestherationaleforthis the facts,conditions,andcircumstancesofaccident.Chapter mendations shouldbedirectlyrelatedtoandsubstantiatedby ings, causes,andrecommendations.Inotherwords,therecom- cident reportwillexhibitastrongcorrelationbetweenitsfind- ity, and,byinference,thequalityofinvestigations. ports inanattempttodevelopageneralassessmentoftheirqual- cent NationalTransportationSafetyBoardaircraftaccidentre- of theinvestigation,thispaperexaminedaselectedgroupre- final reportisonemeanstoevaluatethequalityandeffectiveness effort putintothem?Usingthepremisethatqualityof of NTSBinvestigationsiscommensuratewiththeexpenseand tremely safe,canonenecessarilyinferfromthatthequality there isnodoubtthattheU.S.airtransportationsystemex- mendations areissuedinitsfinalreportoneachaccident.While major civilairlineaccidents.Normally,theBoard’ssafetyrecom- charged withconductingandreportingontheinvestigationsof Transportation SafetyBoard(“NTSB”or“Board”)istheagency If theabove-describedrelationshipsareprevalentinacci- The underlyingconceptinthiseffortisthatahigh-qualityac- metric forjudging thequalityofNTSBreports, theworkload that accountingforthesediscrepancies wouldprovideanother study toidentifyorquantifythose findings.Whileitisrecognized enumerated asformalfindings, noattemptwasmadeduringthis appear inthetextofNTSB reportsandyetnotbeexplicitly sions” sectionofeachNTSBreport. Althoughmanyfindingsmay by thisstudy In allcases,thefindings,causes,andrecommendationsanalyzed Preliminary dataanalysis characteristics, arepresentedinTable1. These final16accidents,alongwiththeirpreviouslytabulated into terrain(UCFIT),oneinflightfire,andthreeicingevents. controlled flightintoterrain(CFIT),threeuncontrolled flight dents. These16accidentsincludedfiverunwaycollisions, four tion. Usingthiscriterion,thestudyscopewasreducedto16 acci- sults fromtheserepeataccidentstypesovertheperiodin ques- be someadditionalbenefitinpossiblycomparingtheNTSBre- ber ofrepeataccidenttypes,anditwasdecidedthattheremight umn ofthetabulationinAppendix1revealsasignificantnum- headers andthedataisalsoincludedinAppendix1. tabulation appearsinAppendix1.Anexplanationofthecolumn accident andcorrespondingreportweretabulated.Thisoriginal select thesefinalaccidents,certaindefiningcharacteristicsofeach to twoaccidentsfromeachyearoftheoriginalscope.Inorder source constraintsofthisauthor,itwasdecidedtolimitthestudy ough analysisofsuchalargegroupreportswasbeyondthere- date accidents.Sinceitwasrecognizedthatthecriticalandthor- selection process. fatalities resultedwerenotexplicitfactorsintheinitialorfinal • • scope ofaccidentsforthisstudywereasfollows: the “BlueCover.”Theotherparametersusedtoinitiallylimit aircraft accidentreport,whichisknownthroughoutindustryas in thisstudywasthattheinvestigationresultedafullNTSB tion. Insummary,theinitialqualifierforinclusionofanaccident and afewoftheseaccidentshavebeenincludedinthefinalselec- to thepointwheretheycommandsignificantNTSBresources, personnel, thefollow-oninvestigativeactivitiessometimesexpand investigations initiallyinvolveonlyalimitednumberofNTSB Washington, D.C.However,whilesomeNTSBregionaloffice manpower responsebypersonnelfromNTSBheadquartersin the “MajorInvestigations,”whichinitiallyinvolveasignificant time andfinancialresources.Typically,butnotalways,theseare investigations thatutilizedlargeamountsofNTSB’spersonnel, of theaccidentinvestigations,itwasdecidedtolimitscope step process.Sinceonegoalofthisstudywastoassessthequality The selectionofthefinalsubjectaccidentreportswasamultiple- Selection ofstudyaccidents Methodology study oftheNTSB. parison withsomeresultsfromtheRANDCorporation’srecent conclusions, andrecommendationsofthisstudy,aswellacom- cludes detaileddiscussionsofthemethods(includinglimitations), Inspection ofthedatain“ACCIDENTCATEGORY”col- The preliminaryselectionprocessyieldedagroupof44candi- It shouldbenotedthatthetypeofaircraftandwhetherornot T Date ofAccident:Accidentoccur ype ofOperation:LimitedtoF only includethoseexplicitlylisted inthe“Conclu- AR P r ed between1990and1997 art 135or121operations The ValuJet592report hadthemostfindings(47), whilethe 16 reportscontainedatotal of 401findings(refertoTable2). the NTSBfindingsfrom 16 subjectaccidentreports.These The firststepintheprimary analysisconsistedofsummarizing NTSB findings Analysis results These rawdatamatricesarepresentedinAppendix3. relationship, anda“0”denoteslackofthatdefinedrelationship. safety recommendation.Inthismatrix,a“1”denotesthedefined and whethereachprobablecontributingcauseresulted ina each probableandcontributingcausewassupportedbyafinding, recommendations. Inthiscase,thegoalwastodeterminewhether probable andcontributingcauseswithboththefindings safety tion arepresentedinAppendix2. agonal linesintheboxes.Thesematricesofrawdatacorrela- have correspondingsafetyrecommendationsaredenotedbydi- the boxes.Similarly,findingsofsafetydeficienciesthatdonot have correspondingFindingsaredenotedbyhorizontallinesin in therespectivematrix.Safetyrecommendationsthatdonot tion wasdetermined,andthiscorrespondenceisnotedbya“1” dence betweeneachindividualfindingandsafetyrecommenda- most columnofthematrix.Foreachaccident,anycorrespon- outline boxes.Thesafetyrecommendationsarelistedintheleft- deficiencies (asdeterminedbytheauthor)aredenotedbold are listedacrossthetoprowofmatrix,andfindingssafety ber astheyappearintheNTSBaccidentreport.TheFindings matrix listsallthefindingsandsafetyrecommendationsbynum- two two-dimensionalmatrices(tables)werecreated.Eachlarge findings, causes,andsafetyrecommendations.Foreachaccident, the specificrelationshipsexistingbetweeneachaccidentreport’s sented inthetablesthispaper. ings, andsimilarcalculatedvalues.Thesederivedvaluesarepre- tion, thepercentageoffindingssafetydeficienciestototalfind- data. Thesederivedvaluesincludethedurationofinvestiga- sisted ofcomputinganumberderivedvaluesfromtheraw tative analyseswereconducted.Initialquantitativeanalysiscon- Once thefinalaccidentswereselected,quantitativeandquali- Primary dataanalysis relayed totheflightcrew”wouldqualifyasasafetydeficiency. ing of“Presencethemicroburst-inducedwindshearwasnot not becategorizedasafindingofsafetydeficiency,butfind- microburst thatresultedinsignificantlow-levelwindshear”would hypothetical findingof“Thethunderstormproducedastrong ered tobeafindingofsafetydeficiency.Asanexample,the “manmade” event,andthatseemstobecorrectable,wasconsid- Generally speaking,afindingthataddressesanintentionalor sary todeterminewhichfindingsrepresentedsafetydeficiencies. deficiencies. Therefore,somejudgmentbytheauthorwasneces- not explicitlycategorizeorevenalwaysstateitsfindingsassafety several non-U.S.transportationsafetyagencies,theNTSBdoes umn “FNDNGDEF”)arenecessarilysubjective,becauseunlike ception. Thequantitiesofthefindingssafetydeficiencies(col- 1 andAppendix1)isobjectivedata,thereonesignificantex- associated withthiswasbeyondtheresourcesofauthor. Also foreachaccident,amuchsmallermatrixcorrelates the The qualitativeanalysisconsistedoftheefforttodetermine While mostoftheinformationinAccidentSummaries(Table ISASI 2003 Proceedings • 109

ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS ing, theNTSB’sprobablecause findingisthe“ballgame.” for amajoraviationaccident, bythemediaorinalegalproceed- halls oftheNTSB.Interms theassignmentoffaultandblame process. However,probablecausereverberates farbeyondthe cumulative fact-findingand analysis oftheNTSBinvestigative the statementofprobablecause….This statementreflectsthe most controversialresultoftheNTSB’sinvestigationprocess is are thefollowingremarksregardingprobablecause: “The Perhaps statedbestbytheRANDreport(Lebow,C.,etal,p. 42) or addressedbyrecommendations Correlation ofcausessupportedbyfindings recommendations couldalsobemoreconsistent.” port (Lebow,C.,etal,p.41),whichnoted,“Thepreparation of tions thatitissues.ThisobservationisechoedintheRAND re- safety deficienciesthatitidentifiesandtherecommenda- porting processdoesnotensurehighcorrelationbetweenthe standard deviationindicatesthattheNTSBinvestigativeandre- addressed byacorrespondingsafetyrecommendation.Thelarge identified safetydeficienciesinanygivenaccidentwillevenbe on theseresults,thismeansthatapproximatelyonlyhalfofthe and thestandarddeviationwas24percent.Insimpleterms,based ciencies, andnocorrespondingsafetyrecommendations. the AIA808accident,whichyieldedeightfindingsofsafetydefi- ratio of93percent.Attheextremeoppositeendthisscalewas “hits,” thisratiowassignificantlybetteredbytheCMR3272hit However, althoughVJ592hadthehighestabsolutenumberof ommendations, foracorrelationvalue(“hitratio”)of74percent. Twenty ofthese27VJ592Findingswereaddressedbysafetyrec- cies rangedfromahighof20(VJ592)tolowsix(TWA427). tabulations ofthiscorrelationarepresentedinTable3. was notobserved,andtheresultsdisplayedverywidescatter.The quality accidentinvestigationandreport.Suchidealcorrelation cies andsafetyrecommendationswouldbeexpectedfromahigh- one-to-one (orbetter)correspondencebetweensafetydeficien- ciencies wereaddressedbysafetyrecommendations.Ideally,a study, sinceitquantifieshowmanyoftheidentifiedsafetydefi- This analysisisconsideredtobethemostimportantpartof and safetyrecommendations Correlation offindingssafetydeficiencies ciencies andyieldedastandarddeviationofonly10percent. age of54percentthetotalfindingsweresafetydefi- ing betweenahighof78percentandlow38percent,anaver- age ofthetotalfindings,resultsaremuchmoreconsistent.Rang- when thefindingsofsafetydeficienciesarepresentedasapercent- the average,indicativeofawidelyscattereddistribution.However, 13, andagainthestandarddeviation(6)wasapproximatelyhalfof least (6).Theaveragenumberoffindingssafetydeficiencieswas Eastern 111report(7)wasedgedoutbytheTWA427for of safetydeficiencies.AgainValuJet592hadthemost(27),but distribution. Inthiscasetheauthoridentifiedatotalof215findings the numberoffindingsforeachaccidentwaswidelyscattered. findings was25,butthestandarddeviationof11indicatesthat Eastern 111reporthadtheleast(12).Theaveragenumberof 110 However, most safetyprofessionalsagreethat theNTSB’scon- The averagehitratioforthestudyaccidentswas54percent, As notedpreviously,thenumberoffindingssafetydeficien- Examination ofthefindingssafetydeficienciesshowsasimilar • ISA SI 2003 P r oceedings Example: formal NTSBsafetyrecommendations. On apositivenote,suchfindings almostalwaysalsoappearedas tial.” Inessence,thesefindings weresafetyrecommendations. usually containedtheword“should” orthephrase“itisessen- tently withthewordingof bulkoftheNTSB’sfindings,and gestions orimperatives.These findingswerenotwordedconsis- Several NTSBreportscontainedfindingsthatwerestatedas sug- Findings intheformofrecommendations accident beginswith“HadtheCessna441pilotvolunteered….” Example: ferred toequipmentnotinstalledorproceduresfollowed. findings werealsoofacause-and-effectnature,andtypically re- with thewords“if”or“had.”Mostcommonly,theseconditional Some findingswereofaconditionalnature,frequentlybeginning Conditional findings reduced thetimeavailableforasuccessfulemergencyevacuation.” damaged intheinitialcollision….Theacceleratedfiresignificantly the releaseofoxygenfromflightcrewsystemthatwas accident states,“Thepropagationofthefire…wasacceleratedby Example: accident report. separate, unrelatedfindingswerelistedasasinglefindinginan seemed tobestatementsofcauseandeffect,butoccasionallytwo actually containedmultiplefindings.Mostfrequentlythese Some findingslistedbyindividualnumbersintheNTSBreport Multiple findingslistedasasinglefinding quality andconsistencyoftheNTSBaccidentreports. here becausetheydoprovidesomeadditionalinsightsintotheoverall analytic resultsofthisstudy.Nevertheless,theyhavebeenincluded counting fortheseitemswouldsignificantlyalterthequantitative slightly lessrigid,theauthordoesnotbelievethatmeticulouslyac- While thesevariationsmadetheanalysisslightlymoredifficultand of somethevariationspresentinindividualNTSBreports. employed. Theseobservationsprovideseveralqualitativeexamples but whichfelloutsidethestatedscopeofanalysisprocessbeing during thegatheringandanalysisofsubjectNTSBinformation, This sectionbrieflydiscussesseveralobservationsthatweremade Additional observations addressed byformalNTSBsafetyrecommendations. probable causesand63percentofthecontributingare continuity hasdiminishedsignificantly.Only76percentofthe causes andsafetyrecommendations(Table5)isexamined,the ported byNTSBfindings.However,whenthecorrelationbetween able causesand93percentofthecontributingaresup- stantiated initsfindings.Anaverageof97percenttheprob- of ensuringthattheprobableandcontributingcausesaresub- passive…. Preventingrecurrencerequiresaction.” the samebook,IraRimsonisquotedstating,“Assigningcauses sis isneededtoidentifythemostpracticalremedialactions.”In that “…causesaremerelyconvenientcubbyholes….Furtheranaly- fact, itiscounterproductive.C.O.Miller(Weir,1999,p.227)notes able causes”drawsthefocusfromrealsafetyissues,andin tinued practice(whichislegislatedbyCongress)ofissuing“prob- A reviewofTable4showsthattheNTSBdoesathoroughjob NTSBfindingnumber19fromtheUSAirFlight1493 NTSBfinding number23fromtheAtlantic Southeast NTSB findingnumber11fromtheTWAFlight427 TABLE 1:FINALLISTOFSTUDYACCIDENTS vestigation process. is anareaforpotentialimprovementintheNTSB’saccidentin- the accidentinvestigationprocess,theseresultsindicatethatthis safety recommendation.Sincethereportisanintegralpartof given accidentwilllikelyevenbeaddressedbyacorresponding proximately onlyhalfoftheidentifiedsafetydeficienciesinany ops andwritesitsaccidentreportssincethestudyperiod,ap- determine theunderlyingreasonsfortheseresults. safety recommendationsthatitissues,thestudydidnotattemptto correlation betweenthesafetydeficienciesthatitidentifiesand the NTSBinvestigativeandreportingprocessdoesnotensurehigh port, aswellacrossallthereports.Whiletheseresultsindicatethat ited arelativelylargedegreeofinconsistency,bothwithineachre- mendations. Furthermore,thereportsincludedinstudyexhib- hibit strongcorrelationbetweentheirfindings,causes,andrecom- This studyshowsthattheNTSBaccidentreportsinfrequentlyex- Conclusions senger-carrying aircraft.” ing thedesignofcrashaxesrequiredtobecarriedaboardpas- Flight 529accidentstates,“Thereshouldbestandardsgovern- Unless theNTSBhassignificantlychangedwayitdevel- aim ofremovingthemandateforfindingProbableCause. courage theU.S.CongresstomodifyNTSBlegislationwith (or someothermeansofremedialaction). ated byfindingsandalsoaddressedsafetyrecommendations • • • and noteitherconditionalorimperativeinnature. • comply withthefollowingguidelines: dent reportwritingandreviewprocesstoensurethatthereports an identifiedcausalorancillarysafetydeficiency. • fied causalorancillarysafetydeficiency. • • to ensurethattheresultscomplywithfollowingguidelines: dent investigation,analysis,andsafetyrecommendationprocess The NTSBshouldevaluate,andmodifyasnecessary,theacci- Recommendations The NTSBshouldengagetheparticipationofindustrytoen- The NTSBshouldevaluate,andmodifyasnecessary,theacci- Ensur Sequence findingsinasclosetochr Establish aseparatecategoryforfindingsofsafetydeficiencies. Structur Ensur Develop oneormor Identify allcontributingandancillarysafetydeficiencies. e thatpr e thateachsafetyr e findingssothattheyar obable andcontributingcausesar e safetyr ecommendation issubstantiatedby ecommendations foreachidenti- e cleardeclarativestatements, onological or ISASI 2003 der aspossible. Proceedings e substanti-

N • 111

ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS 112 SAFETY RECOMMENDATIONS TABLE 3:CORRELATIONOFFINDINGSDEFICIENCIESWITH TABLE 2:SUMMARYOFFINDINGS • ISA SI 2003 P r oceedings TABLE 5:CORRELATIONOFCAUSESADDRESSEDBYSAFETYRECOMMENDATIONS TABLE 4:CORRELATIONOFCAUSESSUPPORTEDBYFINDINGS ISASI 2003 Proceedings • 113

ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS 114 National TransportationSafetyBoard.(1994). National TransportationSafetyBoard.(1994). National TransportationSafetyBoard.(1993). National TransportationSafetyBoard.(1993). National TransportationSafetyBoard.(1992). National TransportationSafetyBoard.(1991). National TransportationSafetyBoard.(1991). National TransportationSafetyBoard.(1991). Lebow, C.,Sarsfield,L.,Stanley,W.,Ettedgui,E.,&Henning,G.(undated). References ton, D.C.:NTSB. N334PX Hibbing,MinnesotaDecember1,1993 rain ExpressIIAirlinesInc/NorthwestFlight5719JetstreamBA-3100 04) Washington,D.C.:NTSB. Naval AirStation,GuantanamoBayCubaAugust18,1993 rain AmericanInternationalAirwaysFlight808DouglasDC-8-61N814CKU.S. Alabama June8,1992 rain GPExpressAirlinesIncFlight861ABeechcraftC99N118GPAnniston, March 22,1992 USAir Flight405FokkerF-28,N485USLaGuardiaAirportFlushing,NewYork Brunswick, GeorgiaApril5,1991 Flight 2311UncontrolledCollisionwithTerrainAnEmbraerEMB-120,N270AS 08) Washington,D.C.:NTSB. ternational AirportLosAngeles,CaliforniaFebruary1,1991 1493, Boeing737andSkywestFlight5569FairchildMetrolinerLosAngelesIn- ton, D.C.:NTSB. Airport Romulus,MichiganDecember3,1990 1482 and299RunwayIncursionCollisionDetroitMetropolitan/WayneCounty (NTSB/AAR-91/03) Washington,D.C.:NTSB. Atlanta HartsfieldInternationalAirportAtlanta,GeorgiaJanuary18,1990 Airlines Boeing727Flight111andEppsAirServiceBeechcraftKingA100 Monica, CA:RANDInstituteforCivilJustice. Safety intheSkiesPersonnelandPartiesNTSBAviationInvestigations. • ISA SI 2003 (NTSB/AAR-93/02)Washington,D.C.:NTSB. P r oceedings (NTSB/AAR-93/03)Washington,D.C.:NTSB. (NTSB/AAR-92/03)Washington,DC:NTSB. Runway CollisionofUSAirFlight (NTSB/AAR-94/05)Washing- (NTSB/AAR-91/05) Washing- Takeoff StallinIcingConditions Uncontrolled CollisionwithTer- Northwest AirlinesInc.Flights Atlantic SoutheastAirlinesInc Controlled CollisionwithTer- Controlled CollisionwithTer- Runway CollisionofEastern (NTSB/AAR-94/ (NTSB/AAR-91/ APPENDIX 1 Santa Wood, R.&Sweginnis,(1995). Weir, A.(1999). National TransportationSafetyBoard.(2000). National TransportationSafetyBoard.(1998). National TransportationSafetyBoard.(1997). National TransportationSafetyBoard.(1997). National TransportationSafetyBoard.(1996). National TransportationSafetyBoard.(1996). National TransportationSafetyBoard.(1996). National TransportationSafetyBoard.(1995). Endeavor Books. don, UK:Simon&Schuster. 1997 Korean AirFlight801Boeing747-300HL7468NimitzHill,GuamAugust6, ton, D.C.:NTSB. N265CA Monroe,MichiganJanuary9,1997 Uncontrolled CollisionwithTerrainComairFlight3272EmbraerEMB-120RT Florida May11,1996 Terrain ValujetAirlinesFlight592DC-9-32N904VJEvergladesNearMiami, Illinois November19,1996 press Flight5925andBeechcraftKingAirA90QuincyMunicipalAirportQuincy, AAR-96/06) Washington,D.C.:NTSB. 529 EmbraerEMB-120RTN256ASCarrollton,GeorgiaAugust21,1995 Forced LandingandCollisionwithTerrainAtlanticSoutheastAirlinesIncFlight ton, D.C.:NTSB. East Granby,ConnecticutNovember12,1995 Approach AmericanAirlinesFlight1572McDonnelDouglasMD-83N566AA 31, 1994 Transport Regional(ATR)Model72-212N401AMRoselawn,IndianaOctober Loss ofControlSimmonsAirlinesd.b.a.AmericanEagleFlight4184Avionsde November 22,1994 World AirlinesFlight427andSuperiorAviationCessna441Bridgeton,Missouri (NTSB/AAR-00/01) Washington,D.C.:NTSB. (NTSB/AAR-96/01) Washington,D.C.:NTSB. The TombstoneImperativeruthAboutAirlineSafety (NTSB/AAR-95/05)Washington,D.C.:NTSB. (NTSB/AAR-97/06)Washington,D.C.:NTSB. (NTSB/AAR-97/04) Washington,D.C.:NTSB. Aircraft AccidentInvestigation Runway CollisionInvolvingTrans (NTSB/AAR-98/04) Washing- (NTSB/AAR-96/05) Washing- Inflight LossofPropellerBlade Controlled FlightintoTerrain Runway CollisionUnitedEx- Inflight FireandImpactwith Inflight IcingEncounterand Inflight IcingEncounterand Collision withTreesonFinal Casper,WY: (NTSB/ Lon- ISASI 2003 Proceedings • 115

ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS 116 • ISA SI 2003 CORRELATION OF P r oceedings FINDINGS ANDSAFETYRECOMMEND APPENDIX 2 A TIONS (RA W D A T A) ISASI 2003 Proceedings • 117

ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS 118 • ISA SI 2003 P r oceedings ISASI 2003 Proceedings • 119

ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS 120 • ISA SI 2003 P r oceedings ISASI 2003 Proceedings • 121

ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS 122 • ISA SI 2003 P r oceedings ISASI 2003 Proceedings • 123

ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS 124 • ISA SI 2003 P r oceedings Key toCorrelationTables APPENDIX 3 ISASI 2003 Proceedings • 125

ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS 126 half andtotalseriousinjuries bynearly100percent. tured. Occupationalsafetydata increasetotalfatalitiesbyabout ground vehicleandsomeother rampequipment,arenotcap- involved. Eventsinvolvingtwo ormoregroundvehicles,a incidents, regardlessoftheir severity,inwhichaircraftarenot Finally, noneofthedatabases usedforthispapercaptureramp reported toinsurancecompaniesduehighdeductible costs. even insurancedataisquitelimited,asmostrampeventsare not cant eventsthatneithertheFAAnorNTSBidentified. Yet, tured bytheFAAdata.DatafromAirclaimsidentified19signifi- Even then,manyincidentsinvolvingaircraftinfactarenot cap- the datacaptureonlythoseeventsthatdirectlyinvolveaircraft. data areabitmorebroadlybased,butlimitedaswellin that database islimitedtoeventsinvolvingintendedflight.The FAA captured bygovernmentalorothercentraldatabases.The NTSB only asmallshareofevents,becausemostrampincidentsare not ramp operationalsafety. database. Thepaperalsocitesselectedresearchonthesubjectof carrier intheUnitedKingdom,asreportedUK’sAAIB recorded byAirclaimsandonerampaccidentinvolvingaU.S. Federal AviationAdministration(FAA),butitalsoincludesevents from theNationalTransportationSafetyBoard(NTSB)and cost oftheseevents. and damagetoaircraftotherproperty.PartIIIexaminesthe tified eventsforcommonfactors,typicalparticipants,injuries, issue andtherampenvironment.PartIIexamines679iden- T not representtheviewsoffederalAviationAdministration. The viewsexpressedinthispaperrepresenttheauthor’sandmay at theU.S.DOT. and severalyearsasanaviationanalystfortheOfficeofSecretary the OrganizationofEconomicCooperationandDevelopmentinParis, of Transportation(DOT),consultingwiththeEuropeanUnionand years innationaltransportationlegislationwiththeU.S.Department FAA andindustry.Hispreviousprofessionalexperienceincludes9 involved inCAST,FOQAandothercooperativeeffortsbetweenthe In theend, datausedinthispaperrepresent onlyabout Despite usingallthesesources,thepaper,atbest,addresses The analysisisbasedprimarilyonaccidentandincidentdata • ISA Involving U.S.Carriers,1987–2002 from 1987through2002.PartIoutlinesthescaleof 565 identifiedrampincidentsinvolvingU.S.aircarriers his paperpresentsananalysisof144rampaccidentsand SI 2003 By RobertMatthews,Ph.D.,OfficeofAccidentInvestigation,FederalAviationAdministration,USA Ramp AccidentsandIncidents P Robert Matthews past 8years.Mostrecently,hehasbeenheavily analyst intheOfficeofAccidentInvestigationfor the FAAsince1989,wherehehasbeenasafety University ofMaryland.Dr.Matthewshasbeenwith Analysis andisanAssistantProfessor,Adjunct,atthe Tech’s CenterforPublicAdministrationandPolicy r oceedings earnedhisPh.D.atVirginia air carrieraccidentsintheUnitedStates.Exhibits1and2 show aircraft inconfinedspaces,etc. gage, aircrafttowingandpushback,accessegressoflarge water andtoilets,loadingunloadingpassengersbag- marshalling, chocking,refueling,cleaning,catering,servicing around betweenlandinganddeparture.Theseactivitiesinclude tions associatedwithservicinganaircraftduringanormalturn- fully succinctdefinitionoframpactivitiestoincludeallopera- to andegressfromthegate.JohnsonMcDonaldofferause- mediately adjacentarea(the“alley”)thathandlesaircraftaccess around thegateonairportsideofterminal,plusim- This paperdefines“ramp”astherelativelysmallareaatand Part I:Scaleandrampenvironment do well. tions fortheseareastypicallyareinexpensivebutverydifficultto are procedures,training,andorganizationalculture.Interven- • erty, cancelledflights,andotherindirectcosts. billion annuallyininjurycosts,damagetoaircraftandotherprop- • ries andfivetosixminorinjuriesperyear. Aviation dataalsoreportanaverageofthreetofourseriousinju- year, ascapturedinotherpublicdatabasesonoccupationalsafety. aviation databases,plusanadditionalfatalaccidenteveryother issue. Onaverage,theycauseonefatalityperyear,ascapturedin • and conclusions: available data,theanalysisisabletoreachfollowingfindings ports on75percentofthecases.Despitelimitations just morethan20percent,andFAAdatawasthesourceofre- was thesourceforabout4percent;NTSB and/or intentofflight.Ofthe679eventsanalyzedhere,Airclaims high percentageofthemoreseriouseventsthatinvolveaircraft 2.5 percentofallevents,oraboutonein40,butshouldincludea Ramp accidentspersistentlyaccountfor20to30percentof all The principalcausesandpossibletar Ramp accidentsandincidentscostU.S.aircar Ramp accidentsandincidentsconstituteasignificantsafety 1 gets forcor riers ar r ective action ound $2 fined, andbusy areas.Theactivitybecomes stillmoreintense out confinedspaces.Intheend, rampareasarecomplex,con- their activitiesasverylargeaircraft moveto,from,andthrough- personnel addtotheramppopulation. Allthesepeopleconduct and others.Finally,FAAairports, fightstandards,andsecurity construction workers,aircarrier andairportengineers,planners dates airportoperationsand maintenancestaffs,airportpolice, ment, andpeopleontheramp.Therampareaalsoaccommo- volve additionalcontractors. eling, aircraftcleaning,catering,andlavatoryserviceoften in- or byairportservicecompanies.Specializedservices,such asfu- those servicesmaybeconductedundercontractbyothercarriers environment. Atstationswhereacarrierhaslimitedpresence, and pushbacktoensuretheoverallcontrolsafetyof ramp which providebaggagehandling,marshalling,aircrafttowing, Generally, largeraircarriershavetheirownrampdepartments, Ramp operations strophic eventatagateremainsreal. attention fromgovernmentandindustry,whiletheriskofacata- equal toormoreseriousthansomeeventsthatattractsubstantial dents andincidents,thesenumbersshowthatrampeventsare ditional ninefatalinjuriesand75seriousinjuries. Occupational SafetyandHealthAdministrationindicateanad- Data foranearlyidentical16-yearstudyperiodfromtheU.S. which sixweredestroyedand113incurredsubstantialdamage. reported minorinjuries.The679eventsinvolved821aircraft,of of whichinvolvedsinglefatalities,plus53seriousinjuriesand83 the distributionbetweenaccidentsandincidentsvariesabit. that thenumberofeventshasbeenfairlystablesince1988,though EXHIBIT 2.RAMPEVENTS,AIRCRAFTDAMAGE,ANDINJURIES. All theseactivitiesputaircraft,surfacevehicles,otherequip- Even ifwelimitthedataonlytoaviationdatabases,rampacci- The 679accidentsandincidentsincluded17fatalevents,all the aircraftorthereisnointention forflight. late safetyforrampworkerswhen nocrewmembersareonboard intended, whileOSHA,theairport authority,ortheStatesregu- the FAAregulatesworkersafety ontheramponlywhenflightis standards withintheirrespective regulatorydomains.Intheory, where otherfederalagencies chosetodevelopandenforcesuch for occupationalsafetythroughouttheeconomy,exceptin fields bor. TheOSHAActof1970establishedOSHA’sbasicauthority Health Administration(OSHA)intheU.S.Department of La- is dividedbetweentheFAAandOccupationalSafety and the carrierhasarelativelylimitedpresence. some orallrampoperationsatselectedstations,particularly where An aircarrierthenmaycontractwithathirdpartytohandle through localleasingagreementsorotherformalmechanisms. the responsibilityforrampsafetytoindividualaircarriertenants New Jersey).Howevermostairportoperatorsdelegatemuchof more thanoneState(suchasthePortAuthorityofNewYorkand the variousStates,includingsomeauthoritiescreatedjointlyby der tomaintaincertificationrequirementcompliance. That subpartrequiresanairportself-inspectionprograminor- supplemental guidanceonsafefuelprograms,lighting,etc. Part 139addressestherampenvironmentonlyindirectlywith operations inaircraftthatseatmorethan30passengers.Yet,FAA port thatservesscheduledorunscheduledaircarrierpassenger sets minimalsafetyrequirementsforthecertificationofanyair- safety atcommercialairportsintheUnitedStates.FARPart139 and interdependentschedulesaremaintained. more pressuretoturnaircraftaroundandensurethatcomplex during peakperiods,withsharpincreasesinaircraftvolume, In additionto the rampoperator,airtrafficcontrol (ATC)is Federal regulatoryauthorityforworkersafetyintheramparea Airport operatorstypicallyarespecialauthoritiescreatedby Generally, localairportoperatorsareresponsibleforramp ISASI 2003 Proceedings • 127

ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS Three illustrates, theriskoffatalandnon-fatal injuryrestsespe- posed tosomeriskatthegate orontheramp,but,asExhibit flightcrew memberillustrate thateveryoneinvolvedcanbeex- fell fromthecateringdoor.Thetwopassengersand single inside theaircraftduringboarding. Hecontinuedwalkingand passenger fatality,ablindelderly manwasleftalonebrieflyjust the crew,thenfellwhileexiting asmallturboprop.Intheother ties involvedaone-armedpassengerwhodeclinedassistance from the cargodoorduringloading.Thetwopassenger fatali- and oneflightcrewmember.Thememberfellfrom of scenarios,includingtwohulllossesduetoafireatthegate. ries involvedmoreaircraftdamageandavariedcollection instances ofincidentaldamagetoprops.Caseswithserious inju- the 17fataleventsinvolvedaircraftdamage,exceptforseveral 18 ofthe53seriousinjuries(35percent).Note,too,thatnone of period, wereinvolvedinnineofthe17fatalities(53percent) and counted for30percentofU.S.aircarrierdeparturesinthestudy disproportionately involveturboprops.Turboprops,whichac- events withfatalandseriousinjurieshaveadifferentprofile. rival. had foundthatmostrampaccidentsandincidentsoccuronar- gers, theyarelargelyassociatedwithdeparture.In1995,NASA ways fromothergateevents.First,exceptforinjuriestopassen- craft. The70casesinvolvingfatalorseriousinjurydifferinseveral involved seriousinjuries.All70oftheseeventssingleair- Of the679accidentsandincidents,17involvedfatalities53 Events withfatalandseriousinjuries aircraft-jetways, andaircraftintootherequipmentstructures. examined first,followedbyeventsinvolvingtwoaircraft,then dents, or“events.”Eventsinvolvingfatalandseriousinjuriesare Part IIexaminesthecharacteristicsoframpaccidentsandinci- acteristics andfactors Part II:Accidentsandincidentsontheramp,theirchar- are responsibleforthesafetransportationoftheirpassengers. responsible forthesafeoperationoftheiraircraft,andaircarriers workplace. Intheend,however,underFARs,pilotsremain which theregulatoryrelationshipsareatleastascomplex the localATCfacility. of decisionsonmovementareasarelefttolocalauthoritiesand in thenumberofenplanementsandphysicallayout,details no formal“non-movement”areasexist.Sinceairportsvarygreatly control handlesalloperationsontheairport(includingpushback), ramp areaaspartofthemovementarea.WhereATCground control ishandedofftoATConadifferentradiofrequency. alleyway, andexitfromthealleywaytosomedefinedspot,where radio frequencyandcontrolspushback,movementwithinthe airport. Typically,theaircarrier’srampcontroloperatesitsown A hubcarrieroftenprovidesrampcontrolforallcarriersatan ers definehandoffproceduresbetweenATCandrampcontrol. the ramparea.LocalagreementsbetweenATCandaircarri- have theirownrampcontroltowers,whichalltrafficwithin port movementarea.Offthearea,someaircarriers responsible forthemovementofaircraftandvehiclesinair- 128 The 17fatalitiesincluded14groundworkers,twopassengers In additiontooccurringmostlyondeparture,fatalinjuries The bottomlineisanintenselybusyandcloseenvironmentin ATC hastheauthoritytodesignatesomeorallofairport • 2 ISA Thisanalysisdoesnotconfirmthatfinding.Ataminimum, SI 2003 P r oceedings animal dropped byachilduponboarding, onlytofindthatthe to doagoodjob,suchasopening thedoortoretrieveastuffed involved flightattendantswho weremakinglast-minuteattempts Not unlikeseveralfatalinjuries togroundworkers,twoofthefive failure tofollowproceduresin openingcabindoorsatthegate. tor wasfailuretofollowprocedures (eitherbydriversorcabincrew). fant, astroller,andseveralbags.Inallfourcases,theprimary fac- crewmembers helpedtheboy’smother,whowascarrying an in- 4-year-old childwhoexitedviathecateringdoorwhile three ation inwhichapassengerwasinjured.Thefourthcaseinvolved a struck anaircraftontaxiout,causingafireandensuing evacu- gers boarded,causingpassengerstofall.Inathirdcase, avan jets. Intwoofthecases,surfacevehiclesstruckaircraftaspassen- bundles ordecliningassistancethatwasoffered. as passengersdisembarkingwhilecarryingexcessivenumbers of gers astheyexited.Passengernegligencealsoplayedarole,such bottom stepoftheaircraftdoor,andfailuretomonitorpassen- handrails, improperplacementofauxiliarystepsbeneaththe in mostcases.Proceduralissuesincludedairstairswithoutproper senger injuries,notwithstandingthepassengers’contributions passenger’s hand. attendant closingthecabindoorprematurely,breakinga while disembarkingorboarding,thoughonecaseinvolvedacabin props. Inmostcases,passengersfellfromorslippedonairstairs Of the11seriousinjuriestopassengers,seveninvolvedturbo- attendants infiveevents,andaflightcrewmemberoneevent. 36 injuries.Passengerswereseriouslyinjuredin11events,flight ible lightning). management oftools,etc.),failuretohaltoperationsduringvis- ment availabletoworkers(headsetswithlimitedcords,shift wings ornearengines,inadequatestaffing,equip- manuals andproceduresthatdidnotaddressmovingbeneath dicated moresystemicissues,suchasinadequateornotraining, waiting toboardflights,etc.Otherproceduralshortcomingsin- get last-minutebagsonboard,off-dutyworkershelpingoutwhile without wingwalkers,approachinganaircraftunannouncedto a betterjob,suchaschoosingtoguidelargejetfromthegate ACCIDENTS, ANDREPORTEDINCIDENTS. EXHIBIT 3.DISTRIBUTIONOFINJURIES,RAMP volves somevisualdifficultyindetectingarapidlyrotatingprop. head injury.Theissue,again,isoneofprocedurebutalsoin- for fiveseriousinjuries,includingaseveredhandandsevere ing wasinadequate(orutterlyabsent).Propstrikesalsoaccounted were inadequateor,morefrequently,notfollowed,ortrain- tating props,usuallyatnight.Inall14cases,procedureseither cially withsurfaceworkers. Of thefivecabinattendants seriously injured,fourinvolved The remainingfourseriousinjuriestopassengersoccurred on Failure tofollowprocedureswasaprimaryfactorinmostpas- Of the53seriousinjuries,groundworkersaccountedfor“only” Procedural shortcomingsofteninvolvedworkerstryingtodo Of 14fatallyinjuredgroundworkers,eightwerestruckbyro- dural shortcomings. four casesisthatsurfaceworkers arenotthesolesourceofproce- no procedureforhandlingthe situation.Therealpointofthese procedures, whenatowbardisconnected andtheoperatorhad who failedtofollowprocedures andoneexampleofinadequate was destroyed.Thefourcases alsoinvolvedtwogroundcrews age tootherequipmentanda fireatthegateinwhichanaircraft to severeinjuries(lossoflimbsorcrushinginjuries),plus dam- engine-start procedures(causingjetblast).Thesefourcases led cedures onpushback,whiletwootherflightcrewsfailedto follow procedures. Intwocases,flightcrewsfailedtofollowbraking pro- them routinelytooperatetoocloselynosewheels. in whichshortcordsrestrictedworkers’movementsorcaused equipment, asmarshalersorwingwalkersworkedwithheadsets collapsed tailstrut).Severalothercasesincludedinadequate tributable tofaultyequipment(aloaderwithbrakes anda involve rampprocedures.Thosetwocaseswereexclusivelyat- ries and/ordamagetoaircraftandotherproperty. that influencedmanyotherrampeventsinvolvingminorinju- ety ofproceduresandotherissuesthataffectrampsafety cedures. However,therangeofcausalfactorsillustratesvari- inadequate rampproceduresorsomeone’sfailuretofollowpro- was destroyed. airport fireserviceextinguishedthequickly,butaircraft the flightcrewandthreepassengersincurredminorinjuries.The attendant orderedanevacuationwithoutcommunicatingwith to afirebeneaththeaircraft’sengine.Inthatevent,cabin and struckagroundpowerunit(GPU)withoneprop,whichled crew. Theaircraftranforward,crushedagroundworker’sfoot, other hulllossinvolvedaprematureenginestartbytheflight jumped thechocksandpoweredintoterminalbuilding.The thrust” positionandbrakesnotset.Theaircraftimmediately chanics startedtheenginesatgatewiththrottlein“full ating anaircraftbeforetheday’sfirstflight.Inthatcase,me- occurred ondepartureandtheotherinvolvedmechanicsoper- substantial damagetoaircraftandtwohulllosses.Oneloss different frominjuriesnotedabove,16ofthe53casesinvolved (five) involvedissuesuniquetoturboprops(propstrikes).Also ing injuries. reports citedseveredlegsorarms,whileothersseverecrush- ally weremoreseverethanother“serious”injuries.Atleastfour This numbermaybesurprisinglylow,butthoseinjuriesgener- paperwork, limitingtheirvisuallookout. yield toanaircraft,whiletheflightcrewtaxiedcompleting follow procedure.Thebusdriverranastopsignandfailedto NTSB citedboththebusdriverandflightcrewforfailureto sion withanemployeebusastheaircrafttaxiedinalley.The compliance. stop, knockingdownaflightattendantwhowascheckingseatbelt and thegroundcrewsignaledtoflightforanemergency In thatcase,atowbarbecamedisconnectedduringpushback injuries toflightattendantsinvolvedanyrolebyrampsupport. jetway hadalreadybeenmovedback.Justoneofthefiveserious Four ofthe36casesinvolvedflightcrewswhofailedtofollow Just twoofthe36seriousinjuriestogroundworkersfailed Serious injuriestogroundworkersalmostuniformlyinvolve Unlike theinjuriesnotedabove,arelativelysmallshareofcases The remaining36seriousinjuriesinvolvedgroundworkers. The loneseriousinjurytoaflightcrewmemberinvolvedcolli- events aredescribe, below. aircraft andotherproperty, minorinjuries.Severalclassesof The largemajorityoframp incidents involveonlydamageto Events withminorinjures and aircraftdamage dents (nineof17)andinserious injuriestopassengers. including props.Turbopropswereover-represented infatalacci- they typicallyresultfromseverefallsorbeingstruckbyanaircraft, to rampworkers.Fatalinjuriesseldomincludeaircraftdamage; (eight of14).Injuriesbypropsalsoexplainedfiveseriousinjuries common fatalscenarioinvolvedrampworkersbeingstruckby props serious injuriestopassengersoccurredduringarrival.The most ramp workers(36of53seriousinjuriesrecorded).However, most preparation fordeparture,aswasthecasewithseriousinjuries to serious injuries.Mostworkerfatalitiesoccurredondeparture or Ramp workersaccountedfor14ofthe17fatalitiesand36 of53 Summary offatalandseriousinjuries most pervasiveissueintheentiredataset. ture isbeyondthescopeofthispaper.Yetculturemaybe area aspartof“flight.”Clearly,atreatiseonorganizationalcul- flect aculturalvalueinwhichwesimplyfailtotreattheramp Injuries toflightattendants,thoughfewinnumber,alsomayre- to reduceinjuriesflightcrews,cabinandpassengers. workforce, whilesimultaneouslyworkinghard(andproperlyso) cate awillingnesstoabsorbinjuriesthatsegmentofthe inherent costsofdoingbusiness.Worse,theymightevenindi- a certainsensethatrampeventsandinjuriesmaybeoneofthe are notsuggestedherelightly. values andbehaviors.Simplyput,thesearenoteasytasks vene andactuallybringtheorganizationtoprescribedsetof (sometimes remarkablyarrogantnotions),and(3)wecaninter- culture shouldchangeandthecharacteristicsthatitadopt the existingculture,(2)wecanidentifydirectioninwhich least thefollowing:(1)wecancometounderstandandarticulate Conscious effortstochangeanorganization’scultureassumeat ture toooftenisgliblysuggestedasanavenuetoasafersystem. notion ofcultureiscitedherealmostapologetically,becausecul- workers impliesafundamentalculturalissueintheindustry.The The repeatedissueofproceduresandlimitedtrainingramp The roleoforganizationalculture drove toofast,oroutsideofdesignatedareas. Drivers oftensimplyfailedtoyieldtherightofwayaircraft, tors ofcateringtrucks,buses,fuelbaggageloaders,etc. injuries involvedproceduresandtrainingfordriversopera- procedures werefoundtobeinadequateorabsent.Otherserious to aircraft.However,inmanyofthesecases,standardoperating workers hadnotclearedthearea,orsimplyoperatingtooclosely for communicatingwitheachother,failuretoobservethatother procedural failurestypicallyinvolvedfailuretofollowprocedures most commonissuesinvolvedmarshalersandwingwalkers.Their baggage loaders,buses,maintenancevehicles,andothers.The half thecases)tooperatorsoftugs,cateringtrucks,fuel of groundactivities,frommarshalersandwingwalkers(about inadequate training—proceduralissuesappliedtoabroadrange or failuretofollowthoseprocedures.Manycasesalsoinvolved Nevertheless, commonandrepeatedproceduralissuesimply The remaining30casesinvolvedgroundoperatingprocedures ISASI 2003 Proceedings • 129

ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS ever, 62percent ofthisgroupinvolvedflight crewprocedures remainder wereevenlysplitbetween arrivalanddeparture.How- ing bymechanicsaccountedfor 20percentofthecases,while Jet blast during boardingordeplanement). surface workers(suchastug or truckoperatorsstrikingjetways Company procedureswerean issueinseveralcases,aswereother 9 percentinvolvingrampconditions,especiallysnowand ice. operators, with11percentinvolvingflightcrewprocedures, and However, 15percentalsoinvolvedproceduralfailurebyjetway position, misjudgingclearances,andcommunicationfailures. every foursucheventsarerelatedtomarshalers’being out of mon (andexpensive)scenario(54ofthe679events).Three of Aircraft andjetways mechanics. percent ondeparture,and10duringrepositioning by designed gateareas,etc.Halfofthecasesoccurredonarrival, 40 trained mechanicstomarshalaircraft,operatingininadequately combined 25percentofthisgroup).Issuesincludedusingun- volved rampconditionsandinadequatecompanyprocedures(a dures, failingtosetbrakes,etc.).Theothercommonfactorsin- crews (followinstructionsofgroundguides,engine-startproce- third oftheseeventsinvolvedroutineproceduralfailuresbyflight ing clearances,communications,etc.Besidesmarshalers,one- marshalers outofposition,failingtoconfirmall-clear,misjudg- with issuessimilartothoseinaircraft-vehiclecollisions,suchas damage. substantial damage(qualifyingasaccidents)and234withminor events (282aircraft)producednohulllosses,but34aircraftwith rarely leadtoinjuriesorsevereaircraftdamage.The141such ramp area.Theseeventstypicallyproducesubstantialcostsbut captured intheaviationdataisaircraft-to-aircraftcontact Aircraft-to-aircraft contact of issuesinherentinrampsafety. vice vehicles.Thisvariedlistisagoodindicationofthebreadth loader operators,busdrivers,and,finally,driversoflavatoryser- operators, cateringtruckoperatorsofothervehicles, ority ranking,tugoperators,fuelingbaggagecart workers wereprominentinthisgroupaswell,including,pri- failure tofollowchockingprocedures.Howeverotherground ing clearances,failingtofollowcommunicationsprocedures,and to ensureall-clearbehindorbesideamovingaircraft,misjudg- By far,themostcommonfailuresinvolvedmarshalersbyfailing volved proceduralandtrainingissuesamongsurfaceworkers. imply culturalissuesasmuchanythingelse. sent companyprocedures.Asnotedabove,thesefiguresmight ramp area.Stillanotherfivepercentinvolveinadequateorab- In mostcases,thecarrierisresponsibleforconditionof ice andsnoworclutter,inappropriateequipmentbeingused. surface workersarenottheonlyissue. inappropriate engine-starts).Again,thismakesthepointthat ure tofollowguides,failuresetbrakes,misjudgingclearances, 271 events,15percentwererelatedtoflightcrewprocedures(fail- studies, 271(40percent)wereaircraft-vehiclecollisions.Ofthe involves asingleaircraftandsurfacevehicle.Ofthe679events Aircraft andsurfacevehicles 130 Again, marshalingproceduresaccountedforhalfofallcases, Yet, asignificantmajority(nearlytwo-thirds)ofallcasesin- Another 15percentinvolveeitherrampconditions,suchas • ISA . Jetblastsaccountedforjust41 oftheevents.Reposition- SI 2003 P r oceedings . Aircraftstrikingjetwaysistheremainingcom- .

The nextmostcommontypeofevent . Themostcommonscenario,byfar, case thatthetotal costoframpaccidentsand incidentsisonthe side withanassumedratioof just4to1andstilleasilymakethe other goverymuchhigher.We canremainontheconservative DOT, andothers).Qantashasestimated ratiosof7to1,while scale ofdirectcosts(seeFlight SafetyFoundation,BoreneratU.S. or settlinglitigation,etc. costs toothercarrierssharingaterminal,theofdefending tomer base,temporaryorevenextendedlossofgates,possible tions, permanentlyforegonetripsbytravelers,somelossof cus- craft, thecostofreplacementstafforothertypesrealloca- for aircraft(oftenextensive),thecostofleasingreplacement air- such asnetworkcostsassociatedwithcancelledflights,down time However, thesesumsaremodestwhencomparedtoindirect costs, for strandedpassengersintheeventofcancelledflights, etc. and otherproperty,thecostofstaffovertime,hotels any injuries,damageandrepairstoaircraft,structures,vehicles that sameamountjustamongU.S.aircarriers. of thatera. about halfoftheindustry’slossesduringeconomicdownturn in theearly1990samongaircarriersofWesterncountries,or ramp accidentsandincidentsequaledabout$2billionannually In 1994,theUnitedKingdom’sAAIBestimatedthatcostof Part III:Costoframpaccidentsandincidents each categoryhadaslightlydifferentprofilefromtheothers. of thescenariosoutlinedabovesharedmanycharacteristicsbut conditions andcompanyproceduresaresignificantissues.Some nant, bothforgroundworkersandflightcrews.However,ramp without risk.Clearly,proceduralissuesareoverwhelminglydomi- property, ratherthanfatalorseriousinjury,buttheyarehardly The largemajorityofeventsinvolvingdamagetoaircraftand Summary ofdamageaccidentsandincidents dures (6percent)accountedfortheremainder. percent), equipmentfailures(6andcompanyproce- nication, inappropriateengine-start,etc.).Rampconditions(11 the cases(misjudgingclearance,followguide,setbrake,commu- of thisgroup.Flightcrewprocedureswereissuesin30percent dlers, tugoperators,truckandothersalsowerepart wing walkerswerethemostcommongroup,butbaggagehan- and/or trainingissuesamongsurfaceworkers.Marshalersand 679 events).Three-quartersoftheseeventsinvolvedprocedural porary structures,lightpoles,etc.,wasfairlycommon(91ofthe of otherproperty,includingterminal,constructionfacilities,tem- Aircraft andotherproperty aircraft, jetways,hangars,vehicles,andothergroundequipment. terminals (glassandstructures),damageto20otheraircarrier cant. Theeventsidentifiedinthisanalysisincludeddamageto bris. Asaresult,thedamageassociatedwithjetblastsissignifi- cases. to theaircraft).Carrierproceduresalsoexplainedseveralof workers (marshallingerrorsandinjuriesfromwalkingtooclose 14 percentofthesecasesinvolvedtheperformancesurface tions, confusionuponenteringrampareas,etc.Arelativelymodest engine-starts, addingthrustinclosespaces,notfollowinginstruc- (not double-countingmechanicoperators)duetoinappropriate Modest estimatesputindirect costsatthreetofivetimesthe Direct costsarefairlystraightforward.Theyincludethecostof Jet blasts,ofcourse,canmoveallsortsequipmentandde- 3 Thispaperfindsthatrampeventsnowcostabout .

Similarly, aircraftstrikinganassortment the industry. All theseissuesimplyamorefundamentalculturalissuewithin and inadequatecompanyproceduresalsowerecommonissues. 25 percentofallcases.Rampconditions(especiallyiceandsnow) common, butflightcrewproceduresalsowereinvolvedinabout sues. Proceduralfailuresamongsurfaceworkerswereespecially another, whiletrainingandinexperiencealsowerecommonis- dressed inthispaperinvolvedproceduralfailuresofonesortor impose verysubstantialcostsontheindustry. Accidents andincidentsinthisenvironmentoccurfrequently threats forpassengers,crew,and,especially,surfaceworkers. people aremovingabout.Consequently,rampsposerealsafety seemingly endlessvarietyofaircraft,vehicles,equipment,and Ramp areascanbeintenselybusy,confinedspaces,inwhicha Conclusions stantial figure.However,itisverylikelyontheconservativeside. rect costsforeventsnotcapturedinthisdocument. average thatthemorecommonestimateofabout$250,000indi- tive intheestimatedcosts,wecanalsoassumeamoremodest seem outofline.However,intheinterestsremainingconserva- a result,theaveragedirectcostof$600,000inthisdatasetdoesnot where aswelldonotcaptureaircraft-to-aircrafteventsverywell.As Note, too,thatmostdatabasesintheUnitedStatesandlikelyelse- FAA databasesaremorelikelytocapturethesevereoutcomes. of averagecosts,butitisreasonableifwerecognizethatNTSBand costs ofabout$600,000.Thisismuchhigherthanmostestimates higher ratiocouldincreasetheestimatedcostsproportionately. order of$2billionayearintheUnitedStatesalone.Assuming The question,though,iswhattodoaboutrampaccidentsand Very nearlyallofthe679rampaccidentsandincidentsad- In theend,$2billionannuallytoU.S.carriersisaverysub- Either way,theeventsidentifiedforthispaperaverageddirect 3 2 1 Footnotes costs associatedwiththem. and theirappropriateactionsrequiredtoreducetherisk sumably, betterunderstandingofthecharacteristicstheseevents reliable database.Thatwouldenablebetteranalysisand,pre- better reportingofeventsandthedevelopmentamuchmore undertook afterits1994conference.Thatis,committoensuring ticularly theFAA,couldcopyeffortthatUnitedKingdom that endforseveralyears).Ataminimum,bothagencies,par- needs tobeclarified(thetwoagencieshavebeenworkingtoward At aminimum,theregulatoryrolebetweenFAAandOSHA ing services,etc.,aswellairportauthoritiesandothers. service providers,suchastoiletservices,fuelingcater- agement, butalsotheemployeesandmanagementofallsorts of playersareinvolved,includingaircarrieremployeesandman- costly place.Thepaperindicatedrepeatedlythatawidevariety to takesomeactioniftherampisbemadealessriskyand and theirrespectivecharacteristics. and gateoperations,asillustratedbythevarietyofbasicscenarios and trainingissuesalsoneedtoaddressthefullrangeoframp mit thefiscalandsafetyvalueofprocedures.Theprocedural they cantraintothoseprocedures,andbegintrans- they doso,canestablishmeaningfulandclearprocedures, fronts, providedthatcompaniessustaintheeffortovertime.If needs tobecarefuloutthere.”Yet,actioncantakenonthese ture change”oftenisnomoreusefulthanfindingthat“everyone incidents. Advocatingbetterprocedures,training,and“cul- See“Rampsafety”in AviationSafetyReportingSystem,NASA;April1995. Donald,NickandFuller,Ray,“TheManagementofSafetyontheAirport Ramp,” AviationPsychologyandPractice,page68. Finally, government,too,needstoimproveitsroleinthefield. However, carriersarefarfromtheonlyorganizationsthatneed Aerospace , March1995,page8. N ISASI 2003 Proceedings • 131

ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS 132 • ISA SI 2003 P r oceedings SESSION VI ter, whichisaiming tofurtherimproveaccident prevention. planes. it involvesdomesticorinternational flights,civilormilitary incident investigationperformed inBrazil,regardlessofwhether ter) isthetopsupervisorfor everysingleaircraftaccidentand systemic basis. (SIPAER jargon),reporttoCENIPAandoneanother ona organization structure.Allthoseoffices,called“safety links” Bases, andsoon,arerequiredtohaveasafetyofficein their muter, aswellaircraftmanufacturers,flyingschools,AirForce cratic drawbacks.Allairlines,commercial,regional,or com- allowing anexpeditiousflowofinformationwithoutbureau- Force Commander. and isundertheChiefofStaff,whoreportsdirectlyto Air (CENIPA) isSIPAER’scentraloffice.CENIPAlocatedinBrasília The AeronauticalAccidentPreventionandInvestigationCenter tigation System,whichisinchargeofallsafetymattersBrazil. SIPAER standsforAeronauticalAccidentPreventionandInves- II. SIPAER(BrazilianSafetySystem) where almost70percentoftheaccidentstakeplace. Finally, itwillbeshownhowBrazilisimprovingsafetyatairports, gation boardmembersare,whopaysforthecosts,andsoforth. jurisdiction foraccidentinvestigation,includingwhotheinvesti- signments. Here,onewillseetheadvantagesofhavingregional for safetypurposewillbediscussed,alongwiththeiras- Following that,themainsevenareasinwhichBrazilisdivided tured, showingthesafety“links”spreadthroughoutcountry. accident investigation.Atfirst,wewillcoverhowSIPAERisstruc- words, SIPAERpavesthewayforacheaperandmoredynamic to helpsafetyinvestigatorscopewithlocalcharacteristics.Inother posed byadevelopingnationreality. carry outtheirtask,nottomentioneconomicconstraintsim- United States,Brazilposesachallengeforsafetyinvestigatorsto sq miles(8,511,180km),alittlebiggerthanthecontinental second onlytotheUnitedStates.Embracinganareaof3,286,170 Brazil hasthesecondlargestcorporateaviationfleet,ranking million passengersgoingthroughBrazilianairportsin2002. A lotofaccomplishmentshave beenattainedattheSafetyCen- In summary,CENIPA(hereafteralsocalled theSafetyCen- As asystem,SIPAERhasdynamicandmodernstructure, The BrazilianAviationSafetySystem(SIPAER)wasdesigned Accident InvestigationInBrazil Latin America.Countrywide,therewere75 port, inSãoPaulo,isthebiggestofitskind ies, SãoPaulowithmorethan20millionpeople. also hasoneofthetopfivemostpopulatedcit- ing thelargestrainforestinworld,Brazil Brazil isacountryfullofcontrasts.Whilehav- I. Introduction By Col.MarcusA.AraújodaCosta,ChiefAeronauticalAccidentPreventionand As toaviation,GuarulhosInternationalAir- Investigation Center(CENIPA),Brazil,KeynoteSpeaker occur inthetakeoff, initialclimb,finalapproach andlanding Sixty-eight percentofaccidents involvingcivilcommercialjets IV. Safetyintheairport close toairports,CENIPAhas decidedtofocusonthatarea. are headedbyAirForceofficers. Giventhatmostaccidentsoccur mand resources.BothCENIPA andtheCivilAviationDepartment tions arepaidbythegovernment,heavilyusingAirForce Com- and followsAnnex13fromICAO.Costsforaccidentinvestiga- six members,includingpilots,aviationdoctors,andpsychologists, Central andSouthAmerica. 5,000 students,withrepresentativesfrom18States,mainly from craft accidents.Uptonow,CENIPAhasgraduatedmore than in thecrashlaboratory,whichreproducesaround8actualair- nal fee.Studentsarefacedwithwreckageinvestigationtechniques course isfreeofchargefornationals,andforeignerspaya nomi- force pilots,engineers,psychologists,airportpersonnel,etc).The day. Therearemorethan40differentinstructors(airlineandair and takethesamecourse. dent investigation,sinceallinvestigatorsgraduatedfromCENIPA and militarypilotsinBrazilsharethesameknowledgeacci- civilians andmilitaries.Atthispoint,itisworthnotingthatcivil ways athandinallowingfacilitiesforsafetymeetings. and professionalrelationshipwithairportmanagers,whoareal- tions, operators,flyingschools,etc.Furthermore,thereisagood speeches tolocals.Safetysurveysarealsoconductedinrepairsta- ics intheirregion,theRSOscanproperlyaddressseminarsand the aviationcommunity.Knowingmostofpilotsandmechan- majority ofaircraftaccidents,theRSOsplayanimportantrolein compulsory inBrazil.Besidesbeingchargeofinvestigatingthe controlling andsupervisingallsafetyrecommendations,whichare board forfurtheranalysis. the investigation,reportcanbereturnedtoinvestigation is thefinalstep.ShouldanyagencyinAICbeunsatisfiedwith through anAccidentInvestigationChain(AIC),inwhichCENIPA until theDACtakesoverprocess.Allaccidentreportsgo ated aircraft,investigationsareinitiatedbytherespectiveRSO there isno“GoTeam.”EvenforaccidentswithRBHA121oper- the DAC.SinceRSOsaredistributedthroughoutcountry, under RBHA121(equivalenttoFAR121)thatareinchargeof respective areas,exceptoccurrencesinvolvingaircraftoperating The RSOsareresponsibleforinvestigatinganyaccidentintheir one mainsupervisorattheCivilAviationDepartment(DAC). fices—RSOs (locatedattheCivilAviationRegionalDivisions)and For civilaviation,SIPAERhassevenmainRegionalSafetyOf- III. Theinvestigationprocess The requiredAccidentInvestigationBoardhasaminimum of The courseisa7-weektrainingprogramwithsevenclasses As totheinvestigationteam,boardiscomposedofboth CENIPA issolelyresponsibleforissuingthefinalreportand ISASI 2003 Proceedings • 133

ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS are spentyearlyworldwideasaconsequenceofbirdstrikes,not in theneighborhoodofaerodromes.Infact,millionsdollars word “bird”popsupinourminds,sincemostbirdstrikesoccur eration hasthreeormoresafetystaffgraduatedfromCENIPA. Brazilian airportshandling97percentofthetotalaviationop- than 800peoplehaveattendedthatcourse.Today,eachofthe65 management, basicinvestigationtechniques,etc.Sofar,more driving, dangerousgoodhandling,apronsafety,hazardreport port emergencyplan,airportsafetysurvey,birdstrike,defensive once ayear.Subjectscoveredincludebutarenotlimitedtoair- most allBrazilianairports. specialists. Ahealthysafetyculturehasbeenimplementedinal- would probablyhavegoneunreported,comefromairportsafety has receivedrelatedtohazardsandincidents,whichotherwise manager. Toillustratethis,mostreportsthattheSafetyCenter benefits receivedfromairlines,passengers,andbytheairport or intheairportwillbehandledaccordinglyfrombeginning. ing so,theSafetyCenterensuresthatmishapsoccurringaround close towherenearly70percentofaccidentstakeplace.Bydo- developed asafetycourseforairportpersonnel,sincetheywork actions rightaftertheevent.Keepingthisinmind,CENIPAhas A goodandthoroughinvestigationrequirescompleteprompt venting otheroccurrences,thusmakingairtransportationsafer. dents and44percentofallfatalities. proach andlandingaccidents)accountingfor56percentofacci- phases, whichmeansinthevicinityofairports,withALAs(ap- 134 In Brazil,maybeelsewhere,whenwetalkaboutairports,the The course,AccidentPreventionCourse—Airports,isoffered As toaccidentprevention,itisimportantrememberthe Accident investigation,asweallknow,hasthepurposeofpre- • ISA SI 2003 P r oceedings pecially forStateswithlimitedresources. sustained asingularfatalitytopassengersinthelastsevenyears. Force halveditsaccidentrateinthepastdecade,majorairlines have benefitedthemostfromsuchauniquesystem.WhileAir despite itschallengingeconomicreality.Civilandmilitaryaviation cant progressintheaccidentpreventionandinvestigationarena, and forIFRairportstheASAradiusis20km. certified. ForVFRoperatingairports,theASAradiusis13km, varies insize,dependingonwhethertheairportisVFRorIFR including, landfills,slaughterhouses,tanningindustries,etc.ASA ment ofanyactivitythatattractsbirdsinthevicinityairports, tuted theAirportSafetyArea(ASA),makingillegalestablish- by theMinistryofEnvironmentin1995.Thatresolutioninsti- ment todatewashaveaResolution(similarlaw)enacted National BirdStrikeCommitteewascreated.Itsmajorachieve- to correcttheproblem,butnothingworkeduntil1995,when ground deterrentdevices,likegascannons.Alothasbeentried far. Theuseoffalcons(“falconry”)didnotwork,neither ing highaltitudes. wingspan andweighsabout1.6kg,flyinginthermalsreach- quent customeratlandfillsanddumps.Ithasa1.5maverage ing animportantroleintheecologicalsystemandbeingafre- deal with.Itsfavoritedishisspoiledordeterioratingmeat,play- counted inourstatistics,theurubuismostdifficultoneto bird, “urubu”(CoragypsAtratus).Althoughotherbirdshavebeen to mentionliveslost. SIPAER hasproventobeaneffectiveandefficientsystem,es- Having justonehybridsafetysystem,Brazilhasmadesignifi- The urubuhasbeenresistanttoallcountermeasuresusedso The maindomesticproblemisrelatedtoaone-of-a-kindblack N in dailyoperations. tion ofcorrectiveactionstothose accidentprecursorsthatoccur toward aprocessinvolvingthe identificationandimplementa- to moveawayfromemploying traditionalinvestigationtechniques progressive processes,theairline safetyprogramshavebeenable mous amountofdataonadaily basis.Withtheadventofthese clude sophisticatedprocessesinvolvingtheanalysisofan enor- gators ofhighlyvisibleeventstotheprogramstodaythat in- Airline safetyofficeshaveevolvedfromstrictlyincidentinvesti- Introduction carrier involvedinanaccidenthaveanyoftheseprograms. accident investigatormayeffectivelyusethisinformationshould a what thefuturemighthold.Thepaperwillalsodiscusshow the data, currentindustryinformationsharingefforts,andalook at vention, limitationsandbarrierstoeffectiveutilizationofthe safety tant safetyprograms,theircontributionstoaccidentpre- with systematicallytobeeffective. brought withitalargeincreaseininformationthatmustbedealt significant sophisticationintheairlinesafetyprogramsandhas assurance (FOQA).Theinitiationoftheseprogramshasenabled tion safetyactionprogram(ASAP)andflightoperationsquality employee self-reportingphilosophy.Theseprogramsaretheavia- implemented voluntarysafetyprogramsbasedona the FAAandemployeelabororganizations,airlineshave our complexanddynamicoperations.Throughcollaborationwith venting orlimitingthecontributionsofsystembreakdownsin tion ofmultiple-incidentinvestigationswiththeintentpre- honing accidentinvestigationskillsbutratherintheinvestiga- now arandomevent.Airlinesafetyofficesdonotconcentrateon Airline safetyhasevolvedtoalevelwherehulllossaccidentis Abstract and flightdataanalysisprogramdevelopment. recorder specificationsandrequirements,flightsafetydataexchanges, has participatedinnumerousindustryeffortsinvolvingflightdata technology, andinnovation.Healsoholdsaprivatepilot’slicense. Washington University,withemphasisinthemanagementofscience, engineering fromOhioStateUniversityandanMBAGeorge Logan holdsabachelorofscienceinaeronauticalandastronautical Prior to1983,LoganwasaBoeingFlightTestAnalysisEngineer. Airline SafetyData:WhereAreWeand Along withthese newabilitieshavecome concernsabout This paperintendstodiscussthedevelopmentoftheseimpor- By TimothyJ.Logan,Director,FlightOperationalSafetySouthwestAirlinesCompany Association intheAccidentInvestigationDepartment. From 1983to1992hewaswiththeAirLinePilots Safety andQualityAssurancefrom1992until2001. he heldpositionsofManagerandDirectorFlight previously wasemployedbyNorthwestAirlineswhere Operational SafetyforSouthwestAirlines.He Timothy Logan Where AreWeGoing? iscurrentlytheDirectorofFlight to self-reportviolations byprovidingthereporting crewmember ASRS providedcrewmembers aplacetosubmitsafetyissuesand NASA AviationSafetyReporting System(ASRS) reports. limiting thesubmission,for mostpart,toonlythoserequired grams inlimitingthequality ordetailofthereportsaswell tendency towardenforcementworkedagainstthereporting pro- dual purposeforthereportslimitedtheireffectiveness.The FAA’s cial ifplacedinadatabasewithappropriatecategorizations, the ceived bythecrewmember.Whiletheseprogramswerebenefi- NTSB. Thesecondpurposewastoreportsafetyissuesas per- hicle forreportingofoccurrencesasrequiredbytheFAAand the that servedtwopurposes.Thefirstpurposewastoprovide ave- Most airlineshadimplementedflightcrewreportingprograms Pilot reportingprograms the accesstoandeffectivenessofthisinformation. limited accessanduseofFDRdatathat,insomecases,reduced In addition,pilotassociationcollectivebargainingagreements contained onlyminimalparametersthatlimiteditsusefulness. lend itselftotheprocessingofmultipleflights.FDRinformation readout softwarewasnotsophisticatedand,therefore,did cident investigationsonanas-neededbasis.MostairlineFDR Manual on theFDRandwhatwasstatedinairline if any,informationontheactualoperationoutsidewhatwasseen (FDR) analysiswasdoneonindividualevents,sotherelittle, Prior totheintroductionofFOQAandASAP,flightdatarecorder Flight datarecorders line industrythatweexperiencedinthelate1980sand1990s. ing andwasnotconducivetotherapidpaceofgrowingair- type ofprocesswasandisverylaborintensivetimeconsum- prevention wasbasedonthefindingsoftheseinvestigations.This performed post-occurrenceinvestigationsofincidents.Accident line flightsafetyofficeswere,ineffect,miniNTSBprogramsthat For themostpart,upuntilapproximately1995,U.S.air- Incident investigations Past airlineflightsafetyofficepractices to NTSBincidentandaccidentscenariosprocedures. from anindustrysafetyperspectiveinadditiontohowitrelates use oftheinformationfromtheseprogramswillbediscussed and ASAPprogramsfromanairlinesafetyofficeperspective.The the ultimatebenefitsofthesetwoimportantprograms. part developedindependentlyofeachother,somewhatlimiting addition, thetwoprograms,ASAPandFOQA,haveformost the handlingandanalysisofthisdataonaregularbasis.In This paperwillhighlightthedevelopmentandfutureofFOQA (FOM).FDRreadoutswereaccomplishedtosupportin- ISASI 2003 Flight Operations Proceedings • 135

ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS In mostcases, agreements weremadeunder thecollectivebar- Company/pilot associationagreements ing theseprogramsatmostairlines. to bedone,butsignificantprogress hasbeenmadeinestablish- evolving positionsbyallparties involved.Thereisstillmuchwork new safetytoolsdidnotcome withoutalotofhardworkand new levelofsophisticationtoairlinesafetyprograms.But these The developmentofASAPandFOQAprogramshasbrought a ASAP/FOQA programdevelopment agement beforeASAPandFOQAcouldbecomeareality. paradigm shiftwasneededwithintheFAAandcarrier’s man- were alsosignificantlegalissuestoconfront.Inaddition,a huge had providedaroadmapforthedevelopmentofASAP,butthere associations thattheriskwasworthtaking. conquered intheUnitedStatestoconvincecarriersandpilot but significantworkonlegalandotherissuesremainedtobe nized. FOQAwaswell-developedoutsideoftheUnitedStates, dropped theneedformoresophisticatedtoolshasbeenrecog- airlines andthetravelingpublicwell,butasaccidentratehas the outcomeislimitedtoonlythoseATAmembersinattendance. cess ishighlyvaluedbytheFSCmembers,itveryinefficientand tential problemsthatmightaffecttheiroperation.Whilethispro- representatives tosharetheirexperiencesandlearnaboutpo- dent andincidentoccurrences.Thisprocessenablesairlinesafety approximately 20years,involvesthedetailedpresentationofacci- Committee (FSC)quarterlymeetings.Thisprocess,inplacefor conducted attheAirTransportAssociation(ATA)FlightSafety gram, airlineindustrysafetyinformationsharingwasandstillis Along withsafetyinformationbeingreportedviatheASRSpro- ATA FlightSafetyCommittee debrief results. ally notapartytothedebriefprocessbutmayhavehadaccess concerns ofpossibleFAAenforcementsincethewasgener- types ofprogram,whileeffective,stillwerelimitedbycrewmember crewmember toprovidetheneededdetailsofincident.These came withrelieffromcompanydisciplineasamotivationforthe tigations. Inmostcases,participationinapost-incidentdebrief in anattempttogainmoreinformationfollowingincidentinves- Some airlinesdevelopedflightcrewmemberdebriefprograms Post-incident crewdebriefs and balanceinformation,weather,ATCinformation. data thatmayhavebeenrelatedtoaneventsuchasaircraftweight ability oftheairlinesafetyofficesfromgatheringtime-critical Also, thesubmission limits theusefulnessonspecificissuesassociatedwithcarrier. identification involvedintheaggregatedataavailablefromASRS fication processinvolvedwiththeprogram.Inaddition,de- ASRS reportsfromtheircrewmembersbecauseofthede-identi- development ofASAPand,inasmallway,FOQA. amount ofdatatobeanalyzed.Italsoservesasthebasefor program wasandisahugesuccessdidprovidesignificant some administrativerelieffromFAAenforcement.TheASRS 136 The USAirwaysAltitudeAwarenessProgramandNASAASRS The programsandprocessesdescribedabovehaveservedthe Unfortunately, thecarriersafetyofficesnevergottosee • ISA SI 2003 P r oceedings 1 time frameofNASAASRSreducesthe usually ajointdecision betweentheairlineand thepilotassocia- ranted. AswithASAP,thedecisiontocontactacrewmember is and issignificantinconsequence thatfurtherexplanationiswar- event cannotbeexplainedthrough reviewoftheavailabledata complished onaggregatetrend information. grams. Themajorityofwork withintheFOQAprogramisac- Contact withflightcrewmembersisnotthenorminmost pro- the responsibilityofidentifyingindividualflightcrewmembers. representative. Thispositionistheonlypersonentrusted with designated “gatekeeper”positionnormally,apilotassociation tion. Inmostprograms,identificationinformationislimited toa in anagreementbetweentheaircarrierandpilotassocia- issues pertinenttoprogramdatahandlingarenormallyoutlined tion anddatahandlingprovisions.Theseprovisions other confidentiality offlightcrewsthroughexhaustivede-identifica- One ofthekeyaspectsFOQAprogramsismaintaining of FOQA formulate correctiveactions. viding thesafetyofficeincreasedinformationonwhichto lished. Inaddition,thequalityofreportsisimproved,pro- ported thathadnotbeenreportedbeforeanASAPwasestab- grams reportabetter-than-90-percentincreaseinincidentsre- incidents that,before,wouldhavegoneunreported.Mostpro- or riskbeingremovedfromtheprogram. the factthattheymustcomplywithERTrecommendations of correctiveactions.Theemployees’accountabilitystemsfrom ing employeetotheextentpossibleandforensuringfollow-up also responsibleforprotectingtheconfidentialityofreport- than riskingthelossofprogram.Theserepresentativesare drives thethreepartiestoseekresolutionofeventrather recommended correctiveactionortheprogramshutsdown.This of theevent.Thekeyfactoristhatallpartiesmustagreeto agree tocorrectiveactionthatmusttakeplacebasedonthefacts participate intheprocessingofeventinformationandjointly three stakeholders,normallycalledtheEventReviewTeam(ERT), association oremployeegroup,theaircarrier,andFAA.These ested-party system.Thisinvolvesrepresentativesfromthepilot employee(s). action recommendedwithintheprogramtoreporting means. Inmostcases,theresultisalesseningofcorrective differently thanthoseeventsthatarediscoveredthroughother handling ofsole-sourcereportswithintheprogramsistreated Since theintentofprogramistoencouragereporting, provide theonlysourceofinformationthataneventoccurred. ing sole-sourceevents.Sole-sourcereportsarethosethat One specificissuethatispertinenttoASAPtheactionsinvolv- ASAP individual employeesarenotspecificallyidentifiedwithevents. are outlinedalongwithconfidentialityrequirementstoensure rective action.Possibleactionsagainstindividualcrewmembers data willbecollected,andwhatprocessusedtotakecor- ated agreementshighlighttheboundariesofprogram,what and FOQAprogramswillberunateachoperator.Thenegoti- gaining processthatoutlinetheprovisionsunderwhichASAP Contact withlinecrewsisusually onlyaccomplishedwhenthe The uniqueprovisionsofASAPfosterincreasedreporting ASAP hasbeensetupasaprogrambasedonthree-inter- 2 even themilitar y. to otherlargeaircraftoperators, suchasgeneralaviationand ment oftheseprogramswithin theaircarrierindustryandalso 2003. Thesedocumentsprovide thebaselinefordevelop- velopment andpublishingof theFOQAA/Cisslatedforlate nership. AcorrespondingA/C forFOQAprogramsisunderde- how aprogramcanbedeveloped andestablishedasanFAApart- A/C outlinestheprovisionsofanASAP,alongwithprocedureson The FAAhaspublishedanadvisorycircular(A/C)onASA P. The Advisory circulars Part 193designatedprogram. used inappropriately.ItishopedASAPwillsoonbeincluded asa derived fromFOQAprogramsforfearoftheinformation being very difficultfortheairlinestoagreesharesafetyinformation prove flightsafety.Withoutthisprovision,itwouldhave been the pilotassociations.TheintentofFOQAprogramsis to im- velopment ofFOQAprogramssoughtbyboththeairlinesand tion isprovidedtotheFAA. fication ofindividualairlineFOQAinformationwheninforma- individual flightcrewmembers,alongwithpreventingtheidenti- The intentofthisdesignationistoensuretheconfidentiality sharing ofvoluntarilyprovidedFOQAinformationwiththeFAA. Freedom ofInformationAct(FOIA).Thisdesignationfostersthe grams, undertheprovisionsof14CFRPart193,exemptionfrom On June30,2003, Part 193FOIADesignation data onacontinuousbasis. and personnel,wellover1,000aircraftarenowproviding lines haveinvestedmillionsofdollarsinequipment,software, programs atallmajorairlineswithintheUnitedStates.Theair- ported. ThisrulehasallowedthedevelopmentofvoluntaryFOQA the UnitedStatesandonecarrierspilotassociationssup- is animportantstepinthedevelopmentofFOQAprograms erator oritsemployeesunderanapprovedFOQAprogram.This aggregate FOQAdatainanenforcementactionagainsttheop- acts, theAdministratorwillnotuseanoperator’sFOQAdataor for approvedFOQAprograms.Exceptcriminalordeliberate ance programs.Thisruleoutlinestheenforcementprotections tion (FAR),14CFR13.401,onflightoperationalqualityassur- On Oct.25,2001,theFAApublishedFederalAviationRegula- FOQA Rule13.401 ment totheseimportantprograms. of FOQAprogramsandshouldbecommendedonitscommit- The FAAhastakenconsiderableactiontoassistinthepromotion FAA actions analyzed, butindividualflightscannotbeidentified. mally 7-10days.Thereafter,aggregatetrendinformationcanbe matically andpermanentlyerasedafteraperiodoftime,nor- To ensureconfidentiality,theidentificationinformationisauto- agement Office(CMO)areconducted. airline, thoughregularbriefingswiththecarrierCertificateMan- tion membersoftheFOQAteam. This designationwasanotherimportantmilestoneinthede- In manyprograms,identificationinformationhasashelflife. The FAAisnotadirectpartytotheFOQAprogramwithin 3 theFAAdesignatedapprovedFOQApro- the FAA.The group hasalsodraftedaFOQA advisorycircular and mannerinwhichaggregate FOQAdatawillbesharedwith 14CFR 13.401,specificallythe paragraphreferringtotheform and concerns. government andindustryon FOQAregulations,policy,issues, to FOQAisneeded.TheARC alsofunctionsasaforumbetween pare recommendationsonwhetherfurtherrulemakingapplicable to providetheFAAwithadviceonFOQApolicyand pre- The FOQAARCwascharteredasagovernment/industry forum FOQA ARC tion sharinganddataconfidentialityissues. ARC hasalsoworkedcloselywiththeFOQAoninforma- and theFAAonprogramsthatarejustbeginning.The ASAP tion, theARCprovidesasourceofinformationforbothindustry ticed atallcarriersandacrossoperationaldisciplines.In addi- that theprovisionsofASAParecorrectlyandconsistentlyprac- standardization ofASAPprocessesandproceduresthatensure on ASAPpolicyandhasbeendirectlyinvolvedinpromotingthe pate. TheCommitteewasformedtoprovideadvisetheFAA ciplines sothatpilots,mechanics,anddispatcherscanpartici- to fosterthedevelopmentofASAPacrossairlineoperationaldis- of theASAPA/C.TheARChasalsoworkedwithinindustry The ASAPARC ASAP ARC opment. in interpretingtheFOQAruleandalsofosteringASAPdevel- The AdministratorhasdesignatedtwoARCtoassisttheFAA Aviation RulemakingCommittees(ARC) grams isveryencouraging. industry environment,thevaluethathasbeenplacedonpro- siderable trustamongallparties.Inlightoftherecentairline relief. ThedevelopmentofeitheraFOQAorASAPrequirescon- sponded positivelytotheFAAactionsonenforcementandFOIA ers. Theoperatorsandassociatedemployeeunionshavere- ployee groupsincludingpilots,mechanics,andflightdispatch- 40 ASAPprogramsacross28operatorcertificateholder’sem- approved FOQAprograms.Inaddition,theFAAhasaccepted As ofJune1,2003,thereare12U.S.operatorswhohaveFAA- Airline programs Current status the engineeringandmaintenancedisciplines. the U.S.militaryandalsoextendedcarrierFOQAprogramsto carriers. DemoProjhasalsohelpedpromoteFOQAprogramsto ity ofFOQAprogramsinexistenceatbothlargeandregionalU.S. has beenhighlysuccessfulasreferencedbythenumberandqual- specific programrequirements,andlessonslearned.DemoProj representatives meettodiscussFOQAprogramdevelopments, lar forumswhereaircarrier,pilotassociation,andotherindustry programs. Inaddition,theFAAhasandcontinuestosponsorregu- them topurchasehardwareandsoftwarejumpstarttheirFOQA programs. TheFAAprovidedfundingtoseveralcarriersenable provide seedmoneytoassisttheU.S.airlinesinestablishingFOQA In themid1990s,FAAestablishedFOQADemoProjto FAA FOQADemonstrationProject(DEMOProj) The FOQAARChasfocused oninterpretingtheFOQArule, 4 wascrucialinthedevelopmentandpublication ISASI 2003 Proceedings • 137

ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS FDR readoutand CVRtranscript.Inaddition, weoftenlackair of FOQAinformationandhundreds ofASAPreports,notasingle be obvious.Thedifferenceisthat wemaybedealingwithterabytes recorder andcockpitvoice andFOQAASAPshould also whyithappened.Thesimilarities betweentheflightdata way thatthesafetypersonnel notonlyknowwhathappenedbut suring oftheeffectivenesssafetyenhancements. ing environmentaffectingtheindustryandalsotoenable mea- to becontinuallymonitoredandadjustedaccountforthechang- the informationtoeffectimplementation.Safety has cedural developmentaspectsofourindustrymustbeprovided tified, thesubject-matterexpertsindesign,training,and pro- where thefailurecanormayoccur.Oncethesefailuresareiden- fact thatthepurposeofcollectingdataistoidentify areas and measuresafetyenhancements.Wemustalwaysfocuson the ning howwearegoingtoclassify,analyze,identify,implement, a giantdatabaseofFOQAand/orASAPinformationwithoutplan- want toaccomplish.Thebiggestmistakewecanmakeiscreate each ofthetwoprograms.Wealsomusthaveanideawhatwe discusses howtoaccomplishthesharingprocess. able obstacles,justfactorsthatmustbeincludedastheindustry information sharingacrosscarriers.Thesearenotinsurmount- programs, itprovidesanobstacleinimplementationofsafety assists greatlyinfosteringanentrepreneur-likeflexibilitythese grow theseprogramstofittheirrespectivecultures.Whilethis grams intheUnitedStatesisthatairlineshavebeenableto the keyaspectsofdevelopmentASAPandFOQApro- works directlyagainstfacilitatingindustrydatasharing.Oneof classification, andissuedetection.Thelackofstandardization resulted inalackofindustrystandardsondatacollection,issue airlines isconcerned. when theideaofsharingorcombiningsafetyinformationacross formation isachallenge.Thisproblemmagnifiedsignificantly information sources,collectivelyusingthisnew-foundsafetyin- rity reasons,alongwithsimplythespecificcharacteristicsof for themostpart,independentlyfromeachother.Fordatasecu- ing basis.Unfortunately,bothoftheseprogramshavedeveloped, more datathantheyareabletodefinitivelyanalyzeonanongo- from theirprogram,theyalso,toaperson,willtellyouhave an ASAPorFOQAmanagerwhocan’tpointtomajorsuccesses greatly exacerbatedthescopeofthisproblem.Whilethereisn’t ways thechallenge.ThedevelopmentofFOQAandASAPhas Collecting safetyinformationiseasy.Analysisofthisdataal- Industry datasharing try safetyissuesinaproactivemanner. the additionalinformationprovidedbyFOQAtoaddressindus- tant, istoestablishaprocesswherebytheindustrycanleverage aggregate FOQAdatatotheFAA.Thesecond,andmoreimpor- provisions oftheFOQArulerequiringoperatorstosubmit ommend totheAdministratoramethodforcompliancewith identified byFOQA.TheARChastwogoals.firstistorec- the sharingofsafetyinformationregardingindustryissues in late2003. that hasbeenpresentedtotheFAA.Itisintendedforpublication 138 Information fromFOQAmust beintegratedwithASAPina So howdoweit?Wefirstmustunderstandthespecificsof The independentdevelopmentofbothASAPandFOQAhas The ARCisnowconcentratingonestablishingaprocessfor • ISA SI 2003 P r oceedings tion analysisinvolves collectivelyanalyzingall ofthedatainaspe- eter valuestoastatisticaldistribution-type analysis.Thedistribu- from justmeasuringexceedences fromspecificpointsorparam- lyzed. FOQAanalysishasevolved inthelast10yearsmovingaway (feet andknots,etc.).Onceconverted, thedatacanthenbeana- verted fromdigitalformat(zeros andones)intoengineeringunits tion. Oncethedatareach FOQAGDRAS,theymustbecon- digital flightdataandanalysisoftheintouseableinforma- ibility oftheFOQAprogram. Lack ofqualitycontrolcanadverselyaffectdataanalysisand cred- ware thatmaybeloadedintheaircraftandalsoforGDRAS. mented. AlsoanychangesintheLFLmustbeincluded soft- of thedatarecordedonaircraftmustbemaintainedand docu- data fromtheaircraft.Thelogicalframelayout(LFL)orformat readouts arethemaintenanceandqualitycontrolofrecorded any dataanalysis.Justasimportanttoaccidentinvestigation FDR for errorsintheFOQAprogramthatmustberecognizedduring fice andtheairlineengineeringgroup.Thisisafertileground a continualeffortrequiringcoordinationbetweentheFOQAof- FOQA dataanalysiscantakeplace. bor-intensive andcostlyfunctionresultinginatimedelaybefore nology evolves,butcurrentlythisisstill,forthemostpart,ala- will eventuallybereducedorpossiblyeliminatedaswirelesstech- The datatransmissionissuesandrelatedcostsinFOQAprograms significant portionoftheoperatingcostsaFOQAprogram. Moving datafromtheaircrafttoGDRASforanalysisposesa have continuouslyoperatedduringthedatatransferprocess. an issueregardingpracticaltacticalapplicationtoaircraftthat may be5to10plusdaysoldbeforeitisanalyzedsotimeliness transportation ofthedatamayalsotakeadayortwo.FOQA nance. SincethismayoccurawayfromtheFOQAofficelocation, for downloadofthedataoccursinscheduledovernightmainte- due tothenatureofairlineoperationsinthatopportunity the GDRASusuallyinvolvessometimedelay.Thisisprimarily cases itisanexactcopyofwhatbeingrecordedontheFDR. cases, theparameterscanbeselectedbyoperator;inother Aircraft ConditioningandMonitoringSystem(ACMS).Insome recorded onquickaccessrecorders(QAR)orasafunctionofan System (GDRAS).TheflightdataformostFOQAprogramsare portation ofthedatatoGroundDataReadoutandAnalysis ings totheoperationalenvironment. and thethirdinvolvesdataanalysisapplicationoffind- data acquisition,thesecondinvolvesprogramqualitycontrol, gether forasuccessfulprogramtofunction.Thefirstinvolves are threecriticalaspectsofFOQAprogramsthatmustcometo- from aircraftatregularintervalsduringlineoperations.There FOQA consistsofthecontinualreviewflightdatadownloaded FOQA portant safetyprograms. tempt tohighlightthebenefitsandlimitationsofthesetwoim- Let ustakealookattheindividualprogramspecificsinanat- analysis takesonawholedifferentdimensionwiththeseprograms. craft maintenancerecordstoassistusintheinvestigation.Safety traffic controltranscripts,currentweather,flightreleases,orair- The thirdaspect,dataanalysis,involvestheconversionof the Quality controlmanagementofthedatainFOQAprogramsis In mostprogramstransporting,thedatafromaircraftto Data acquisitioninvolvestheflightdatarecordingandtrans- factor isnormally accomplished.Examples oftheclassification Some sortofclassificationeach eventintothecategoryorcausal tional databaseforrecordkeeping andtoperformdataanalysis. ers withintheprogram,itdoes adverselyaffectdataanalysis. helps tomaintainthecriticalrelationship betweenthestakehold- tion onceaneventhasbeen closedoutbytheERT.Whilethis most programsinvolvesomeaspectofpermanentde-identifica- possible topreservetheconfidentialityofinformation.In fact, sis ofASAPinformation. known, butitmustbeconsideredwhenaccomplishingany analy- percent ofactualevents.Thepercentagewillnever be mates arethatASAPreportsreceivedmayconstituteonly 10-50 dents butinmostcasesasubsetofactualeventsoccurring. Esti- ASAP dataalsodonotrepresenta100percentsetof inci- are subjectiveandinvolvetheobservationsofsingleindividuals. ment actionintheformofadministrativeaction.ASAPreports provided amotivationtofilethereportthroughreducedenforce- their mindscompromisesafety.Thereportingemployeesare volved inpossibleFAAviolationsorexperiencingeventsthat ASAP involvesthefilingofwrittenreportsbylineemployeesin- ASAP curity issuesandpotentialformisuseofthedata. ment, theFAA,andpilotassociationbecauseofdatase- program alsorequiresapartnershipamongtheairlinemanage- engineering, informationtechnology,andlinemaintenance.The tion withinandacrossairlinedepartmentsfromflightoperations, year. Italsorequiresasignificantamountofcontinualcoordina- mation downloadedandanalyzedinvolvesterabytesofdataper over thecountryorallworld,365daysayear.Theinfor- cur withinasystemconsistingof100plusaircraftoperatingall if needbe,furthermodifytheactionspreviouslyimplemented. to followuponcorrectiveactionsjudgetheireffectivenessand, continual monitoringofthelineenvironmenttoenableSME airline policyandprocedures.Akeypartofthisaspectisthe specialists, andotherswhoareinvolvedinthedevelopmentof airline engineeringmaintenance,lineemployees,ATC ment correctiveaction.TheseSMEsmayincludeflighttraining, within theorganizationwhocanactonfindingsandimple- of thesafetyinformationtosubjectmatterexperts(SME) entire distributionasopposedtotheindividualevent. benefit isderivedthroughmonitoringandmanagementofthe programs, butexperiencehasshownthatthemoresignificant tions fromtheexpectedperformancecanbedealtwithinmost that deviatefromthenormalizeddistribution.Significantdevia- in thelargeamountofaggregatedata,notindividualflights throughout theflight.Thereforebenefitofinformationis very powerful,enablinghundredsofdatapointstobetaken gram andeffectchangedirectly.TheGDRASsoftwaretodayis they relatetothecarrierpolicies,procedures,andtrainingpro- analyst tocontinuouslymonitornormaldistributionsandhow tributions thatcanassistinidentifyingincidentcausalfactors. tional patternsandthenlookfordeviationsfromnormalizeddis- cific areaofinterest,let’ssaytakeoff,toidentifynormalopera- ASAP reportsinmostprograms, ifnotall,arestoredinarela- As inFOQA,ASAPinformationisde-identifiedtotheextent These aspectsofaFOQAprograminlargeairlinemustoc- The remainingportionoftheanalysisaspectinvolvesfeeding This typeofanalysisisverypowerfulinthatitenablesthe airframe orengine manufacturerswiththe sameaccountability FAA? Also,howdoesthissafety informationgettransferredto with accountabilityandfollow uptotheotherbranchesof that involvesATC,airports,or aircraftcertificationgettransferred tered intheFAAFlightStandards branch.Howdoesanissue is implementedattheindustry level?ASAPandFOQAarecen- tified andwhohasaccountability toensurethatcorrectiveaction addressed iswhattrackdoesanissuetakeonceithasbeen iden- important aspectofindustrysafetyinformationsharing tobe discussed above.Whiletheseshouldnotbetrivialized,anequally gram tobesuccessful.Mostofthetechnicalissueshave been cal andproceduralissuesthatmustbeaddressedforthe pro- sounds likeagreatidea.Inpractice,therearesignificanttechni- to eliminateorreduceaccidentprecursors. can provideapromisingsourceofinformationthatbe used applies theinformationfrommultiplecarriersonaregular basis fied ATC,aircraft,andairfieldissues.Collectively,aprocessthat rective action.Already,individualFOQAprogramshaveidenti- appropriate SMEattheindustrylevelforimplementationofcor- multiple carriersidentifyingaspecificissuecanberaisedtothe readily bechangedbytheairlines.Collectively,informationfrom airport issues,oraircraftsystemdesignandoperationcannot cedures, ortrainingprograms.IssuesinvolvingATCprocedures, lines can’tfixthroughmodificationtotheirexistingpolicies,pro- issues. Thereareissuesidentifiedintheseprogramsthattheair- that thereisvalueinusingthisinformationtoaddressindustry As FOQAandASAPhavedeveloped,theindustryhasrealized Implementation ofindustrysafetyinformationsharing aggregation acrossairlineoperatorsattheindustrylevel. dardization ofeventclassificationworksagainstfacilitationdata ture andresourcesofeachairlineoperation,thelackstan- independently. WhileitisimportantthateachASAPfitthecul- the airlineshaveformostpartdevelopedthisaspectofASAP sification oftheASAPorhazardreportingprocesses.Therefore cation schemestoidentifyissues. line ASAPinformationbecauseitplaceslessemphasisonclassifi- sis. Itmayalsobeinstrumentalinassistinganalysisacrossair- tion withexistingclassificationprocessestoassistindataanaly- information. Textminingcanbeusedonitsownorincombina- received onamonthlybasiscontaininglargeamountsoftextual ise inassistingtheanalysisofhundredsASAPreports perimentation isinitspreliminaryphase,itdoesprovideprom- ASAP reportstoidentifytrendsorrelationships.Whilethisex- ing software,developedwithintheintelligencefieldtoanalyze ASAP. Severalairlinesareexperimentingwithapplyingtext-min- sis ofthewrittentextprovidedbyreportingemployeesin and toconserveresources. sessment practicestoprioritizecorrectiveactiondevelopment fied andimplemented.Someairlinesapplyformalizedrisk-as- be appliedtotheoperationandcorrectiveactioncanidenti- the mostsignificantissuesbeingreported.Thefactorscanthen cations canthenbeusedtoaccomplishdataanalysisidentify distractions airlines gofurtherandaddcausalorcontributingfactorssuchas would include In theory,theaggregationofindustrydataforsafetypurposes The industryhasnotsuccessfullydevelopedastandardforclas- A somewhatuntappedsourceofinformationinvolvestheanaly- or blocked frequency rejected takeoff,goaround,andaltitudedeviations todescribetheevent.Theclassifi- ISASI 2003 Proceedings . Some • 139

ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS airframe, butindividualflightanddateinformationwillnotbe FOQA informationcanbeusedtotrackaspectsofanindividual De-identification processinalloftheprogramswillpreventthis. information totracktheperformanceofindividualflightcrews. realize thattheywillbeunabletouseexistingFOQAandASAP programs totheaccidentinvestigator.Theinvestigatorsmust ment ofFOQAandASAPdirectlyaffecttheusefulnessthese The dataconfidentialityrequirementssocrucialtothedevelop- FOQA, ASAP,andtheaccidentinvestigator successfully implementsuchaprocess. long-term commitmentfromgovernment,industry,andlaborto of safetyinformationisnotgoingtooccurovernight.Itwilltakea the initialdevelopmentofFOQAandASAP,successinsharing can taketoeffectivelyusethisinformationeffectchange.Like carriers alongwithdevelopingaprocessorpaththattheindustry the technicalbarrierstoaggregationofsafetyinformationacross safety issuescanbeaddressed.TheARCisattemptingtoaddress base butmoreimportantlycreatingaforumwheretheindustry to createsuchaprocesswiththeemphasisnotoncreatingdata- industry safetyinformationsharingprocess. identified andaddressed.Thisshouldalsobethegoalof how muchdatawecancollectbutmanysafetyissuesare safety program,thegoalforindustrydatasharingshouldnotbe tion andaccountabilityaspectsofanindustryprocess.Asinany but itwillbeawastedeffortunlesswecansolvetheimplementa- taining FOQAand/orASAPinformationfromallofthecarriers, and follow-up?Theindustrycandevelopagiantdatabasecon- 140 Currently, throughtheFOQAARC,industryisattempting • ISA SI 2003 P r oceedings 2 1 Footnotes development oftheseimportantsafetyprograms. ing blocksonconfidentialitythathasbeensoimportanttothe prevention andatthesametimerespectingfundamentalbuild- tigation processbuttoproperlyusetheinformationinaccident der Annex13investigations.Theintentisnottoinhibittheinves- dards ondisclosureofvoluntarysafetyprograminformationun- place attheICAOleveltodeveloprecommendedpracticesorstan- and ASAPinformation.Inaddition,aparalleleffortneedstotake address properandresponsibleaccidentinvestigationusesofFOQA and theNTSBtogettogetheronthisissueassoonpossible and whatinformationisplacedinthepublicdocket. tor usestheinformationwithincontextofinvestigation involved inanaccident.Whatisimportanthowtheinvestiga- formation fromacarrier’sprogramsshouldanairlinebecome authorized authority,willhaveaccesstotheFOQAadASAPin- that accidentinvestigators,specificallyfromtheNTSBoranother honoring theconfidentialityaspectsofprogram.Itisexpected programs throughmisuseoftheinformationornot been uploadedtotheGDRAS. available afterapproximately7-10daysoncetheinformationhas 3 AviationSafetyActionPlanAdvisoryCircular120-66B,11/15/02 FlightStandardServiceOrder1110-129,ASAP,July2,2001 Federal Register 15468] Federal AviationAdministration14CFRPart193[DocketNo.FAA-2003- lations] [Page38594-38598]DEPARTMENTOFTRANSPORTATION It isimportantfortheaircarriers,laborassociations,FAA The accidentinvestigatorcandogreatharmtoFOQAorASAP : June30,2003(Volume68,Number125)][RulesandRegu-

N assembly canalter theinternalarrangement ofthepartsothat components awayfromtheir accidentpositions.Immediatedis- lead todamagingthepartor shiftingthepositionsofinternal altered forlaterpartsofthe sequence.Immediatetestingcan less ofwhichpathischosen first, thepartbecomesirrevocably sembly areactivitiesthatcan help theinvestigation;butregard- first, thenreassembledandfinally tested.Bothtestinganddisas- mediately andthendisassembled,ortheycanbedisassembled cally followoneoftwopaths.Thepartscaneitherbetested im- sonably intactpartsarerecovered,systemsinvestigations typi- part wasmalfunctioningatthetimeofaccident.Once rea- mary goaloftheaircraftsystems’investigatoristodetermine ifa condition afteritisrecoveredfromanaccidentscene.The pri- gation hascomeaboutfromaneedtodeterminepart’s exact The useofcomputedtomographyimaginginaccidentinvesti- Introduction applications ofCTscanningusedduringNTSBinvestigations. The paperalsopresentstheresultsofspecificaircraftsystems with emphasisonthebenefitsanddrawbacksofCTscanning. down first”questions. terest andmakebetterdecisionsregardingthe“testfirst”or“tear ter understandtheinternalconditionofcomponentsin- nents. Theuseoftheseimageshasallowedinvestigatorstobet- to provideimagesoftheinternalworkingsselectedcompo- mally knownascomputer-aidedtomographyorCATscanning) recently startedusingcomputedtomography(CT)scanning(for- technological solutiontohelpwiththisdecision. results oftheotherevent.Investigatorshavelonglookedfora reassembly andtesting.Eitherchoicewillirreparablyalterthe then disassembletheunits,orthemfirstfollowedby among investigatorsiswhethertotestthecomponentsfirstand craft systemscomponentsarerecovered,arecurringdebate When involvedinanaircraftaccidentinvestigationwhereair- Abstract University andisagraduateoftheU.S.NavyTestPilotSchool. degree inaeronauticalandastronauticalengineeringfromPurdue and conductingflighttestprogramsfortheU.S.Navy.HehasaB.S. Warren cametotheSafetyBoardafterspending11yearsdeveloping In thispaper,abriefoverviewofradiologicalprocessesisgiven, Investigators fortheNationalTransportationSafetyBoardhave Imaging inAccidentInvestigation Use ofComputedTomography tions includingthespaceshuttle has beeninvolvedinnumerousaccidentinvestiga- Safety Board,wherehehasworkedsince1997.He Systems InvestigatorsattheNationalTransportation Scott Warren Flight 185,andEgyptAirflight990investigations. Paul Wellstone’saccident,TWAFlight800,SilkAir By ScottA.Warren,LeadAerospaceEngineer,AircraftSystemsInvestigator, istheTeamLeaderforAircraft National TransportationSafetyBoard,Washington,D.C. Columbia, Sen. Figure 1 the part.InFigure1,internalarrangementofascrew-type of imagecanbeinterpretedtoshowtheinternalarrangement density material.Inmostcomponents,thisvariationtype rial, andthelighteritemsrepresentlessX-rayattenuationorlow- graph representhigherX-rayattenuationorhigh-densitymate- (as opposedtomedicalradiographs),darkeritemsinaradio- variations withinthepart.Generally,inindustrialradiographs dimensional projectionoftheX-rayattenuation(ordensity) ergy thanlow-densitymaterials.Theresultingimageshowsatwo- sity materialswithinthecomponentwillabsorbmoreX-rayen- rays aftertheyemergefromtheotherside.Ingeneral,high-den- using anX-raysourceandmeasuringtheattenuationofX- graph) ofabrokenbone.Itismadebyilluminatingcomponent age mostoftenusedbydoctorswhentheyorderanX-ray(radio- which thegeneralpublicismostfamiliar.Thistypeofim- A standardX-rayimageorradiographisthetypeofwith Radiograph Basics ofradiology through digitalprocessing. herent inthatprocessandtheimageenhancementsavailable investigator. Thisisduetothegreatlyimprovedresolutionin- scanning hasallowedforaquantumleapininformationthe nal conditionofapart.TheusecomputedtomographyorCT do notallowaninvestigatortogetacompletesenseoftheinter- known asaradiograph tor whoneededtolookinsideapartwassimpleX-ray,also part asitwasrecovered. subsequent testingafterreassemblyisnotrepresentativeofthe Previously, theonlytechnologicalaidavailabletoaninvestiga- 1 . Whileusefulinmanycases,radiographs 2 graph usesaphoto-detector dium used.Adigitalradio- involves therecordingme- ence betweenthetwo tal radiograph.Thediffer- tional radiographoradigi- duced eitherasaconven- gears andshafts. distinguish itemssuchas image, itismoredifficultto guished. Atthebottomof pins canbereadilydistin- connector, andtheconnector actuating screw,wiresforthe the image,itemssuchas tem) canbedetermined.In A300 directionalcontrolsys- actuator (fromtheAirbus A radiographmaybepro- ISASI 2003 Proceedings • 141

ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS 142 cally, imagesusedby theNTSBhavepixelsizes ontheorderof view ofthedetectorandnumber ofpixelsintheimage.Typi- attenuation values.Thepixel sizeisdependantonthefieldof value toeachimagepixel(picture element)basedontheX-ray view ofasingleslicetheobjects. perimposed images,whilethe CTimagecontainsan“overhead” clear. Theradiographproducesashadowgraphcontainingsu- differences inviewpointsbetweenthetwoimagingmethods are can befurtherexplainedbyreferringtoFigure4.Inthisfigure, of theitembeingscanned(seeFigures2and3). jection image.TheresultingCTsliceimageisathincrosssection age basedontheinformationcontainedineachindividual pro- completed, thecomputerreassemblescompleteCTslice im- projection. Afteracomplete360-degreerotationoftheobject is ent direction rotated slightlytoproduceanotherimagefromadiffer- nated atanyonetime.Aftereachimageistaken,theobject beam ofX-rayssothatonlyasmallslicetheobjectisillumi- ray intensity.TheX-raysourceisdesignedtoproduceaverythin the object,andthenadetectorisusedtorecordresultingX- to producearadiograph.AnX-raysourceisusedilluminate Allan MacLeodin1979. and hesharedaNobelprizeforthisworkwithphysicsprofessor tion ofCTimagingwasdevelopedbyDr.GodfreyHounsfield, ematical frameworkfortheconcept.Thefirstpracticalapplica- 1900s. TheAustrianmathematician,Radon,providedthemath- tions hasbeenunderstoodonatheoreticallevelsincetheearly process ofreconstructinganimagebasedonmultipleprojec- tions takenfrommanydifferentanglesaroundanobject.The age isproducedbyassemblingalargenumberofX-rayprojec- Computed tomography(CT)scanningisaprocesswhereanim- Computed tomography nal componentssuperimposedoneachother. dimensional “shadowprojection”ofthepartwithallinter- different componentsfromtheimage.Theimagepresentsatwo way todeterminethecompletespatialrelationshipsbetween computer software. but adigitalradiographcanbeprocessedandenhancedusing recorded onfilm.Theresultingimagesaresimilarinmanyways, to recordtheX-rayintensitieswhileaconventionalradiographis Figure 2:CTimagecreation. In creatingtheimage,computer assignsadigitalgraylevel The differencesbetweentheradiographandCTimages A CTimageisproducedusingequipmentsimilartothatused In anycase,thelimitationofaradiographisclear—thereno • ISA SI 2003 3 . Eachimageisstoredinacomputerassingle P r oceedings CT image(bottom). Figure 3:AirbusA300servoactuator(top),axialslice example ofaCT imageofanAirbusA300rudder servoactuator struct thefullobject’svolume. TheupperimageinFigure5isan volume oftheobjectbeingscanned, softwarecanbeusedtorecon- mensional imagecanbecreated. Sinceeachslicerepresentsathin NTSB haveslicethicknesses on theorderof1mmorless. voxel providebetterresolution. Typically,imagesusedbythe contained withinthem,sothegraylevelvaluesassignedto each slice. Thinnersliceshaveasmallerrangeofmaterialdensities given voxelbasedonawiderangeofdensitiescontained in the created usingthicksliceswillhavebrightnessvaluesassigned toa properties throughouttheentirethicknessofslice.Images tional imagearebasedonanaverageofthatlocation’smaterial erties (graylevelvalue)foreachlocationwithinthecross sec- even afractionofmillimeter)isdesiredsincetheimage prop- various thicknesses.Athinslice(ontheorderofmillimeters or used insteadofthetermpixel. images, theterm“voxel,”meaningvolumeelement,iscommonly brightness valueassignedtoeachpixel.WhendiscussingCT the pixelarea,createsavolumeofmaterialrepresentedby of theobjectbeingscanned.Theslicethickness,combinedwith thickness, eachpixelintheimagerepresentsaverysmallvolume 0.25 millimeters.SinceaCTimagerepresentssliceoffinite Figure 4:CTandradiographimagecreation. By combiningmanyoftheslice imagestogether,athree-di- The CTscanequipmentcanbeadjustedtocreateslices of est densityparts (partswiththehighestlevels ofX-rayattenua- ored red,andthemanifoldhousing iscoloredgreen other lowX-rayattenuationitems) intheservoactuatorarecol- cutaway viewofanobject.In Figure7,thehydraulicfluid(and digitally slicethroughanobject andlettheinvestigatorviewa through theuseofcolorand theuseofcutplanes,which The resultingimageisshowninFigure6. subtracted fromtheimage,steelinnermechanismsremain. tuator picturedinFigure5.Ifthelow-densityitemsaredigitally ray attenuationvaluethanthesteelinnermechanismsin ac- housing andotherlowerdensityitemscreateaverydifferent X- age basedontheirmaterialproperties.Thealuminummanifold ences canbeusedtoclassifyandselectdifferentpartsofthe im- ent X-rayattenuationlevelsatthedetectors,andthese differ- the scan.Differentmaterialsinscannedobjectcreate differ- enhance theinvestigator’suseofinformationcontained in the insideofobjectthatmakesCTimagessovaluable. useful inaninvestigation.Itistheabilitytocreateviewsof However, aviewoftheoutsideobjectisnotparticularly as electricalwires,wireclips,andsafetycaneasilybeseen. lution anddetailthataCTimagecanprovide.Smallitemssuch The CTimageshowninFigure5demonstratesthelevelofreso- Digital enhancementoftheCTimage tograph ofthesameser The viewoftheinsidean object canbefurtherenhanced Since theCTimageiscreateddigitally,softwarecanbeusedto vo actuator fromaslightlydifferentangle. image inFigure5isapho- millimeters thick.Thelower age wasapproximately0.95 ages. Eachsliceinthisim- ing morethan250sliceim- that wascreatedbycombin- photograph (above). slice image(left)and structed fromindividual servoactuator recon- Figure 5:AirbusA300 4 . Thehigh- digitally removed. Figure 6:CTimageofservoactuatorwithlow-densitymaterials tor tryingtodetermine eitherifthereare blockages inthehy- can becreatedinanimage.This canbeofvaluetoaninvestiga- based onthatrange,thecomplete setofhydraulicfluidpassages with aspecificrangeofattenuation values.Byprocessingtheimage with CTimaging.Hydraulic fluid isrepresentedonCTimages The hydraulicfluidpassages in aservoactuatorcanbevisualized Hydraulic fluidpassages Case studies housing, thelevelofhydraulicfluidcanbedetermined servovalve spoolandsleeve.Lookingcarefullywithinthe piston removed. Thepistonhousingandarevisibleasisthe main servoactuator appearswhenthefronthalfofunitisdigitally tion) arecoloredwhite.TheviewinFigure7showshow the Figure 7:Servoactuatorwithfrontportiondigitallyremoved. ISASI 2003 Proceedings 5 , 6 , • 7 . 143

ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS 144 ered gears rectional controlsystemthatwasdrivenbyelectricmotor pow- tion ofagearsetcanallbedeterminedinCTscan. the gears,absenceofteethinagear,androtationalposi- to examinethedetailsofapart’sgeartrain.Thealignment One ofthebenefitsCTscanning’shighresolutionisability Gear trainexamination a viablerangemuchmoredifficult. ence ofexcessivenoiseintheimagewillmakejobcreating ently givingtheappearanceofablockage.Obviously,pres- age intheimage,andtoosmallarangecouldleadtoinadvert- Too largearangecouldleadtoinadvertentlyincludingtheblock- of valuestouseforhydraulicfluidshouldbecarefullyconstructed. noise isdigitallysubtractedfromtheimage.Inaddition,range done. Low-densityblockagescanbeinadvertentlyremovedwhen investigator mustbeawareofwhatdigitalprocessingisbeing 8, 9,and10. The visualizationofhydraulicfluidpassagesisshowninFigures of fluid)orifthereareanycracksleaksintheservoactuator. draulic passages(whichmayappearintheimagesasanabsence servoactuator. Figure 8:HydraulicpassageintheAirbusA300rudder the screwhousing oftheactuator,and field ofview(yellow Figure 16,the“camera”(yellow arrow)hasbeenplacedinside vestigator tovirtuallyviewthe componentfromanyangle.In bility touseacameraandviewing vectorsystemtoallowthein- CT imageviewingsoftwareused bytheNTSBcontainscapa- tremely difficulttoreach,even ifthepartwasdisassembled.The they allowforobservationsfrom viewpointsthatwouldbeex- spacings downto1mmcanbeseenintheimages. ined. AsshowninFigures13,14,and15,individualteeth with on thegeartrain,individualteethofeachcanbeexam- ponents ofthegeartrainarevisible(seeFigure12).Zooming in 11. Oncethelowdensityhousingisdigitallyremoved, com- The NTSBexaminedascrew-typeactuatorfromtheA300 di- When tryingtolookforhydraulicfluidpassageblockages,the An additionalbenefitofthedigitalnatureCTscansisthat • ISA SI 2003 8 . TheoverallviewoftheactuatorisshowninFigure P r oceedings Figure 9:Close-upview. A300 rudderservoactuator. Figure 10:ViewofanetworkhydraulicpassagesintheAirbus continually keep inmindthateventhough theimagesarepho- eral hoursorevenonetwo daystoperform. schedule. Finally,thescansthemselves cansometimestakesev- tion doingthescanthenhas tofitthecomponentsintotheir must sometimesbetransported longdistances.Theorganiza- nizations withthecapability toperformthesescans,theparts required toacquirethescan.Sincetherearenotverymany orga- gation. Oneoftheprincipaldrawbacksisamount time There aresomedrawbackstousingCTscansinaccidentinvesti- Drawbacks ofCTscans determine ifitisfastenedproperly. tigator todetermineifthefastenerispresent,andpossibly the fastenerontopofscrew.Thisviewpointallows inves- lines spreadingoutfromtheyellowarrow)hasbeenpointed at Once thescansarecomplete, thereviewinginvestigatormust information. Itisimportantthat theinvestigatorusingCTimag- the imagestandout,canalso inadvertentlyfilteroutimportant specific views.Theseenhancements,whilemakingsomeparts of graph, theCTimageshavebeendigitallyenhancedtoprovide tographic innature,theyarenotphotographs.Unlikeaphoto- screw housing. Figure 16:CTimageofendfastnerassemblyfrominsidethe Figures 13,14,15:CTimagesofgeartrainshowing3,2,and1Mtoothspacing. and samewithlow-densityhousingremoved,right. Figure 11,12:CTimageofA300variablestopactuator,left, 8 7 6 5 4 3 2 1 Footnotes priate courseofaction. testing thepartfirstordisassemblingisappro- gather significantlymoreinformationwhentryingtodecideif trains. TheuseofCTimaginghasallowedtheinvestigatorsto as electricallydrivenscrew-typeactuatorswithcomplicatedgear pabilities toexaminehydraulicallydrivenservoactuatorsaswell X-rays orradiographs.NTSBinvestigatorshaveusedtheseca- aging providessignificantbenefitswhencomparedtostandard capability touseCTimaginginaccidentinvestigations.im- The NationalTransportationSafetyBoardhasdevelopedthe Summary digital manipulationsbeingdonetotheimage. ing takethetimetounderstandprocessandbeawareof Alsovisibleinthemiddleandupperportionsofpicturearesomestreaks ThisisthesameA300variable stop actuatorshownintheradiograph TheCTscanofthisactuatorfocusedonlyonthemanifoldhousingarea of Thisactuatorwasremovedfromanaircraftaccidentsceneinsomewhat InFigure7,itemsinredareintendedtodenotelow-densityitems,prima- Thesameeffectcanbeobtainedbykeepingtheobjectstationaryandrotat- TheNTSBhasamemorandumofagreementwiththeU.S.ArmyResearch Radiographisthetermusedtodescribewhatgeneralpublicrefers presented asFigure1. thickness ofthebottompistonhousinganditself. represent noiseintheCTimage,andtheyarearesultofhighmaterial that appeartoextendtheleftandrightofactuatoritself.Thesestreaks due totimeconstraints. the actuator.Theremainingportionsofpistonhousingwerenotscanned ton housingduetoleaksintheactuator. damaged condition.Thehydraulicfluiddoesnotcompletelyfillthe pis- lic fluidcoatingtheservoactuator. layer ofmaterialcoloredred.Thisisduetoathindirtandhydrau- rily hydraulicfluid.Muchoftheservoactuatorhousingiscoveredina thin in medicalCTscanners. ing theX-raysourceanddetectoratsametime.Thistechniqueisused Graphics, GMBH,inHeidelberg,Germany. cially availablesoftwarepackage,VGStudioMax,producedbyVolume equipment. TheimagesarethenprocessedbytheNTSBusingacommer- Laboratory inAberdeen,MarylandtogeneratetheCTimagesusingtheir the radiationthatproducesanimage. between usingthetermX-raytoreferanimageandit public, thetermradiographwillbeusedinthispapertoavoidconfusion as anX-ray.WhilethetermX-raymaybemorefamiliartogeneral N ISASI 2003 Proceedings • 145

ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS ticular seemsto bebecomingprogressively moreconstrained. sources frequentlyareatapremium, andon-scenetimeinpar- opportunity tooccur. the firstplace,lossoflife,injuries, anddamagewillnothavethe the firstcenturyofpoweredflight: ifanaccidentispreventedin in thecourseofaninvestigation haschangedlittlethroughout of theevent.Therationaleformakingthesekindsjudgments other potentiallyfruitfullinesofinquiryrelatedtothe cause oftheundesiredevent,oftendoingsoatexpense of tendency ofairsafetyinvestigatorsistofocusontheproximate tigation mustbesimilarlydirected.However,alltoooften, the vestigative activitytakingplaceinthecontextofabroader inves- similar eventsinthefuture.Byextension,anycontributory in- investigation istoseekmeansofpreventingtherecurrence of The principalpurposeofeveryaircraftaccidentandincident Introduction Aircraft RescueandFireFightingRequirementsWorkingGroup. has servedontheFAAAviationRulemakingAdvisoryCommittee’s cial AviationSafetyTeam(CAST)process.SinceOctober2001,healso programs, andgovernment-industrycooperationthroughtheCommer- of theflightoperationsqualityassuranceandaviationsafetyaction flight andcabincrewoperationaloccupationalsafety,theevolution papers inthefieldofaviationsafety.Hiscurrentresponsibilitiesinclude Association, andistheauthorofnumerousarticlesprofessional on theBoardofDirectorsInternationalAviationFireProtection Farrier isamemberoftheAmericanSocietySafetyEngineers,serves participation intheInternationalSocietyofAirSafetyInvestigators, Committee. FarrierjoinedATAinApril2001.Inadditiontohis member oftheNASAAviationSafetyReportingSystemAdvisory Air TrafficControllersAssociation,duringwhichtimehealsowasa Most recently,heservedasNationalSafetyCoordinatorforthe retirement, hewaselectedtotheU.S.AirForceSafetyHallofFame. among theAmericanAirForces(SICOFAA).Uponhismilitary Committee (PREVAC)oftheSystemforInternationalCooperation served astheU.S.AirForceDelegatetoAccidentPrevention Safety onflight,ground,weapons,andexplosivessafetymatters.Healso the Pentagon,whereheworkedforthreeconsecutiveAirForceChiefsof 146 This reasoningisbynomeans unjustifiable.Investigativere- • Aircraft Accidents:RevisitingthePast ISA SI 2003 Investigating SurvivalFactorsin P awarded theAirMedal.Hisfinaltourofdutywasat over theNorthAtlanticOceanforwhichhewas saved asarescuepilot,includingonenotablemission his militaryflyingcareer,hewascreditedwith17lives commercial, andgeneralaviationsafetyissues.During in aviationsafety,spanningabroadarrayofmilitary, Thomas Farrier r oceedings To LooktotheFuture By ThomasA.Farrier(MO3763) hasmorethan17years’experience Air TransportAssociationofAmerica,Inc. consequences more commonly treatedasstand-aloneareas ofstudy.Theyare broadly definedasembodying threeseparateareasofconcern For thepurposesofthispaper, theterm“survivalfactors”is Survival factorsdefined through survivalfactorsinvestigations theprimacytheydeserve. quent study,andpossiblemeans ofaccordingissuesidentified tors investigations,informationthatshouldbegatheredfor subse- ber ofrecommendationsregardingthegoalsfuturesurvival fac- critical tooccupantsurvival.Thepaperwillconcludewitha num- tors andotherprofessionalstogaininsightintothefactors most offer examplesofsomethemethodsusedbyairsafetyinvestiga- area ofinquirywithintheinvestigativeprocess,anditwill then explore thereasonswhysurvivalfactorsgainedprominence asan review theprioritiesofearlyaircraftaccidentinvestigations; itwill of theeffectsaccidentsnotsuccessfullyprevented.Itwillbriefly consider notonlythepreventionofaccidents,butmitigation tions astheyfirstacknowledged,andthenembracedtheneedto advantages thatmodeoftravelprovided. misfortune—while significant—usuallywasoutweighedbythe relatively hazardousmeansoftransportationwheretherisk that commercialairtravelwas,initsinfancy,understoodtobea War IIaviationoperations.Italsomaybetraceabletothefact of thetrial-and-errorprocessthatcharacterizedmostpre-World disparity mayseemsurprising,itactuallywasanaturaloutgrowth during andintheimmediateaftermathofacrash.Whilethis atic analysisofwhathappenedtocrewmembersandpassengers for morethanfourdecades,therewaslittleinthewayofsystem- matter ofmasteringthebasicprinciplesaviation.Bycontrast, est effortsatpoweredflight,andfailureshadtobeassessedasa Accident investigationswereessentialadjunctstotheveryearli- the samedegreeoffocusedattentionasaccidentsthemselves. casualties resultingfromaircraftaccidentshaverarelyreceived variety ofreasonswhyinvestigationsintothespecificcauses as ancillarytothemainlineofinquiry. contributed tothetreatmentofsomeinvestigation-relatedtasks mate goalofpreventionfarmoredifficult,andalsohas tigation moreresource-intensive.Thistrendhasmadetheulti- sality leadinguptoeachofthemhastendedmakeinves- rarer events,therandomnessandcomplexityofchaincau- process. Further,asmajoraccidentsbecome(thankfully)ever- tions distinctlysecondarystatuswithintheoverallinvestigative standing patternofaccordingcertainaspectsmajorinvestiga- Nevertheless, concentratingoncausetendstoperpetuatealong- * , DirectorSafetyPrograms, This paperchartstheevolutionofaircraftaccidentinvestiga- Throughout thefirstcenturyofaviation,therehavebeena must remainalert tothepossibilitythatpost-crash survivalissues will notbeexpandeduponin thispaper.However,investigators area moot;accordingly,thisaspect ofsurvivalfactorsinvestigation ing” seemtohaverenderedmost ofthehistoricalconcernsinthis rafts andotherflotationaids“in theunlikelyeventofawaterland- ability ofenginesandcommunications, aswelltheavailabilityof availability andeffectivenessof emergencyresponders.Thereli- vival inthecontextofinvestigationsusuallyrevolvesaround the dedicated raftsandcombinationevacuationslide/raftprovisions. current requirementsforpersonalflotationdevicestoaugment passenger jetsinthe1960s,whichdirectlyledtomost of the also wasreinforcedbyafewisolatedinstancesofditchings of of enginesandthesparsenesscommunicationscoverage. It have beentheresultoftwoprincipalconcerns:unreliability nificant periodoftimeinthemid-20thcentury.Thisseems to among aircraftdesigners,regulatorsandinvestigatorsfor asig- planned orunplannedditchings,wasamajorpreoccupation 3. the peopleexposedtothem. and thermalconditionswithpositivenegativeoutcomesfor tion oftheaircraftanddeteriorationinterioratmospheric be marshaledinsuchawayastoallowcorrelationofthecondi- occupant—both survivorsandcasualties—ineachaccidentmust tors inamoreholisticmanner.Thespecificfactorsaffectingeach However, thereisaneedtodocumentthisaspectofsurvivalfac- nity ports, theyusuallyhavebeenexaminedintermsofthe the issueofevacuationsincontextindividualaccidentre- so mustbedocumented,evaluated,andcorrected. must escapesuccessfully;anythingthatpreventsthemfromdoing least alimitedamountoftime.Occupantswhosurvivecrash ment ofaircraftsubjectedtocrashforcesremainssurvivableforat has beenaconcomitantneedtoensurethattheinteriorenviron- sistance. Asthesurvivabilityofaircraftaccidentshasincreased,there post-crash environment,eitherundertheirownpowerorwithas- the abilityofoccupantstomoveclearpotentiallyhazardous factors inaircraftaccidentsandtheirsubsequentinvestigationis 2. in anaccident. stage processthatresultsinthe vivability” asanenduntoitself,insteadofthefirstpartathree- listed. Second,bylimitingittotheelementsabove,treats“sur- and fixtheattentionofinvestigatorstospecificparameters tion istwofold.First,overtime,thismodelhastendedtodraw sess eachaccidentandrendermeaningfulrecommendations. ume). (i.e., preventionofejectionandpreservationoccupiablevol- and seatattachments);(3)protectionoftheoccupiablearea the humanbody;(2)integrityofrestraintsystems(belts,seats, decelerative forcesnotexceedingtheknowntolerablelimitsof ton, D.C.)identifiesthethreedeterminantsofsurvivabilityas(1) Florida Boeing737thatcrashedjustaftertakeofffromWashing- ample, itsreportonthe“Palm90”accidentof1982(theAir ability initsfinalreportsonmajoraircarrieraccidents.Forex- (NTSB) frequentlyreiteratesitscriteriafordeterminingsurviv- 1. Evacuation andextrication The present-daycodificationoftheconcernforpost-crash sur- To theseextentthatsurvivalfactorsinvestigationshaveexplored However, theproblemwithsuchatightlyconstraineddefini- Survivability Post-crash survival toescapebasedonphysicalobstructionsandelapsedtime. 1 Theseprovideaconsistentbaselineagainstwhichtoas- . TheU.S.NationalTransportationSafetyBoard . Post-crashsurvival,especiallyfollowing . Thesecondcomponentofsurvival actual survivalofpeopleinvolved opportu- 2 “survivability,” asdefinedby theNTSB,isanassessment that host ofequallyimportantconsiderations. Inpart,thisisbecause eas aloneresultsinreducedattention, ornoattentionatall,toa creased urgencythenewaccident suggeststheywarrant. previously expressedrecommendations withreferencetothein- served intheaccidentunder investigation; alternately,reiterate • accidents and/oraddressedinpriorrecommendations. • uted tocasualties. passenger performance,andemergencyresponsethatcontrib- • or passengers. by severityandwhetherthepersonsinvolvedwerecrewmembers throughout theaccidentsequence;tabulatefatalitiesandinjuries • gations, seemstodistilldownafewbroadgoals: reports andpublishedguidanceontheconductofmajorinvesti- thinking astotheseobjectives,evidencedbythecontentof vestigation mustaccomplishmultipleobjectives.Thecurrent investigators tocontinueworkdispelsuchmisconceptions. that aircraftaccidentsareuniformlyfatal; In a2001report,theNTSBtooknoteofpublicperception are essentialelementsofpublicconfidenceinairtransportation. over, successfulevacuations,bothprecautionaryandpost-crash, attempting toreducethetimerequiredforevacuation.” better improvedbyextendingtheperiodofsurvivabilitythan attained foroccupantsafetyasfollows:“Passengermaybe United StatesCongresssummarizedthedesiredendstatetobe ing preparedbytheOfficeofTechnologyAssessment able environmentwithintheaircraftfollowingeachcrash. crash, concernnecessarilyturnedtothepreservationofasurviv- death wasnolongeraninstantaneousandlikelyconsequenceofa able fromthemomentofimpacttopost-crashperiod.Once mulative effecthasbeentoshiftwhenlossoflifeismostprevent- overall survivalexperiencefollowingaircraftaccidents,theircu- • • and furnishings; • aircraft; • • • • • nation byarangeofdesignandretrofitchangesaimedat these conditionshavebeenaddressedindividuallyandincombi- impact anddecelerativeforces,inhalationhazards, principal threatstooccupantsurvivalinanaircraftaccidentare Throughout thispaper,readersshouldbearinmindthatthe The objectivesofsurvivalfactorsinvestigation should bepreparedtoaddressthemasnecessary. could ariseinthecontextofvirtuallyanyoff-airportaccidentand Experience suggests,however, thatconcentratingonthesear- The foregoingdiscussionshowsthateachsurvivalfactorsin- A 1993“backgroundpaper”onthesubjectofevacuationtest- While alloftheseimprovementshavecontributedgreatlytothe Mak Determine whichoftheabovehavebeenobservedinother Identify failur Document whathappenedtotheair r r slowing flame-fr pr impr attenuating crashfor impr r educing radiantheatemissivityoffurnishings. educing sour educing potentialsour eserving a“survivablevolume”withinoccupiedportionsof e newr oving seatr oving physicalcontainmentofinflightfir ecommendations specifictothecir es ofair ces oftoxicby ont movementthr etention; ces; craft equipmentandstructur ces offuelandfir -pr oducts ofcombustion;and ough onboar craft anditsoccupants ISASI 2003 4 itisincumbentupon e sustainment; and es; cumstances ob- fire d components Proceedings es, cr . Overtime, 3 ew and More- • 147

ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS • • of theaircraft’soccupants: four possibleoutcomeswithrespecttosurvivabilityandthefate extraordinary numberofvariables.Foreveryaccident,thereare relies uponspecificcriteriaappliedtoanenvironmentwith 148 • vived; and • killed; was seenasachievable throughthesuccessful resolutionofthese essential toconsistentlysafe operations. Inotherwords,safety niques andmoreeffectivecontrol ofaerodynamicforceswere the aviationcommunityof daythatbetterconstructiontech- pronouncement doesdemonstrate theclearunderstandingof trol surfaces—asubjectnotwell understoodinhisday—Spooner’s use oftheterm“control”referred totheoperationofflightcon- failure ofmaterialsandtheothercontrol.”While his accidents seemattributabletotwocauses,oneofwhich is the observation onthenatureofaccidentsthemselves:“Ingeneral, editor StanleySpoonermadeamoreinterestingandspecific and consequencesofaccidents. dent prevention,withlittlethoughtgivento ters alreadywasbeginningtosettlealmostexclusivelyon acci- safety thoughtinoneoftheworld’sgreatforumsforsuch mat- into thefirstcenturyofflight,focusstructuredaviation evil consequencesthereof.”Inotherwords,lessthan10years must beavoided,sincetherecannosatisfactoryremedytothe have resultedindeath….Itisevidentthatittheaccident Accidents continuetohappen,andoflatemanytheaccidents of thesciencegreaterimportancethanthisquestionsafety. comment onpageB-2:“Itisself-evidentthattherenobranch “the preventionofaccidents,”containedthefollowingeditorial The Jan.11,1913,issueofthismagazine,whichwasdedicatedto publishing as quicklypossible. attention ofpioneeringaviatorsontheneedtomastertheircraft but italsohadapracticalaspecttoaswell:servedfocusthe tude hasbeenremarkeduponbyvariouscommentatorsovertime, ral by-productsofaircraftaccidents.Thissomewhatfatalisticatti- throughout theaviationcommunitywasthatfatalitieswerenatu- man lifeassociatedwithaircraftaccidents,theprevailingsense reality, whiletherewasongoingconcernforthehazardstohu- Almost fromthemomentthatheavier-than-airflightbecamea consideration ininvestigations The evolutionofsurvivalfactorsasa blunt forceorthermalinjuries. furnishings thatperformedasdesiredwithrespecttolimiting lives savedbydesigned-inprotections,emergencyequipment,or when itcomestoobjectivedataregardingthesuccessstories: ate properly,andsoforth.Therecordfrequentlyisquitethin seats thatfailedtobeproperlyretained,exitsoper- the inclinationofinvestigatorsistodocumentnegatives— trying topreservethelivesofanaircraft’soccupants.Toooften, sight intosurvivalfactors,thatis,whatworkedanddidn’tin All four However, laterinthissameessay,RoyalAeroClubfounder/ In 1909,theRoyalAeroClubofUnitedKingdombegan a survivableaccidentinwhichoneormor a survivableaccidentinwhichalloccupantssurvived; a nonsurvivableaccidentinwhichalloccupantswer a nonsurvivableaccidentinwhichoneormor • ISA SI 2003 oftheabovescenariosofferopportunitiestogainin- Flight P r oceedings , billedasthe“firstaeroweeklyinworld.” mitigation e occupantswer e occupantssur- oftheforces e killed. e chairs. Thecruising speedwas90milesan hour….Theaggre- occasionally two.Thecabin contained fromfourto10wicker a cat’s-cradleofexternalbracing. Itusuallyhadoneengine,but cal ofanythingthatwasinso tentative astate,wasbiplanewith “The typicalairtransportof1920, ifonemayusethewordtypi- wich [Vermont]University,Dr.Edward PearsonWarnerobserved, the ground—importantforflyinginreducedvisibility.” rizon, whichshowedpilotstheattitudeofaircraftrelative to climb indicators,compasses,andtheintroductionofartificial ho- also improved,withbetteraltimeters,airspeedindicators,rate-of- and makinglargerfasterplanespossible.Cockpitinstruments Air-cooled enginesreplacedwater-cooledengines,reducingweight many believeitwasthemostinnovativeperiodinaviationhistory. “There weresomanyimprovementstoaircraftinthe1930s that improvements. turn providedtheeconomicimpetusforawholerangeofsafety were aperiodofsignificantgrowthinaircarrierservice,which ing theperiodicseriousaccidentsthatmadeheadlines,1930s dozen wereinvolvedinfatalaccidentseveryyear. the 400to600aircraftincommercialservice,morethanahalf- miles flowngrewsteadilyeveryyearduringthissameperiod,of mercial airtransportwasundeniablygettingsaferandpassenger more thanonein10ofthoseaccidentswerefatal.Whilecom- more than1.5millionin1937.Still,evenduringthebestyears, carrier servicegrewsteadily,fromjustover100,000in1928to “promoting airsafety.” sometimes conflictingtasksof“promotingaircommerce”and was thefirstlegislationtoassignanaviationregulatorybody to licensepilotsandaircraft,investigateaccidents. merce todesignateairroutes,developnavigationsystems, rier operations.Inparticular,itauthorizedtheSecretaryofCom- Department withrespecttotheconductofcommercialaircar- ger servicebyassigninganumberofdutiestotheCommerce law, takingofficialnoteofthegrowthcommercialairpassen- lowing yearthattheAirCommerceActof1926wassignedinto mercial aviationintheUnitedStates,butitwasn’tuntilfol- of thepayloadincramped,noisyandcoldcabins.” rail androadconnections.Forseveralyears,passengerswerepart mail deliverydonebybushpilots,takingpeoplebeyondtheirusual kind ofcargo.Passengerservicewasanoffshootthefreightand the earlyaviationdays,passengersinairplaneswerejustanother port Canadapublicationonaviationhistoryaptlydescribesit,“In regularly scheduledpassengerservice. began toconsiderthepossibilitythatairplanesmightbeusefulin after theArmisticethatbothgovernmentsandprivateconcerns Tampa AirboatLine,whichoperatedbrieflyin1914.Itwasonly passenger serviceisgenerallythoughttobetheSt.Petersburg- passenger service.Theworld’sfirstscheduledheavier-than-air little thoughtwasgiventotheuseofheavier-than-airaircraftin paying passengersinthesummerof1910 early yearsofaviation.Infact,althoughZeppelinsbeganserving safety receivedmuchinthewayofconsciousconsideration (like occupantfatalities)wouldbesolved. two keyconcerns;fixthem,andallotherundesirableoutcomes In 1938,duringaspeechto the studentsandfacultyofNor- According totheAirTransportAssociation’s From 1928through1937,the“milesflownperaccident”inair The AirMailActof1925gaveamuch-neededboosttocom- This isnottosaythatthe“passenger”dimensionofaviation 10 6 Ofcourse,asarecentTrans- 5 , priortoWorldWarI Airline Handbook 9 Notwithstand- 7 11 8 This , • the endofWorldWarI: technical developmentofaircraftinthetwentyyearsfollowing ing hisremarks,preferringtocitefivemajoradvancesinthe • in reliabilityandsafety; • they aresubstantiallymorecomfortable.” rugged asthoseofthePullmancar,andbypopularappraisal well setintothewing’sthickness.Its10to40seatsappearas plane, itslandinggearretractedinflight,twoorfourengines gate of19yearsdevelopmenthascreatedacantilevermono- aircraft, “Considerations ofsafetyhavenecessarily beenlimitedin referring to“thedesignproblem ahead,”notedthat,inmilitary Forces reporttotheChamber ofCommercetheUnitedStates, fronts almostassoonWorld WarIIwasover.A1945ArmyAir senger crashprotectionbegan togainattentiononavarietyof from ahostofmoreurgentsafety priorities,thechallengesofpas- gers gofromplacetoinsafety.” little ways,theairlinestrytoroundouttheirdesireseepassen- plies areavailablewhenneeded.Intheseandhundredsof other firm arecarefullyassistedtoandfromtheirseats.Medical sup- are substantiallybuiltandcoveredwithslip-prooftread.The in- against sharpcorners.Rampsleadingfromthedoortoground signed sopassengerswillnotinadvertentlyknockthemselves weather andduringlandingortakeoff.Airplanefurniture isde- belts areprovidedtokeeppassengersintheirseatsbumpy utmost toseethatpassengersarenotinjuredinotherways. Safety sengers aretransportedinasafevehicle,theairlinesalsodotheir servations regardingpassengersafety:“Besidesseeingthatpas- which investmentsinsafetyshouldbemade. simply recognizedthattherewerefarmorepressingneedsagainst This isnottosayhewasunawareofthepotentialhazards; much inthewayofresourcestowardsurvivalfactorconcerns. engine reliabilityandinstrumentationtorecommenddevoting reer, wasfarmorepreoccupiedwiththenecessitiesofimproved future aviationsafetyrequirementsthroughouthisillustriousca- Even Lederer,whogavetheworlditsmostclear-eyedvisionsof private sector,spoketothesameaudienceatNorwichUniversity. cably transmittedcrashenergiesdirectlytotheiroccupants! flew unrestrainedthroughcabinsduringcrashes,orastheyimpla- fortable piecesoffurnitureundoubtedlywereequallysolidasthey and seatanglesseparatelyadjustable.” lightness gavewaytosolid-lookingpiecesoffurniturewithback ment [since1920].Chairsthathadnoothervirtuethanof structed passenger,weretheimprovementsininteriorarrange- and structure,butevenmoreobviousbeguilingtotheunin- to theaeronauticalengineerthanchangesinexternalform tures ratherthansafety-relateddesignelements:“Lessinteresting regarding theroleofinterioraircraftfurnishingsasmarketingfea- • • in performance; gines, withtheirattendantreductionindragandimprovement Dr. Warnermadenomentionofsurvivalfactorsconcernsdur- Despite whatamountsto40years ofbenignneglectresulting During hisNorwichaddress,Lederermadethefollowingob- In 1939,JeromeLederer,whoatthetimewasworkingin In passing,however,Dr.Warnermakesarevealingobservation understanding andincr the developmentof“totalenclosur the adventofmultiengineair the eliminationofneedfore the impr ovement ofaviationfuel;and easing wingloading; craft, withtheirattendantincr xternal bracingforfuselages. 13 14 e” cowlsforair-cooleden- Manyofthosesolid,com- 12 ease but tookthenovel additionalstepofallowing forthepossibility focusing oninvestigatingfor thecausesofaccidentsasawhole, Investigation tion ofDocument6920-AN/855, the the investigationarenain1949, withthereleaseoffirstedi- accident experienceasaprincipal sourceofsupport. scientifically basedapproachtooccupantprotectionthatdrew upon lar documentsclearlyshowtherapidemergenceofadeliberate, transport categoryaircraftdesignrequirements. by theCAB’sBureauofSafetyRegulation,withrespectto future Board’s BureauofSafetyInvestigation,intendedforconsideration series ofrecommendationsdevelopedbytheCivilAeronautics cific concern.InJuneof1950,itgavewidedissemination toa ning seriesofbulletinsthatincludedseveraldevotedtothis spe- tion andAirportFireProtection”beganpublishingalong-run- National FireProtectionAssociation(NFPA)“Committeeon Avia- bution ofdecelerativeforcesinrelationtoinjuries.” way inwhichdesignfeaturesinfluencethemagnitudeanddistri- DeHaven, anengineerworkingwithmedicalscientists,onthe collaboration towardthisendisaffordedbythestudiesofHugh to besubstantiallyincreased.Anexcellentexampleofdesirable tribution mustbedeterminedifthesurvivalratesincrashesare ing crashes:“Boththemaximumforcesandtheiroptimumdis- discussing thepressingproblemofimprovingsurvivalratesdur- “Crash InjuryResearch”program(ofwhichmorepresently)in • • • standing highdecelerations; • ditching; • to advanceanalysiswithrespectoccupantsurvivalfactors: aircraft itself.” air crewsmaymakeand“everypossiblefaultofthedesign macy, embodiedtwosimultaneousconsiderations:mistakesthat vance analysis.”Thelatter,whichhefeltshouldbegivenpri- dent analysis:post-accidentstudies,“nearaccidents,”and“ad- ing survivalfactorstothepresentday. been followedbyresearchersandinvestigatorsalikeinexamin- Air TransportDesign tions ofhumanbeings.Hisseminal1946work, the inaccessibilityofconcretedataoncapacitiesandlimita- of integrationbetweenaircraftoperatorsandmanufacturers, identified twokeydisconnectsinaviationsafetythinking:alack cal effectsofaviationonhumans.Overthenext20years,he since thelate20shadbeeninterestedinsomeofphysiologi- neering workofHarvardUniversity’sDr.RossA.McFarland,who all builtintotheairplanetomaximumdegreepossible.” tant progresstowardflyingsafetycanbemadeuntilisfirstof ing, however,][f]orthefuture,itisinconceivablethatanyimpor- armor, escapehatches,safetybelts,andthelike….[Incivilianfly- tural strength,adequatepower,protectionfromfire,protective most casestosuchbasicrequirementsassafewingloading,struc- The InternationalCivilAviation Organization(ICAO)entered Working onaparallelbutcomplementarypath,in1946the Dr. McFarlandalsogavedueregardtotheworkofCornell’s Dr. McFarlandidentifiedthreedistinctclassificationsofacci- The immediatepostwareraalsobroughttotheforepio- the r fir the meansofescapeinemer the str the locationofwingsandstr e-pr eliability ofthehydraulicandelectricalsystems. essing ofchairs[sic],safetybeltsandharnessesforwith- evention measur . Thisinitialeffortfollowed the generalpatternof 16 McFarlandsawfiveissuesasbeinguniquelysuited , helpedchartthescientificcoursethathas es; and gency landings; engthening ofthefuselagefor Manual ofAircraftAccident ISASI 2003 19 Human Factorsin Thisandsimi- Proceedings 18 17 • 15 149

ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS • • limiting factorsfacedbyoccupantsofcrashedaircraft,including: vival inAircraftCrashFires,”wasthefirstcleararticulationof nar” thatsamemonth.Thefirstofthese,entitled“HumanSur- noteworthy paperspresentedatits“ThirdAnnualAviationSemi- accident. Forexample,inMayof1954,theNFPApublishedtwo given tothechallengesofemergencyevacuationfollowingan mortem injuryanalysis. to berecognizedas“aviationpathology”protocolsforpost and causes,alsoofferedthefirstdiscussionofwhathavecome vided evenmoreelaborateformsfordocumentinginjurysources vival issues,shouldberecorded. causes ofpassengerdeathandinjuries,aswellpost-crashsur- Injury Research”group,andexpandedsignificantlyonhowthe Injury ReportForms”developedbyCornellUniversity’s“Crash factors issues.Itincludedthefirstversionof“AircraftCrash resented asignificantimprovementinitstreatmentofsurvival devices, etc.” ing fromfaultycockpitorcabindesignandpersonalrestraining tensive researchinprogresstoeliminateinjuriesordeathresult- aircraft accidentsareofgreatvalueinprovidingdatafortheex- tors: “Detailsrelatingtoinjuriessustainedbypersonsinvolvedin effective informationcollectioneffortwithrespecttosurvivalfac- considered causalinitsownright.Italsosetthestageforahighly of materielfailure,orevenafailure“safetyequipment,”tobe 150 FAA’s Officeof AviationSafetyidentified“recurring persistent focus regulatoryefforts.Forexample,in1976areport bythe years seemtohavebeenwritten inanefforttofurtherrefineand whole seriesofstudiesconducted inthemostrecent20to30 sues needingtobeaddressed intheareaofoccupantsafety,a in theregulationandoversight ofallmodestransportation. 1965 book public andpoliticalresponsetoconsumeradvocateRalphNader’s into avirtualmoviefranchise.However,italsoisevidentinthe such offeringsasArthurHailey’snovel ards. Thelattermaybediscernedinthepopularmedia,through rier service,andincreasedpublicsensitivityto“preventable” haz- ing fleetmixfrompropellertojetaircraftincommercialair car- were tworeasonsbehindthisincreasedinterest:thesteadily shift- ment researchstudiesthroughoutthe1960sand1970s. There was developed. as forward-lookingitwasusefulinthecontextwithinwhich present-day accidentsandincidents lenges regardingescapeslideavailabilitycontinuetocropupin and theirrespectiveimpactonevacuationtime.Giventhatchal- dow versusdoorexits,aswelldifferentconfigurationsofslides, includes adetailedanalysisoftherelativeeffectivenesswin- on thesubjectofevacuationunderfireconditions. the CivilAeronauticsAdministration’sOfficeofAviationSafety, ture ofthedayinthisareawasamonographbyBarryG.King • tion; and • Also inthe1950s,seriousprofessionalattentionbegantobe The secondeditionofthisdocument,publishedin1951,rep- Notwithstanding theapparent clearunderstandingoftheis- Survival factorsconcernsbecamemoreprominentingovern- The secondvaluablecontributiontotheprofessionallitera- r skin burning; accumulation ofcarbondioxide. e espiratory systemdamage; • xposur ISA SI 2003 e tocarbonmonoxidecombinedwithoxygendeple- Unsafe atAnySpeed 20 P r oceedings 22 , whichfueledsignificantinterest 21 By1959,thethirdeditionpro- 23 25 , thiskindofthinkingwas Airport anditsspin-off 24 Thisessay crash, Feb.11,1952. Figure 1.CrashkinematicsofNationalAirlinesDC-6 • • • • • • • crews inemergencyprocedurestraining) • vests) • • • • • cabin safetyproblems:” accident investigations. reinvestigating afterthefactor workinginparallelwithongoing of knowledgefromtheinvestigation reports,oftenby that seemedtoassemblethe most usefulandaccessiblenuggets studies conductedbygovernment agenciesandprivateconcerns inroads intotheknownproblems. Moreoften,itwasthespecial tion reportsseemedtocapture allofthedatanecessarytomake ever, itwasonlyinrarecasesthatindividualaccidentinvestiga- tinely reinforcedallofthereportsandstudiescitedabove. How- Investigations ofaccidentsthroughoutthe50s,60s,and70s rou- Model documentationanddatavisualizationefforts source thatisexploredinthebalanceofthispaper. was createdthathassurvivedtothepresentday.Itisre- imaginative approachestoissueidentificationanddescription defended, andintheprocess,alegacyofcreativethinking and cal solutionstomanyoftheseproblemsweredocumented and tailed researchrequirementsandbettersupportfortechnologi- and worthyofconcertedstudyinthe30yearsthatfollowed.De- survival factorsassertedthemselvesasbothcriticallyimportant full. Theauthoroptsforthelatter;after40yearsofinattention, one caneithermakeacasefortheglassbeinghalf-emptyorhalf- When lookingatthisplethoraofapparentlyrepetitivestudies, timing ofpassengerservice alcoholic beverageservice car cr handicapped passengers communications (especiallycabin-to-flightdeckandviceversa) galley equipment cr emer flight attendantseating emer cabin interiors cabin fir ew training(especiallytheseparationbetweenflightandcabin ew dutyday ry -on baggage gency equipmentavailability(ax gency equipmentlocations e, smok e andtoxicgases 26 es, over water rafts,life “analysis” and thus exemptfrompublicdisclosure. of thatinformationundoubtedly havebeenfoundtobeworksof structure toit,andmanyconsolidated orindexedcompilations able onthepublicrecordrarely hasmorethanarudimentary “public docket.”Unfortunatel y, thereamsofinformationavail- of databyplacingasmuchpurely factualdataaspossibleintothe National TransportationSafetyBoard,try toorganizethiswealth accident investigation.Someinvestigationauthorities,such asthe formation regardingsurvivalfactorsinthecourseofanaviation Figure 2.Injuriescorrelatedtostructuraldamage. It iseasytoassembleahuge,virtuallyindigestiblebodyof in- garding aircraft designrequirementsforoccupant survivability. forthright, astonishinglyprescient setofrecommendationsre- makes tothebodyofliterature surroundingsurvivalfactorsisits forces appliedtothefuselage during thecrashsequence. occupants, theareasofcabin floordestruction,andtheexternal pre-crash position),theextent ofinjuriessustainedbythatseat’s actual damagesufferedbyeachseat(withtheseatsplacedin their vided intheprecedingthreediagrams,showingdepictions ofthe • of passengerseats. mid-crash contactwithatree,whichinturndislodgednumber • with atreeduringthecrashsequence. on thispageshowstheeffectsofaircraft’sbroadsidecollision occupants tostructuraldamage;thesmalleroftwodiagrams • from flighttobeingfullyatrest. its occupantsencounteredobjectsthatchangedtheirenergystate cident, thatis,thevariousstagesduringwhichairframeand • vided onthefollowingpages: extremely accessibleform. diagrammatically presentanumberofcomplexconceptsinan but notelaboratelyillustrated,usingavarietyoftechniquesto equally eminentdirector,thelateHughDeHaven.Itiscopiously der thesupervisionofCrashInjuryResearchprogram’s presented bytheAerospaceMedicalAssociationisnamed),un- prepared byA.HowardHasbrook(forwhomapresent-dayaward Airlines DC-6AccidentatElizabeth,N.J.,onFeb.11,1952. and wellaheadofitstime. of survivalfactorsthatwas,withoutquestion,bothgroundbreaking Injury Research”(AvCIR)initiative.Thiseffortresultedinastudy pices ofCornellUniversityMedicalCollege’s“AviationCrash tive art,aparallelinvestigationwasconductedundertheaus- as wellintheinterestsofadvancingstateinvestiga- routine nighttimecrash. particular accidentbeingtreatedassomethingfarmorethana is sometimesthecasewithaircraftaccidents,politicsledtothis Aeronautics Board’sBureauofSafetyInvestigation.However,as than halfofitsoccupants.ThiswasablyinvestigatedbytheCivil after takeofffromNewark,N.J.,crashingwiththelossofmore In thewinterof1952,aNationalAirlinesDC-6crashedshortly “Informative AccidentRelease#15” tive authorities. pretation, andtouseinapproachestheirrespectiveinvestiga- alternate meansofarrayingcomplexdatasetsforreadyinter- studies areofferedasameansofinspiringreaderstoconsider terpretation. Tothisend,thefollowingexamplesofhistorical during survivalfactorsinquiriesremainsoneofretrievalandin- tors andresearcherswhomaywishtobuildupontheworkdone be usedasbenchmarks?Ultimately,thechallengetoinvestiga- should suchreformbepursued,andwhatgoodexamples mation managementthatneedsreform.Thequestionis,how arguments forspecificsafetyinitiatives,thisisapatternofinfor- The mostremarkablecontribution the Several oftheillustrationsfromthissuperbreportarepro- The reportinquestionisthe Given theneedforharddatainmakingincreasinglydifficult F F F F igur igur igur igur e 4bringstogetheressentiallyalloftheinformationpr e 2r e 3showshowtheair e 1illustratesthecrash“kinematics”ofElizabethac- elates theseverityofinjuriese 27 Inresponsetothepoliticalpressure, craft cabinwascompr Crash SurvivalStudy:National ISASI 2003 xperienced byair Crash SurvivalStudy Proceedings omised by 28 Itwas craft • 151 o-

ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS 152 the diagrams that accompanythem. tions… andtherecommendations aregraphicallysupportedby cific teststoverifyproperseat performanceundercrashcondi- tention andenergyabsorption, andtoperformavarietyofspe- and floortomaximizeenergy absorption,todesignseatsforre- They citetheneedtoensure properintegrationofthefuselage front propagation,andexitavailability. Figure 5.Depictionofwindinfluenceonsmoke,flame appearance ofseats,andimpactforcevectors. Figure 4.“Compilation”diagramblendinginjuries,actual Figure 3.Dynamicsofaircraftimpactwithtree. • ISA SI 2003 P r oceedings 29 Beyondthat, the Study injured occupantsandblockedexitsontheright). conditions (smokeandflamepropagationontheleft, Figure 7.Twoexamplesrelatingegressroutestoother Figure 6.Flamepropagationversuselapsedtime. products ofcombustion withinthecabinenvironment. Aswith fire andsmokepropagation, andtheminimizationoftoxicby- of occupantmobility,theavailability ofusableexits,thedelay A successfulevacuationgenerally maybeconsideredafunction Assessing successfulandunsuccessful evacuations ture—survival factorsinvestigation reportsshouldbemeasured. establishes thebaselineagainstwhichallsubsequent—and fu- • • • • • • • • • • • • • • menu ofconsiderations: conducted 4yearsearlier,whichofferedasubstantiallybroader However, theseweresomewhatatvariancewithanFAAstudy • • • • • • • • • a bearingonthesuccessorfailureofanemergencyevacuation. NTSB issuedacomprehensivestudy passenger capacity,andcabincompartmentation. reevaluate aislewidth,numberandlocationofemergencyexits, placarding requirements,post-crashfirecontrol,andtheneedto dations regardingbefore-landinginstructionstopassengers,exit “whys” ofevacuationsuccessesandfailures,offeringrecommen- pants. Thisreportwasthefirsttooffersignificantinsightsinto definitions ofthatterm,yetresultinginthedeaths16occu- ideal forthepurpose,beingcompletely“survivable”bystandard the previoussummer. Airport followinganemergencylandingandrunwaydeparture United AirlinesDC-8atDenver,Colorado’sStapletonInternational completed anin-depthreviewofthepost-crashevacuationa dents whereevacuationissuescameintoplay. stead ofbeingapartthemaininvestigationreportsonacci- evacuations hasbeenaccomplishedbystand-alonestudiesin- the mostvaluableanalyticalworkonsuccessfulandunsuccessful the crashsurvivalstudiesdescribedearlierinthispaper,muchof • • • • • The themesabovewereexpandeduponin1974,whenthe In Mayof1962,theU.S.FederalAviationAdministration(FAA) Physical e Size, number P Secondary e P P P Physical condition W External illumination Cr Cr Condition/availability ofslides Aisle width Seating density Cr Cr P Obstructions toegr Emer Emer Condition/availability ofslides Air W Age Gender T External illumination r r assenger e assenger pr r er eather eather esence offir esence oftoxicby esence ofheat ew e ew training ew pr ew training craft attitude rain gency communications gency lighting xperience ocedur xit cues xperience xplosion(s) , andlocationofe epar e and/orsmok es edness 30 ess Theaccidentsubjectedtothisanalysiswas -pr oducts e Biobehavioral Factors Environmental Factors Procedural Factors Configurational Factors “Man” (HumanPerformance)Factors “Machine” (Aircraft)Factors “Environmental” Factors xits 31 identifying12factorswith ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) 32 • Denver DC-8crashofJuly1961. the aircraftaffectedoccupants’ abilitytoevacuatefromthe • later study: such informationcanbeassembled inameaningfulmannerfor 10 onpage152andthefollowing pagesshowexamplesofhow routes, andotherrelevantsurvivalfactorsdata.Figures5through formation tosupplementnarrativedescriptionsofinjuries, egress depiction ofusefulconfiguration,orientation,andmortality in- cited abovealsoservesasarichsourceofideasregarding the prominence. lists, thatrepetitionisnotnecessarilyanindicationofparticular 3. Whileanumberofissuesshowupinmorethanone these zational schema. in isolation,relyinguponitsownpreferredmodelsandorgani- 2. Eachgroupofexpertsseemstohaveworkedatleastsomewhat tered. have assigneddifferentprioritiestothevariousfactorsencoun- 1. Differentgroupsofexpertslookingatevacuationsovertime males), age,andassignedseat. occupant demographics:gender( Figure 8.Bodylocationswithinthecabin,correlatedwith As inthecaseof1952AvCIRreport,1970FAAstudy What dotheforegoingliststellus? F F igur igur e 5showshowthepr e 6showshow fuel fr 33 evailing windandtheorientation of om aruptur …for females, ISASI 2003 ed tankspr Proceedings ead andig- …for • 153

ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS positions forunsuccessfulevacuations. Figure 9.Alternatemeansofdepictingstartingandending 154 The artandscience ofaircraftaccidentinvestigation hasadvanced Conclusions andrecommendations tation anddatabasemanagement efforts. to marshalinginformationcould belenttomodern-daypresen- accessibility, andthemeansby whichtheirrespectiveapproaches selected forpresentationabove liesintheirinnovativeness, bibliography attheendofthispaper.Theappealexamples reports wereoriginallyissued;someofthebestarelisted in the of bothsurvivabilityandevacuationconsiderationssince these sual presentationoftheresultsanalysis. deal toremindmodern-dayinvestigatorsaboutthevalue ofvi- course ofeverysurvivalfactorsinvestigation,andallhavea great studies citedabovehavemerit, quence andsuccessfulevacuationsbyexit. discrete eventsinatestreplicatingpost-crashevacuationse- tries Associationchartingtheamountoftimeelapsedbetween • ment fromassignedseattobodypositionwithintheaircraft. • demographic informationabouteachcasualty. the locationsoffatalities,withanaccompanyingkeyproviding • post-crash environment. of aircraftoccupantscanbecorrelatedwithotheraspectsthe November 1964showinghowinformationaboutthecondition crash ofNovember1965andtheRomeBoeing707accident • various timeintervalsfromtheinitialbreach. nited bydepictingtheextenttowhichflameshadspreadat A greatdealofusefulresearchhasbeenconductedintheareas The bottomlineisthat F F F F • igur igur igur igur ISA e 9showsanalternatemeansofshowingoccupantmove- e 10showsatimelinecr e 8isan e 7offers SI 2003 P r oceedings xample ofanair xamples fr all ofthefactorscitedbydifferent om theSaltLak all eated bytheAer shouldbeaccountedforinthe craft layoutdiagramshowing e CityBoeing727 ospace Indus- work asdesigned orexpected.NTSBrecords arerepletewith not present.However,itisequally importanttoidentifywhat document whatpreventiveand protectivemeasuresfailedorwere dent inwhichtherearenonetheless survivors,itisessentialto In anysurvivableaccident,as wellasany“nonsurvivable”acci- outsourcing toensurecompleteness. part ofeverymajorinvestigation, evenifdoingsorequires Recommendation 1.Makesurvivalfactorsamoreprominent advantage offortheforeseeablefuture. the issuesofsurvivabilityandevacuationwillneedtobe taken In otherwords,everyopportunityforconcentratedattention to which reallearningcanbedrawnintheareaofsurvivalfactors. century offlight,anditwillreducethenumberaccidents from factors, justthewayneedforpreventiondidinfirst half- sists, willhavetworesults:itpullattentionawayfromsurvival direction ofnonsurvivability.Thattrend,ifitmaturesand per- dents seemstobeundergoingasubtlebutdefiniteshiftinthe aircraft accidents,theratioofsurvivabletononsurvivableacci- 2. Whileitistruethatfarfewerpeoplearedyinginpresent-day point inthefuture. should safetyprofessionalsdeemsuchproposalsessentialatsome to beasthoroughlyandunequivocallydocumentedpossible, investments ineithersurvivalorevacuationcapabilitieswillneed on theoverallsafetyrecord.Justificationofneedformajor were considerednovel,buthadimmediateandpowerfuleffects years ago,wheneventhemostrudimentaryprotectiveprovisions rary safetyandeconomicclimateisfardifferentfromthatof50 much ofwhatneedstobedone,hasbeendone.Thecontempo- 1. Asfarassurvivabilityandaircraftdesigncriteriaareconcerned, flect ontwofactsoflifeintoday’saviationenvironment: dations wouldbeundulyburdensometoimplementshouldre- tury offlight.”Thosetemptedtosuggestthattheserecommen- vival factorsinvestigationgoalsandobjectivesforthenext“cen- investigation reportstoday. than themorecustomary,text-heavynarrativesusedintypical outcomes thatwouldbeoffargreaterusetofutureinvestigators and preparinggraphicaldepictionsofaccidentconditions drawings, correlatingoccupantdatacontainedin“flatfile”tables, technology ismorethanuptothetasksofcreatingengineering they wereexpectedtoprovide.However,moderninformation creating themwasclearlyjustifiedonthebasisofinsights products limitedtheirusetocaseswheretheprohibitivecostsof discussed. Inmanycases,thelabor-intensivenatureofcertain regarding thepracticalityandingenuityofmanytechniques mit meaningfulactiontobebasedonit. is in apubliclyavailableform,andmuchofthedocumentationthat ment tosupportsuchactivitiesisnotwellenoughdocumented survival factorsinvestigation,research,andequipage.Therequire- tant toconfrontthecentralconcern:justifyinginvestmentsin century investigativeenvironment.However,itisequallyimpor- them insurvivalfactorsassessmenttoolkitsforusethe21st serve bothcodificationandmodificationtopermanentlyenshrine diagrammatic investigativetechniquesstillholdupwell,andde- flight, especiallyintheareaofsurvivalfactors.Someoldest by leapsandboundsthroughoutthefirstcenturyofpowered availableistooinsubstantialorinadequatelydefendedtoper- Four specificrecommendationsareofferedwithrespecttosur- Throughout thispaper,therehavebeengeneralobservations did evant survival factors dataascanbecollected inthecourseof cant improvementsintheaccessibility ofasmuchcorrelated,rel- forts. both thesuccess clear andconvincingbodyof investigative evidenceisessentialto prompted theoriginalcallfor action.Inbothcases,though,a tigations effectivelydocumentconditionssimilartothose which not acteduponmaybeworthasecondlookifsubsequent inves- place. Bythesametoken,“closed”recommendationsthat were that ledtotherecommendationhavingbeenmadein first dents failtodemonstratethedegreeofunsatisfactoryperformance Action” statusmaywarrantreconsiderationifsubsequent acci- dents involvingsimilarseatinstallations. also needstobeavailablefutureinvestigatorsofother acci- quence. Thatfactisgermanetotheinvestigationathand, and equate fortheconditionsencounteredduringimpactse- seat function,thatmeanstheexistingprovisionsweread- answer to#4alone.Ifpassengersarenotinjuredasaresultof occupants? 4. Didthatperformanceresultintheexpectedprotectionofseat and restrainingtheiroccupants? 3. Didtheseatsperformasexpectedintermsofdissipatingloads 2. Didtheseatsconformtorecommendedlevelofprotection? the accidentsequence? 1. Whatforcesweretheseatsandoccupantssubjectedtoduring each finalreport: means thefollowingquestionsshouldbeaskedandansweredin cific requirementsforpassengerseatenergyabsorption.This and directlyaddressingtheminsubsequentinvestigations. recommendations cannotfairlybeassessedwithoutconsciously previous investigations.Theeffectivenessandusefulnessofprior figurational andproceduralchangesidentifiedasnecessaryin recommendations, bothopenandclosed,regardingvariouscon- exits basedonexteriorconditions). Figure 10.Evacuationtimelineanalysis(availabilityof This recommendationcannot befullyrealizedwithoutsignifi- Some recommendationsthatremainin“Open—Unacceptable Note thattheseissuesshouldbeexplored For example,theNTSBisonrecordasrecommendingspe- and thecredibilityofanysuchretrospective ef- regardless ofthe is theonlycomprehensive compilationofrelevant dataregard- report onanaircraftaccident renderedbyaninvestigativebody exhaustively inclusiveinstead ofexpresslyexclusive.Thefinal along formulaic,standardized designs,theircontentmustbe tribute totheunderstanding of theevent. dent report,regardlessofthe degreeofilluminationtheycon- wreckage orientationdiagramsareapartofvirtuallyevery acci- aircraft’s occupants.Nevertheless,seatingarrangement and ment post-impacteventswithabearingonthesurvival of the event involvinglossofcomponentsorstructuralintegrity docu- expressly designedtodooneoftwothings:documentaninflight are illuminatingwithrespecttosurvivalfactorswhenthey are and thentonotewholiveddiedwithinthatarea. a portionoftheaircraftfromimpactforcesorpost-crash fire, ever, itishardlyamatterof“analysis”toshowthedestruction of to createthemonthatbasisintheinterestofobjectivity.How- pictions asbeinganalyticalinnature,andforthatreasondecline imprecise process.Governmentinvestigatorsoftencitesuchde- the “why”ofanevacuationsequence…acumbersomeandoften post-crash survivoractionsandmovementstogaininsightinto wreckage andimpactinformationwithnarrativedescriptionsof pant survival.Forexample,itiscommontohavecompare mation theyinclude—indifferentsections—withrespecttooccu- in thelackofcorrelationamongvariouspiecesfactualinfor- One oftheprincipaldefectsmanycurrentaccidentreportsis which thereisevenasinglesurvivororfatality. to survivalfactors,bepreparedforeverymajoraccidentin Recommendation 2.Establishastandardsetofproductsrelated versions shouldprovidedirectlinks. Printed documentsshouldaccomplishthisbyfootnote;electronic members andstaffindevelopingeachsection’sconclusions. the relevantportionsofpublicdocketrelieduponby could andshouldbegreatlystrengthenedbydirectreferenceto different sectionsofthereport).Further,analyticalreport The author’scontentionisthat, ifreportsmustbeconstructed Virtually theonlytimewreckagediagramsanddescriptions at theriskofrepeatingsomefactsin gether allrelevantpiecesoffact(even clusively analyticalinnature,tyingto- reports themselvesshouldbecomeex- into theformalreport.Therefore, major investigationneverfindtheirway contained inthepublicdocketforeach amounts ofrelevantfactualinformation ably, theproductofanalysis.Vast from the“survivalfactors”section.) pact information”isalwaysseparate section devotedto“wreckageandim- document. (Forexample,thefactual from severaldifferentpartsofthesame data pointsrelevanttooccupantsurvival sis—requires userstocullthevarious tion fromconclusionsbasedonanaly- need todistinguish“factual”informa- least partlydictatedbytheperceived NTSB reportformat—whichitselfisat each investigation.Therigidityofthe Each NTSBfinalreportis,inescap- ISASI 2003 Proceedings • 155

ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS pact withfurnishings, impactwithotheroccupants, impactwith h. Sourceofinjuries(multiple selectionspossible,includingim- out, etc.) pression fracture(s),thermal, carbonmonoxide,flail,aortictear- g. Natureofinjuries(multiple selectionspossible,includingcom- f. Occupantstatus(fatal,serious, minor,none) e. Disabledattimeofaccident (Y/N) d. Weight c. Height b. Age a. Gender and populatingthefollowingfieldsasaminimum: 10. Relationaldatabase“flatfile,”listingeachseatbynumber exit, exitoffirstoccupant,lastoccupant. 9. Timeline:initialimpactthroughfinalstop,openingof first elapsed time(keytoChart#3(b)). 8. Firepropagationchart:point(s)oforigin,directionspread, ber (keytoChart#3(b)and/or(c)asappropriate). found withinoradjacenttoeachfuselagesection,byseatnum- 7. Casualtylocationchart:offatallyinjuredoccupants Chart #3(b)). used. Includeavailabilityofeachexitandwinddirection(keyto 6. Egressroutechart:color-codetomatchseatvacatedwithexit Chart #2). whether deathoccurredduringorsubsequenttoimpact(key of injuriesasdefinedbyICAO,effectonmobility,and 5. Fatality/injurychart:mortalityinformation,includingseverity cific seatsaffected. d. Areaswhereoccupiablevolumewasnotretainedandthespe- fected. c. Areaswhereseatswerenotretainedandthespecificaf- the prevailingwind,andlocationofexteriorfire. b. Thepositionofseparatedaircraftsectionsrelativetoeachother, specific seatsaffected. a. Thelocationoffuselageseparation(s)andintrusions,the 4. Cabincompromisecharts:Chart#1,redrawntoshow: place frominitialimpactuntilallmovementhasstopped. promise, andotherenergy-inducingdiffusingeventstaking folded, includingimpactwithgroundstructures,airframecom- 3. Kinematicschart:adepictionofhowthecrashsequenceun- existing disability,weight pact sequence,includingfulldemographicdata(gender,age,pre- 2. Censuschart:whereeachoccupantwasatthestartofim- a templateforChart#2). cally listingseatpitchineachseparatecabin–thisistobeusedas ing thelayoutofaccidentaircraftpriortoimpactandspecifi- 1. Cabinconfigurationchart:seat/aisle/galley/exitdiagram,show- copy andelectronicformats,wouldbeagoodstart: content, butthecreationofstandardportfolio,inbothhard have a“survivalfactorsannex”thatwouldbuilduponthisbasic come wouldbeforeveryaccidentinvestigationfinalreportto rent stateofcomputergenerationandanimation.Theidealout- the historicalrecord,earlyICAOandAvCIRforms,cur- ous injury,orevacuation-relatedinjurytakesplace.Itisbasedon mentation, tobeusedforeveryaccidentinwhichafatality,seri- for establishmentofastandardportfoliosurvivalfactorsdocu- knowledge oftheeventandsubsequentinvestigativeactivities. ing thataccident,assembledbythosewiththemostcomplete 156 The followingisabare-boneslistintendedtoserveasbaseline • ISA SI 2003 P r oceedings 34 , etc.) single specialized team. fully metbeforetheevidence ordatawindsupinthehandsofa phase aswelltoensurecross-functional informationneedsare tivities canandmustoccur during theinformation-gathering vestigative authoritygroupchairs. However,interdisciplinaryac- deferred untiltheanalyticalphase, throughdialogueamongin- various deadlines;interaction amongthevariousgroupsoftenis gathering phaseoftheinvestigationasallpartiesworkto meet “chimneying” ofexpertisefrequentlypersiststhroughoutthe fact- and tasksalongfunctionallinesforeaseofmanagement. This a majorinvestigationistheneedtodivideinvestigativeefforts One oftheinherentlimitationsassociatedwithearlystages of ate targetingofthatresearch. cupant injuryandfatalityinformationtoensuretheappropri- port ongoingresearch.Suchdatamustbecorrelatedwith oc- data aspossibleiscollectedbystructuresinvestigatorsto sup- survival factorsinvestigatorsshouldensureasmuchobjective components andfurnishingscontributestooccupantfatalities, Recommendation 4.Wherethefailureofaircraftstructural no standardfire.” CAMI reportobserves,whenitcomestotoxicityeffects,“thereis which tobuildtheircommoncoreofunderstanding,andasthe multidisciplinary effortscannotsucceediftheylackdataupon plications fromexistingandprojectedresearch.” multidisciplinary skillswillberequiredtodeveloppracticalap- changes fromadescriptivedisciplinetomechanisticone, ment inaircraftaccidentsisclearlywarranted.Asfirescience tal researchinsmoketoxicity,firesafety,andhazardsassess- 1995 CAMIreportmadethepointthat,“Continuingfundamen- requirements andintheexecutionofstudiesthemselves.A research anddevelopmentefforts,bothintheestablishmentof niable needforareliablebodyofdataonthissubjecttoaugment but onewhichmustbespecificallyaddressed.Thereisanunde- This isasensitiveissue,involvingvarietyofprivacyconcerns, tors andresearchers. formation, iscollectedandretainedforusebyotherinvestiga- ures, includingbothsurvivorstatementsandpost-mortemin- jective informationregardingevacuationsuccessesandfail- Recommendation 3.Ensurethatasmuchobjectiveandsub- by ascompleteadatasetpossible. tive factors(e.g.,ageversusinjuryseverity)isgreatlyfacilitated the useofstatisticaltoolssuchasscatterplotstoshowcompara- u. Distancebetweenseatandexitused t. Typeofexit(e.g.,II,III,etc.) s. Exitused(number/location) r. Distancebetweenseatandnearestusableexit q. Distancebetweenseatandnearestexit p. Occupantself-evacuated o. Seatejectedfromaircraft n. Seatwithinsurvivablevolume m. Occupantrestrainedasintendedthroughoutthesequence(Y/N) l. Occupantrestraintinuse(Y/N) k. Occupantprotectedfromimpactforces(Y/N) j. Seatretainedduringimpact(Y/N) i. Survivedimpact(Y/N) intruding material,etc.) The morecomprehensiveadatabaseofthistype,thebetter; 36 35 Such Advisory GroupforAeronauticalResearchandDevelopment. Bibliography 17. Endattachmentfittingsofthesafetybeltsshouldbemountedon swivel- 16. Thestewardessandothercrewmembersshouldwearshoulderharness, 15. Thestewardessshouldbeinthemostaftsectionofcabin. 14. Brittleplasticpartitionsshouldnotbeusedinthepassengercabin. 13. Fireextinguishersandother“lethal”objectsshouldbesecured accord- 12. Overheadhatrackstructureshouldbeofdelethalizeddesignandcon- 11. Ifpractical,buffetsshouldbeusedtopartitionoffthecabininanumber 10. Buffetunitsshouldbeattachedtoprimaryfuselagestructureinsucha 9. Ifpractical,seat-backsshouldbehighenoughtoprovidesomeprotection 8. Seatsandseatanchoragesshouldbetesteddynamically—aswellasstati- 7. Iftheseatsareattachedtobothwallandfloorstructure, 6. Seat-flooranchorageunitsshoulddeformwithoutcompletefailure,upto 5. Seatsshouldbedesignedtoresistforeandaftlongitudinalloadsimposed 4. Seatstructureshouldbeductile,aswellstrong,topermitdeformation 3. Thepassenger“tiedown”(safetybelt,anchorage,portionsoftheseat 2. Thefloorstructureshouldalsobesufficientlyductiletoprovidefailureby 1. Thefloorstructureshouldpreferablybethestrongestpartofentire Cornell UniversityMedicalCollege,October1953),pp.44-45 Recommendations fromA.HowardHasbrook Appendix tive community. and broadviewwarrantallpossiblesupportfromtheinvestiga- studies ofoccupantandlocalstructure. analysis methodologies,studiesofmajorairframestructure,and background data,establishment,verificationandcomparisonof involves separatesubtasksforestablishmentanddistributionof example, theEuropeanCommunity’s“Crashworthiness”Project tinuously inneedofsubstantivedatawithwhichtowork.For Caves, RichardE. Biederman, Paul. Aviation SafetyCouncil,Republic of China. Allward, Maurice. dent Investigation ing anchoragestopreventfracturefailuresduebending. rection theyareseated,i.e.,aft,forward,orside-facing. as wellsafetybelts,duringtakeoffsandlandings—regardlessofthe di- ing toloadfactorsnotlessthanthoseusedforthepassengertie-down. struction. of sections. integration. way astopreventlargescaledisplacementupthepointoffuselagedis- for thetopsofpassengers’heads. sign. cation ofdynamicimpactloads—particularlyiftheseatsarerigidde- deformation ofthefloorandwallstructuresshouldaccompanyappli- cally—on typicalportionsoffloorand/orwallstructure.“Weaving”and will notbreaktheseatsloose.[emphasisinoriginal] their attachmentsshouldbedesignedsothat the pointofdisintegrationfloorstructure. from anypointwithin30 of passengertie-down. without completefailureofmajorportionstheseat,andresultant failure ofbasiccabinstructures. to failcompletelyunderloadslessthanthoserequiredcauseextensive strength, thepassengerseatsandseatanchoragesshouldnotbedesigned load capacityofthesafetybelt.Ifbasicfloorstructurehasagreater ture) shouldhaveastrength,foreandaft(seeitem#5,below)equaltothe which carrythesafetybeltloads,seatanchoragesandbasicfloorstruc- types” offailure. progressive bucklingandcollapseratherthanbyshatteringor“explosive tached—up tothepointofdisintegrationmajorportionsaircraft. fuselage inordertoprovideaplatformwhichtheseatswillremainat- jury Research, Airlines DC-6AccidentatElizabeth,N.J.onFebruary11,1952 lishers, 1982. Number ASC-AAR-02-04-001.Taipei,ROC:AviationSafetyCouncil, 2002. 747-400, 9V-SPK,CKSAirport,Taoyuan,Taiwan, October31,2000 on aPartiallyClosedRunwayduringTakeoff, SingaporeAirlinesFlight006,Boeing Aériens àl’usagedesMédecinsdel’Aviation A varietyofongoingresearchanddevelopmenteffortsarecon-

The U.S.AirlineIndustry:EndofAn Era Safety intheAir Manual forAirSurgeons/Manueld’EnquêtesurlesAccidents Air TransportandItsRegulators: AnIndustryStudy N o ofthelongitudinalaxisaircraft. . London:Abelard-Schuman,Ltd., 1967. . NY:PergamonPress,1961. Aircraft AccidentReport:Crashed , CrashSurvivalStudy:National 37 flexion Activitiesofthisscope ofthewallandfloor . NY:PraegerPub- (NY:CrashIn- Aircraft Acci- , Report . Cam- National TransportationSafetyBoard. Moscow, Alvin. McFarland, RossA.,PhD. Ministry ofTransportandCivilAviation. Lederer, Jerome. Lederer, Jerome. Lederer, Jerome,M.E. King, BarryG.“AircraftEvacuationUnderFireConditions,”in International CivilAviationOrganization. International CivilAviationOrganization. International CivilAviationOrganization. Hasbrook, A.Howardetal.EvacuationPatternAnalysisofaSurvivableCom- Federal AviationAdministration. Department oftheAirForce. Chaturvedi, ArvindK.andSaunder,DonaldC.Saunders. U.S. ArmyTransportationResearch Command Tufte, EdwardR. Smith, C.R. Sadeghi, M.M.andHashemi,S.M.R. Reals, WilliamJ.,M.D.,ed. Pesman, GerardJ.“HumanSurvivalinAircraftCrashFires,” Office ofTechnologyAssessment. Office ofFlyingSafety,HQUSAAF. Office ofAviationSafety. National TransportationSafetyBoard. National TransportationSafetyBoard. National TransportationSafetyBoard. landic Airlines,Inc.,DouglasDC-8-61,JohnF. Hill BookCompany,Inc.,1946. 1955. Comet G-ALYYon8thApril,1954 Court ofInquiryintotheAccidentstoCometG-AYLPon10thJanuary,1954and Nineteenth WingsClub“Sight”Lecture.NY:TheClub,April21,1982. tion, Inc.,1962. Lecture totheASMEAviation&SpaceConference 20, 1939. Jackson CabotProfessorshipLecture letin No.107 Aviation Organization,1957. gation, ThirdEdition. Civil AviationOrganization,1951. gation, SecondEdition Organization, 1949. gation Institute, May1962. mercial AircraftAccident.OklahomaCity,OK:CivilAeromedicalResearch Transportation, October1970. Aircraft, Force, 1February1956. tion—Investigation—Reporting. Oklahoma City,OK:CivilAeromedicalInstitute,February1995. Toxicity, andSurvival:AnOverview, bridge, MA:HarvardUniversityPress,1962. TC-707, TRECT echnical Report61-91, AvCIR61-2).Phoenix,AZ: Avia- February 1961 U.S. ArmyG-91ReconnaissanceJet Fighter Accident,FortRucker,Alabama,1 Graphics Press,LLC.,1983. “Sight” Lecture Cranfield ImpactCentre,1997. ness StudyProject:AEuropeanProgramme cago, IL:CollegeofAmericanPathologists,1968. letin No.106. Congress oftheUnitedStates,September1993 Technology Issues, Washington, D.C.:Headquarters,ArmyAirForces,1945. D.C.: FederalAviationAdministration,December1976. March 2001. NTSB/SR-01/01. Washington,D.C.:NationalTransportationSafetyBoard, 121 U.S.AirCarrierOperations,1983through2000 1974. 3. Washington,D.C.:NationalTransportationSafetyBoard,November gency EvacuationsfromAirCarrierAircraft, August 10,1982. NTSB-AAR-82-8). Washington,D.C.:NationalTransportationSafetyBoard, ton NationalAirport,Washington,D.C.,January13,1982 Inc., Boeing737-222,N67AF,Collisionwith14thStreetBridge,NearWashing- ington, D.C.:NationalTransportationSafetyBoard,December5,1973) Jamaica, NewYork,June23,1973(ReportNumberNTSB-AAR-73-20). . ICAODoc.6920-AN/855.Montreal:InternationalCivilAviation ReportNumberFAA/AM70-16.Washington,DC:Departmentof Safety inAirTransportationOverthe Years—TheEighthWingsClub . Boston,MA:NationalFireProtectionAssociation,May1954. Boston,MA:NationalFireProtectionAssociation,May1954. Tiger onaLeash . (Preparedby,NewYork,ork—ContractDA-44-177- . NY:TheWingsClub,May19,1971. Aviation SafetyPerspectives:Hindsight,Insight,Foresight—The The VisualDisplayofQuantitativeInformation Perspectives inAirSafety:DanielGuggenheimMedalAward ReportNumberOTA-BP-SET-121.Washington,D.C.: ICAO Doc.6920-AN/855.Montreal:InternationalCivil . ICAODoc.6920-AN/855.Montreal:International Safety intheOperationofAirTransportation:AJames Human FactorsinAirTransportDesign A SurveyofAirCarrierCabinSafety Medical InvestigationofAviationAccidents Air ForceManual62-5,AircraftAccidentPreven- . NY:G.PPutnam’sSons,1961. Washington,D.C.:DepartmentoftheAir . London:HerMajesty’sStationeryOffice, . Northfield,VT:NorwichUniversity,April Survival inEmergencyEscapefromPassenger Aircraft EvacuationTesting:Researchand Safety asaFactorintheFutureofAviation. ReportNumberDOT/FAA/AM-95/8. An OverviewoftheAircraftCrashworthi- Survivability ofAccidentsInvolvingPart Aircraft AccidentReport:LoftleiderIce- Special Study:SafetyAspectsofEmer- Aircraft AccidentReport:AirFlorida, Civil AircraftAccident:Reportofthe ReportNumberNTSB-AAS-74- Manual ofAircraftAccidentInvesti- Manual ofAircraftAccidentInvesti- Manual ofAircraftAccidentInvesti- . Cranfield,Bedfordshire,UK:

Kennedy InternationalAirport, ISASI 2003 . NY:FlightSafetyFounda- . CrashInjuryInvestigation: , SafetyReportNumber Aircraft Fires,Smoke, (ReportNumber Proceedings . Cheshire,CT: (NY:McGraw- . Washington, Aviation Bul- Aviation Bul- Wash- . Chi- • 157

ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 * Footnotes Warner, EdwardPearson,Sc.D. U.S. CivilAeronauticsAuthority. 158 16 19 18 17 TheZeppelinLZ7operatedbyDELAG(DeutcheLuftshiffahrtAktien NationalTransportationSafetyBoard, Theviewsexpressedinthispaperaretheauthor’sanddonotreflectoffi- U.S.CivilAeronautics SeetheAirTransportAssociation(ATA)on-lineversionofits “AltitudeisEverything:CanadiansintheSky,”May2003,Ottawa,Trans- OneexampleofthisnewperspectivemaybeseeninBritishaviationhis- NationalTransportationSafetyBoard, OfficeofTechnologyAssessment, Investigators alsomaywishtodifferentiatebetween“contactinjuries”and MA: HarvardUniversityPress,1962),pp.100-102passim. ard E.Caves, forming partofthe“market”driveforbetter,saferaircraft.Seegenerally Rich- dent ofairlineagreements,sofurnishingsthatwerebetter(andsafer)began however, researchanddevelopmentenabledmanufacturerstogrowindepen- that tiedaircraftdevelopmenttothepurchasers.Inpostwareconomy, the jetera,airlinesandmanufacturershadmoreorlessexclusiveagreements early airlinerswastosomeextentboundupinaircarriercompetition.Prior to 1939), p.15.Readersshouldalsonotethatinnovationincabinfurnishings in son CabotProfessorshipLecture Norwich University,February23,1938),p.5. Transportation: AJamesJacksonCabotProfessorshipLecture Publishers, 1982),p.xiii. Printing Office,1938),p.10(TableA). nautics BulletinNo.3,September1,1938 mercial airoperations. an expandeddiscussionofthelegislativehistoryunderlyingearlyU.S.com- book YOUTH%20ACTIVITY%20BOOK%20English%20WEB%20Version.pdf. port Canada,p.13.Availableathttp://www.altitudeis.com/ /www.glias.org.uk/news/198news.html. tions servingLondonweremovedtoCroydonthefollowingyear.Seehttp:/ “London TerminalAerodrome,”andthatcommercialpassengeropera- Heath wasconvertedfromRoyalAirForcetocivilaviationoperationsas ety, Mr.BillFirthnotedthat,inthebeginningofMay1919,Hounslow tory; ina2002lettertotheGreaterLondonIndustrialArchaeologySoci- ing chronologyofthisandotherground-breakingeventsinaviationhistory. 1910. Seehttp://www.adamsaviation.co.uk/new_page_3.htmforaninterest- Gesellshaft, beganservicebetweenFrankfurtandDusseldorfonJune22, 2001), p.7. SR-01/01 (Washington,D.C.:NationalTransportationSafetyBoard,March Part 121U.S.AirCarrierOperations,1983through2000, United States,September1993),p.3. Technology Issues, saia.af.mil/magazine/htdocs/febmag98/crshsurv.htm. by theU.S.AirForceSafetyCenterfollowingyear.Seehttp://www-afsc- Australian AirForce’sDirectorateofFlyingSafetyina1997articlereprinted ated withcrashsequences.ThisusefuldistinctionwasmadebytheRoyal “decelerative injuries”encounteredduringthesuddenstoppagesassoci- ton, D.C.:NationalTransportationSafetyBoard,August10,1982),p.22. tional Airport,Washington,D.C.,January13,1982, Boeing 737-222,N67AF,Collisionwith14thStreetBridge,NearWashingtonNa- ber airlines. cial positionsoftheAirTransportAssociationAmerica,Inc.,oritsmem- wich University,February23,1938. portation: AJamesJacksonCabotProfessorshipLecture Printing Office,1938. nautics BulletinNo.3,September1,1938. July 1961. tion CrashInjuryResearch(aDivisionofFlightSafetyFoundation,Inc.), Office, 1945),p.10. Factor intheFutureofAviation senger SafetyLevel inTransportCategoryAircraft, BulletinNo.57 McGraw-Hill BookCompany,Inc., 1946),pp.566-569passim. JeromeLederer,M.E., Headquarters,U.S.ArmyAirForces,OfficeofFlyingSafety, EdwardPearsonWarner,D.Sc., ATA, PaulBiederman, RossA.McFarland,Ph.D.,

Ibid., Ibid., Ibid., CAB BureauofSafetyInvestigation’s RecommendationsforIncreasingthePas- • (http://www.airlines.org/public/publications/display1.asp?nid=961) for ISA p.32. p.637. p.571. op. cit SI 2003 . Air TransportandItsRegulators:AnIndustryStudy P r oceedings OTA-BP-SET-121 (Washington,D.C.:Congressofthe The U.S.AirlineIndustry:EndofAnEra Safety intheOperationofAirTransportation:AJamesJack- Authority, AircraftAccidentsandCasualties:CivilAero- (Northfield, VT:NorwichUniversity,April20, Technical DevelopmentanditsEffectonAirTrans- Human FactorsinAirTransportDesign (Washington, D.C.:Government Printing Aircraft AccidentsandCasualties:CivilAero- Technical DevelopmentanditsEffectonAir Aircraft EvacuationTesting:Researchand Aircraft AccidentReport:AirFlorida,Inc., (Washington,D.C.:Government Survivability ofAccidentsInvolving Washington,D.C.:Government NTSB-AAR-82-8(Washing- . Northfield,VT:Nor- Safety ReportNTSB/ (Northfield,VT: (NY:Praeger (Boston, MA: (Cambridge, Airline Hand- Safety asa (NY: 21 30 29 28 27 26 25 24 23 22 20 International CivilAviationOrganization,1949),p.42. raises inthepreceding40years. after virtuallynoorganizedattention hadbeenpaidtoanyoftheissuesit basis ofworkdonebythissmallgroup ofresearcherssincetheearly1940s, was preparedafullhalf-centuryago, and(b)itwascreatedpurelyonthe paper—drawn verbatimfromthereport—and reflectonthefactthat(a)it author athttp://www.edwardtufte.com/tufte/. 1983), availablefromavarietyofon-linebooksourcesordirectly the Visual DisplayofQuantitativeInformation puter graphicstothatuse.Agoodinitialworkstartwithwouldbe field whohasdoneexceptionalworkregardingtheapplicationof com- couraged toconsulttheworkofEdwardR.Tufte,atrueinnovatorin the oping moreideasregardingthevisualpresentationofinformationare en- to bothexpertsinaviationandlaymenalike.Readersinteresteddevel- evident, andalsomakestheircontentimportanceextremelyaccessible cal industryandrequirestheintentconstantattentionofallconcerned.” problem ofaviationsafetypresentsthegreatestchallengetoaeronauti- ground,” butconcludedbyleavinginvestigatingtotheprofessionals: “The reducing thedangerandannoyancetopersonsintheirhomeson the need to“continuestudywaysandmeansofimprovingaircraftsafety and days aftertheaccident!)madeafewgracefulobservationsregarding the port, publishedas82ndCongress,2ndSession,ReportNo.1140(justnine the CommitteeonInterstateandForeignCommerce.ThatCommittee’sre- drove aseparatecongressionalinvestigationintotheaccidentconductedby ark; withindaysofthecrash,theyofferedSenateResolution268,which had alreadybeenreactingtoconstituentcomplaintsaboutoperationsatNew- aircraft, diedinthisaccident.NewJerseySenatorsSmithandHendrickson mentioned issuessinceairlinederegulation. lied bythesuccessthathasbeenseeninaddressingsomanyofafore- p. V-1.Thispessimisticbutrealisticappraisalhasbeentosomeextentbe- sideration oftheeconomicimpactversusdegreeimprovement.”Id.at interest groupsseemtoconsistentlyadvocatechangewithoutpropercon- gests whythisknowledgewassoslowtobetranslatedintoaction:“Special space IndustriesAssociationon“TheDesigner’sViewofCabinSafety”sug- An Appendixtothisreport,preparedbyRichardOstlundoftheAero- D.C.: FederalAviationAdministration,December1976),pp.14-26passim. ASC-AAR-02-04-001 006, Boeing747-400,9V-SPK,CKSAirport,Taoyuan,Taiwan,October31,2000, Crashed onaPartiallyClosedRunwayduringTakeoff,SingaporeAirlinesFlight October 31,2000;seeAviationSafetyCouncil, dent involvingaSingaporeAirlinesBoeing747-400atTaipei,Taiwanon inflations, continuetoplagueevacuationefforts,aswasseenintheacci- No. 107 NFPA’s headquartersinQuincy,Massachusetts. in thisseriesareavailableonmicrofilmattheMorganTechnicalLibrary ily lendthemselvestocontemporaryinvestigativeapplications.Bulletins tion, aredatedbutotherwiseexcellentconceptualmodelsthatwouldeas- and temperatureincrease,betweenelapsedtimetoxicgascircula- Pesman’s graphicaldepictionsoftherelationshipsbetweenelapsedtime No. 106 à l’usagedesMédecinsdel’Aviation vestigation ManualforAirSurgeons/Manueld’EnquêtesurlesAccidentsAériens Advisory GroupforAeronauticalResearchandDevelopment,AircraftAccidentIn- arising outoftheadvancesinaviationequipmentandtechniques.”See ness ofthepreventivemeasuresaswelltodetectaccidentforces ending, sincedatamustcontinuetobeassembledassesstheeffective- occur toprovetheaccidentpotentialofacondition…Theprocessisun- strated inaccidentsorincidents,butitisnotnecessaryforanaccidentto served, recorded,andprovenbyanalysis.Suchconditionsmaybedemon- measures canbeexercised,theaccidentpotentialconditionsmustob- “[Data collection]isabasicprerequisiteofprevention.Beforepreventive vocate ofcomprehensivedocumentationinjuriesincurredduringcrashes: vival factorsandtheirinvestigation.AGARDwasaparticularlystrongad- thus bringingtogetheroncemorecivilandmilitaryperspectivesonsur- sory GroupforAeronauticalResearchandDevelopment(AGARD)in1961, tion National FireProtectionAssociation,June30,1950). ICAODoc.6920-AN/855, A.HowardHasbrook, etal., Readersareencouragedtoreview thefulllistinAppendixtothis Thevalueofillustrationsalongthelinesexamplesprovidedis self- Sixpeopleontheground,aswellaformerSecretaryofWaraboard OfficeofAviationSafety, Forexample,theeffectsofwindonescapeslides,aswelluncommanded BarryG.King, GerardJ.Pesman, ThesesameformsandprotocolswereinlargepartadoptedbytheAdvi- ICAODoc.6920-AN/855, (Montreal: InternationalCivilAviationOrganization,1951),pp.30-31. (Boston,MA:NationalFireProtectionAssociation,May1954).Mr. (Boston,MA:NationalFireProtectionAssociation,May1954). Aircraft EvacuationUnderFireConditions,AviationBulletin Human SurvivalinAircraftCrashFires,AviationBulletin (Taipei, ROC:AviationSafetyCouncil,2002),p.224 A SurveyofAirCarrierCabinSafety Manual ofAircraftAccidentInvestigation,SecondEdi- Manual ofAircraftAccidentInvestigation Evacuation Pattern AnalysisofaSurvivableCom- (NY:PergamonPress,1961),pp.67–69. (Cheshire,CT:GraphicsPress,LLC., Aircraft AccidentReport: (Washington, (Montreal: The 33 32 31 providing informationtocabinpreparation).Althoughnotdifferentiated communications problems;andpreparingpassengerstoevacuate(from exits, useofslides[and]communicationsystems;humanfactors,including gases; technicalspecificsofevacuationsystems,e.g.openingemergency areas ofconsiderationproposedinthisstudywerefires,smokeandtoxic l’Aviation Civile(F-DGAC)offeredalistsimilartotheabove.Thefour ation problemsmustbeaddressed,theFrenchDirectionGénéralde rates observedinallthreeaccidents. ing factorthatmighthavehadabearingonthehigherfemalemortality that wentsofarastodocumentparentalstatus(p.46)apossibledelay- exits used,aswelldetailedepidemiologicalanalysesofalloccupants tained ausefulanalysisofpassengerevacuationtimesandnumbersbythe which wasanin-depthreviewofthreeaccidentsfromthe1960s,alsocon- ington, DC:DepartmentofTransportation,October1970).Thisreport, November 1974). Air CarrierAircraft Institute, May1962). mercial AircraftAccident In1999,inidentifyingthevariousdomainsacrosswhichemergencyevacu- FAA/AM70-16, NTSB-AAS-74-3, Survival inEmergencyEscapefromPassengerAircraft (Washington,D.C.:NationalTransportationSafetyBoard, Special Study:SafetyAspectsofEmergencyEvacuationsfrom (OklahomaCity,OK:CivilAeromedicalResearch (Wash- 37 36 35 34 Impact Centre,1997),p.2. thiness StudyProject:AEuropeanProgramme Civil AeromedicalInstitute,February1995),p.4. ity, andSurvival:AnOverview, AccientResearch/600Air-Pre-Bec.html. tion Accident, Veronneau, abstract ofapresentationonthisissue,A.L.PennybackerandS.J.H. addressed, however,theimportanceofthisvariablebecameobvious.An Civil AerospaceMedicalInstitute’s“AircraftAccidentResearchTeam.”Once cifically includedaspartofananalysisperformedinthemid-1990sby eration asafactorinseatfailureduringcrashsequenceuntilitwasspe- actu_gd/sfatsecu/dedale1A.pdf.) 10th, 1999,p.4. raphy Review—Researchstudyonemergencyevacuation,Revisionn the omissionsareequallynoteworthy.SeeF-DGAC, by typesoffactorsasseenintheothertwo,overlapsarestriking,but M.M.SadeghiandS.M.R.Hashemi, ArvindK.ChaturvediandDonaldC.Saunders, Thereislittleevidencethatpassengerweighteverwastakenintoconsid- Ibid., p.3. Description ofFailuresPassengerSeatsinaHighVerticalAccelera- isavailableathttp://www.cami.jccbi.gov/AAM-600/ (Englishversionavailableathttp://www.dgac.fr/html/ DOT/FAA/AM-95/8(OklahomaCity,OK: An OverviewoftheAircraftCrashwor- (Cranfield,Beds,UK:Cranfield ISASI 2003 Aircraft Fires,Smoke,Toxic- Interim Report:Bibliog- Proceedings o 2,March • 159

ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS 160 The databaseaccess menuisasfollows: 3. Contentofthedatabase www.rgwcherry.co.uk www.fire.tc.faa.gov/cabwg.stm accidents, andofthese,textual informationisavailableon742. diagrams aswelltextualandnumericaldata. ments orequivalent.Thedatabasecontainsphotographs and ger seats)andcargoaircraftcertificatedunderPart25require- transport-category passengeraircraft(with19ormorepassen- from accidentinvestigatingauthorities.Recordsarestored for for passengerandcargooperationsovertheperiod1967to 2001. The databaseconsistsofaccidentsinvolvingoccupantinjuries 2. Populationofthedatabase documentation. features ofExcelandWordmaybeusedfordataanalysis ii) ItscompatibilitywithMicrosoftsoftwaremeansthatallofthe with similarfeatures. i) Itswordsearchcapabilityallowsrapidretrievalofaccidents out fortheairworthinessauthorities. safety andhasbeenusedmanytimesinbenefitanalysescarried tool instudyingtrendsandquantifyingimprovementsaircraft establishing aninternationalframework.” hance theeffectivenessandtimelinessofcabinsafetyresearchby pan, Australia,Brazil,andRussia,whosepurposeisto“en- consist oftheaviationauthoritiesNorthAmerica,Europe,Ja- funded bytheCSRTG.TheCSRTGisagroupwhosemembers project carriedoutbyR.G.W.Cherry&AssociatesLimited. Group (CSRTG)wasdevelopedoutofaEuropeanUnionresearch The accidentdatabaseoftheCabinSafetyResearchTechnical 1. Historyofthedatabase offering aconsultancyandcontractualservicetotheindustry. Airworthiness withRaytheonCorporateJets.Heisnowpartofateam Certification EngineerwithBAe,andHeadofProductIntegrity Raytheon, includingHeadofDevelopment&Reliability,ChiefSafety The databasemaybedownloaded fromthefollowingwebsites: At Issue21,thedatabasecurrentlycontainsinformationon2,819 All datahavebeenderivedfromreliablesources,primarily Its primeattributesare The databasehasprovedtobeanextremelyusefulanalytical The databasewassubsequentlyadoptedanditsdevelopment • ISA The AccidentDatabaseoftheCabin SI 2003 Safety ResearchTechnicalGroup P Ray Cherry with HawkerSiddeleyAviation,BritishAerospace,and teams. Hehasheldmanypositionsduringhiscareer the Airbus,BAe146,and125/Hawkerdesign aircraft industryin1968andhasbeenamemberof Institution ofMechanicalEngineers.Hejoinedthe Royal AeronauticalSociety,andamemberofthe r oceedings is aCharteredEngineer,Fellowofthe By RayCherry,R.G.W.Cherry&AssociatesLimited,UK 1 identifier orname,e.g.,“SiouxCity”and“Kegworth.” This fieldisusedwhereanaccidentcommonlyknownby Identifier event, subsequentaccidentswillbeannotated“B,”“C,”etc. accidents, shouldmorethanoneoccuronthesamedate.Inthis The alphanumericcodehasbeenusedtodifferentiatebetween the July25,1991,wouldhavefollowingcode:19910725A. of occurrencetheaccident.Henceanaccidentoccurringon Each accidenthasauniquereferencenumberbasedonthedate Reference General fields reports generatedbyaccidentinvestigatingauthorities. Custom List(Nodatapresent) Exit andAssistMeans Orientation Environment Impact Factors Water Factors Fire Factors Occupant Data Aircraft Data Exit Usage Injury Locations Photographs Screen 3 Screen 2 Screen 1 Figure 1.Example ofaScreen1entryinthe accidentdatabase. Screens 2and3aretextualfieldsprovidingtextextractedfrom of impact. This fielddefineswhetherthe fueltankwasrupturedasaresult Fuel TankRuptured of impact. This fielddefineswhetherthefuselagewasrupturedasa result Fuselage Ruptured dent involvedanimpactwiththegroundorwater. This fieldcontainsYESorNOdependingonwhetherthe acci- Impact Related dent involvedalightingonwater. This fieldcontainsYESorNOdependingonwhetherthe acci- Water Related dent involvedafire. This fieldcontainsYESorNOdependingonwhethertheacci- Fire Related Number ofInjuries Number ofFatalities Number ofOccupants report. This fieldcontainsthereferencenumberofofficialaccident Accident ReportReference Aircraft Operator Location ofAccident Date ofAccident Registration Number E =greaterthan400,000lbs D =250,000lb-400,000lbs C =100,000lb-250,000lbs B =12,500-100,000lbs A =lessthan12,500lbs used: ing categoriesofmaximumtakeoffweight(MTOW)havebeen This fielddefinestheweightcategoryofaircraft.Thefollow- Weight Category This fielddefinestheaircrafttypeandseries,e.g.,B-737-236 Aircraft Type Screen 1containsthefollowingfields: aircraft, andtheinjuriestooccupants. All accidentsonthedatabasehaveaScreen1entry. Screen 1 Screen 1containsthebasicdataregardingaccident, An exampleofScreen1isshowninFigure1. Figure 2.ExampleofaScreen2entryintheaccidentdatabase. The Screen2fieldsareasfollows: a typicalScreen2entry. textual summaryoftheaccident. SeeFigure2foranexampleof Screen 2containsadditional fieldstoScreen1andprovidesa Screen 2 evacuation. This fielddefineswhethertheaccidentinvolvedanemergency Evacuation Occupant Fatalities Occupant Injuries None Minor Substantial Destroyed ing categories: accident. Theaircraftdamagehasbeenclassifiedintothefollow- This fielddefinestheextentofaircraftdamageasaresult Aircraft Damage the aircraftoverrunningrunway. This fielddefineswhethertheaccidentwascausedasaresultof Overrun curred. This fielddefinesthephaseofflightinwhichaccidentoc- Phase ofFlight cinity oftheairfield. This fielddefineswhethertheaccidenttookplacewithinvi- Runway Vicinity or duringthenight. This fielddefineswhethertheaccidentoccurredduringday Day/Night ISASI 2003 Proceedings • 161

ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS 162 in Figure4. The codesusedtoindicatecauses andextentofinjuryareshown sustained byeachoccupant assignedseatposition. This sectionofthedatabase contains adiagramoftheinjuries Injury locations display. tographs anddrawingspertinenttotheaccidentforon-screen This sectionofthedatabasecontainsaselectionavailable pho- Photographs entry isshowninFigure3. appear undereachoftheheadings. from accidentreportsalthoughitwillhavebeenreordered to 2 textfieldandusesthesameheadings.Theistakendirectly The Screen3textfieldcontainsanexpandedversionofthe Screen 3 Conclusions Injuries toOccupants Environmental Conditions Aircraft Factors Evacuation Fire Impact Résumé cident underthefollowingheadings: The Screen2textfieldcontainsahigh-levelsummaryoftheac- Screen 2Text minor/no injuriesforthecrew,passengersandtotal. fatalities, numberofnon-fatalseriousinjuries,and The databasehasagroupoffieldscontainingthenumber Number ofFatalitiesandInjuries compassing flightcrew,cabinandpassengers. This fieldcontainsthetotalnumberofoccupantsonboarden- Total Aboard Figure 3.ExampleofaScreen3entryintheaccidentdatabase. • ISA SI 2003 P r oceedings

An exampleofaScreen3 accident database. Figure 5.Typicalexampleofinjurylocation diagraminthe database. Figure 4.Codesusedforinjurylocationsintheaccident Number ofPassenger Seats High Wing/LowWing Weight Category Engine Type Engine Configuration Number ofEngines Aircraft pear onthedatabase.Thefieldsareasfollows: to theaircraft.Figure7showsaircraftdatafieldsasthey ap- This sectionofthedatabasecontainsbasicinformationpertinent Aircraft data usage diagramisshowninFigure6. by eachoccupantseatposition.Atypicalexampleofan exit This sectionofthedatabasecontainsadiagramexit used Exit usage ure 5. A typicalexampleofaninjurylocationdiagramisshowninFig- Figure 7.Aircraftdatafieldsastheyappearonthedatabase. database. Figure 6.Exampleofexitusagediagramintheaccident pants areannotated as“N”inthisfield. evacuation capabilityareannotated witha“Y.”Allotheroccu- Passengers havinganyform of infirmitythatmayimpairtheir Infirmity Weight Height Sex Age purposes. senger namesarereplacedwithanasteriskfordatapresentation is utilisedsolelyaspartofthedatapreparationprocessand pas- The occupantnamesarenotincludedinthedatabase.This field Name sengers. The occupantsareclassifiedintoflightcrew,cabinandpas- Type occupant references. The flightcrew,cabinandpassengerseachhaveunique Occupant reference The fieldscontainedinthissectionareasfollows: shows theoccupantdatafieldsastheyappearondatabase. to theoccupants—passengers,cabincrew,andflightcrew.Figure8 This sectionofthedatabasecontainsbasicinformationpertinent Occupant data Maximum SeatsperRow Number ofAisles contains aquestionmark“?”.] being surroundedbysquarebrackets)orunavailable,thefield seats. Ifthenumberofpassengerseatsisestimated(indicatedby the numberofpassengersaboardbypassenger Load Factor[Thisfieldisnotstoredbutderivedfromdividing Figure 8.Occupantdatafieldsastheyappearonthedatabase. ISASI 2003 Proceedings • 163

ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS 164 This fieldindicates theseatpitchbetween occupantsseatand Seat Pitch approved toandtheclassofseat. This fieldindicatesthestandard ofrequirementthattheseatwas Seat Standard/Class result oftheaccident. This fieldindicateswhethertheoccupantsseatwasdamaged asa Seat ImpactDamage tened atthetimeofaccident. This fieldindicateswhethertheoccupanthadtheirseatbelt fas- Seat Belt This fieldindicatestheseatthatoccupantwasoriginallyallocated Seat Allocated exit route. This indicateswhethertheoccupantusedseatbacksasan Seat Backs This indicatestheexitusedbyoccupant. Exit Used injury severitythehigherAISlevel. Injury (AIS)Scorefornineareasofthebody.Thehigher There areninefieldsavailableforindicatingtheAbbreviated AIS Level 75. Thegreaterthenumberhigherinjuryseverity. score inthreedifferentbodyregions Score (ISS).TheISSisthesumofsquareshighestAIS Occupant InjuriesmaybeindicatedbyBaker’sInjurySeverity ISS Level nally orexternallytotheaircraft. ing causeofinjuryandwhethertheinjuriesweresustainedinter- Occupant Injuriesareindicatedbyseven2-digitcodesindicat- Injury Code(OccupantData) Figure 9.Firefactorsfieldsastheyappearonthedatabase • ISA SI 2003 P r oceedings 2 . ISSscoresrangefrom1to . . This fieldindicateswhetherthe occupantutilisedaliferaft. Life Raft This fieldindicateswhetherthe occupantutilisedalifevest. Life Vest is annotatedwith“Y”;nofloorfailure“N.” in theimmediatevicinityofoccupantsseat.Floordisruption This fieldindicateswhethertherewasanydisruptiontothe floor Floor Failure occupants seatrowtothenearestexit. This fieldindicatestheminimumaislewidthininchesfrom the Aisle Width tion tothenearestaisle. This fieldindicatesthenumberofseatsfromoccupantsloca- Number ofSeatstoNearestAisle the seatinfront,orabulkhead,inches. Figure 11.Impactfactorsfieldsastheyappearinthedatabase. Figure 10.Waterfactorsfieldsastheyappearinthedatabase. Premeditated Ditching Aircraft TotallyorPartiallyonWater defines thenatureofevent. This fielddefineswhethertheaccidentwaswater-related and Water Related The fieldscontainedinthissectionareasfollows: factors fieldsastheyappearonthedatabase. to thefireforwater-relatedaccidents.Figure10shows water This sectionofthedatabasecontainsbasicinformationpertinent Water factors aircraft involvedintheaccident. This fieldspecifieswhetherseat-blockinglayerswerefittedtothe Seat Blockingfitted cabin materials This fieldspecifiestherequirementstandardapplicableto Standard ofCabinMaterials Hydraulic FluidType Fuel Type the cabin. This fieldgivesdetailsofthemechanismforfirepenetration Fire PenetrationofCabin This fieldgivesdetailsofthemediainvolvedinfire. Fire Medium the impact. This fieldgivesdetailsofwhetherthefueltankwasrupturedby Fuel TankRuptured This fieldgivesdetailsoffireorigininafire-relatedaccident. Fire Origin related accident. This fieldgivesdetailsoftheignitionsourcefireinafire- Ignition Source craft. This fieldindicateswhethertherewasafireexternaltotheair- Fire/Smoke ExternaltotheAircraft and definestheextentoffire. This fielddefineswhethertheaccidentwasfireorsmokerelated Fire Related tained inthissectionareasfollows: Factors fieldsastheyappearonthedatabase.Thecon- to thefireforcabinfire-relatedaccidents.Figure9showsFire This sectionofthedatabasecontainsbasicinformationpertinent Fire factors of theimpact. This fieldspecifieswhetherfloor disruptionoccurredasaresult Floor Failure result oftheimpact. This fieldspecifieswhetherbulkhead disruptionoccurredasa Bulkheads uncontained asaresultoftheimpact. This fieldspecifieswhetherthestowagecontents were Stowage Contents as aresultoftheimpact. This fieldspecifieswhetheroverheadstowagedisruptionoccurred Overhead Stowage result oftheimpact. This fieldspecifieswhetherwardrobedisruptionoccurredasa Wardrobes of theimpact. This fieldspecifieswhethertoiletdisruptionoccurredasaresult Toilets as aresultoftheimpact. This fieldspecifieswhetherthegalleycontentswereuncontained Galley Contents of theimpact. This fieldspecifieswhethergalleydisruptionoccurredasaresult Galleys Percentage ofStowageNon-Retained Percentage ofOverheadStowageDetached The fieldscontainedinthisSectionareasfollows: database. Figure 11showstheimpactfactorsfieldsastheyappearon Impact factors Section SupportedbyWater given foreachsectionoftheaircraft. as aresultoftheimpact(groundorwater),flotationtimeis minutes. Iftheoccupiableareaofaircraftbrokeintosections, This fieldspecifiestheestimatedflotationtimeofaircraftin Flotation Time Distance ofAircraftfromShoreline water. etrated asaresultofimpactwithgroundpriortoalightingon This fieldspecifieswhethertheaircraftpressurehullwaspen- Damage DuetoImpactwithGround etrated asaresultofimpactwithwater. This fieldspecifieswhethertheaircraftpressurehullwaspen- Damage DuetoImpactwithWater ISASI 2003 Proceedings • 165

ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS ing thefirstandlastseatrow ineachsection. aircraft, boundedbymajorruptures ofthefuselage,byprovid- This fieldindicatesthesizeand locationofeachsectionthe Area BetweenSeatRows impact. the aircraftboundedbymajorrupturesoffuselagecaused by This fieldassignsareferencenumbertoeachofthesections of Section Reference The fieldscontainedinthissectionareasfollows: tabase. Figure 13showstheorientationfieldsastheyappearon da- Orientation the accident. This fieldindicatestheotherweatherconditionsatsceneof Other WeatherConditions dent. This fieldindicatesthewindconditionsatsceneofacci- Wind This fieldindicatestheprecipitationatsceneofaccident. Precipitation This fieldindicatesthevisibilityatsceneofaccident. Visibility time oftheaccident. This fieldsummarisestheprevailingweatherconditionsat Main WeatherConditions fire. This fieldindicateswhetherrescuerswereimpededbywateror Impediment toRescuersbyWaterorFire logical conditions. This fieldindicateswhetherrescuerswereimpededbymeteoro- Impediment toRescuersbyMeteorologicalConditions pants externaltotheaircraft. This fieldindicateswhetherinjuriesweresustainedbytheoccu- External totheAircraft The fieldscontainedinthissectionareasfollows: database. Figure 12showstheenvironmentfieldsastheyappearon Environment and longitudinalaxes. impact, asreportedintheaccidentreport,vertical,lateral, These fieldsindicatethepeak“g”levelsencounteredduring Peak “g”Levels or structuralfailure)occurredasaresultoftheimpact. This fieldspecifieswhetherseatdisruption(includesdistortion Seat Failure 166 • ISA SI 2003 P r oceedings This fieldindicatesforwhom theexitisintended. Who For This fieldreferenceseachof the exits Exit Reference The fieldscontainedinthissectionofthedatabaseareasfollows: the databaseisshowninFigure14. An exampleoftheexitandassistmeansfieldsastheyappear on Exit andassistmeans section oftheaircraft. This fieldindicatesthecauseofadverseorientationfor each Cause ofAdverseOrientation enced sectionoftheaircraft,indegrees. This fieldindicatesthepitch,roll,andyawangleforeachrefer- Pitch, Roll,andYawAngle injury. Figure 12.Environmentfieldsastheyappearonthedatabase Figure 13.Orientationfieldsastheyappearonthedatabase. This fieldindicatestheminimumaislewidth,ininches,from the Minimum WidthinAisle exits, orthestepdownheightfornon-floorlevelexits. with theaircraftsupportedbyundercarriageforfloor level This fieldindicatesthedistancefromdoorsilltoground Sill Height This fieldindicatestherelativedispositionofexits. Distance fromNose JAR/FAR 25.807. This fieldindicatesthewidthofexitininchesasdefined Width JAR/FAR 25.807. This fieldindicatestheheightofexitininchesasdefined Height of theexitisgiveninpounds. For exitsthatrequireliftingtoopenbytheoccupantsweight Weight floor exit. This fieldindicateswhethertheexitisafloorlevelornon- Floor/ Non-Floor This fieldindicatesthetypeofexitasdefinedinJAR/FAR25.807. Type database. Figure 14.Exitandassistmeansfieldsastheyappearonthe 2 1 Footnotes structed. This fieldindicateswhetherthereferencedassistmeanswas ob- Obstructed (AssistMeans) exit failed,andifso,themodeoffailure. This fieldindicateswhethertheassistmeansatreferenced Failed exit wassuccessfullydeployed. This fieldindicateswhethertheassistmeansatreferenced Deployed Deploy Attempted Assist MeansFitted Number ofOccupantsUsingExit structed. be openedandindicateswhetherthereferencedexitwasob- This fieldisapplicabletoexitsthatwereopenedorattempted Obstructed (Exit) although anattemptwasmadetoopenit. This fieldindicateswhetherthereferencedexitfailedtoopen Failed occupants. This fieldindicateswhetherthereferencedexitwasopenedby Opened to openthereferencedexit. This fieldindicateswhetheranattemptwasmadebyoccupants Open Attempted floor levelexits. This fieldindicatesthewidthbetweenseats,ininches,fornon- Width BetweenSeats floor levelexits. This fieldindicatestheminimumaislewidth,ininches,fornon- Minimum WidthinCrossAisle referenced exittothenextnearestexit. TheAbbreviatedInjuryScale”1990Revision-AssociationfortheAdvance- UnitedStatesFederalAviationAdministration,www.fire.tc.faa.gov/cabwg.stm ment ofAutomotiveMedicine N ISASI 2003 Proceedings • 167

ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS 168 recovery vehicle) systemrecoveredanatomic bomblostoff ments. WhentheU.S.NavyCURV (cablecontrolledunderwater government andcommercial sectorstomeetvariousrequire- development ofthesesystems tookplaceinthe1960sboth played apivotalroleinsearch-and-recovery operations.Initial Side scansonarandremotely operatedvehicles(ROVs)have Background I willtellyouhowwedoours. culties. So,ratherthantellyouhowtodoyouraspectofthe job, tion accidentinvestigationspresentuniquechallengesand diffi- surprised bythecapabilitiesofequipment.Overwater avia- tary andcivilianaircraft.Inmanyinstances,theinvestigators were the techniquesthatweapplytolocationandrecoveryof mili- search-and-recovery operationswithaircrashinvestigators. bered themanyconversationsonback-deckofshipsduring by thesheermagnitudeofsubjectmatter.Then,Iremem- 10 space-relatedcraftfromthedepthsofocean. than 110militaryaircraft,16commercialcivilaviationand our currentoperationspersonnelhavelocatedandrecoveredmore while workingforothercompaniesthatheldtheNavycontract, accident investigationsinthelast2years.Overpast28years, formed search-and-recoveryoperationson10militaryaviation tract. Underthiscontractoverthepasttwoyears,wehaveper- the searchandrecoveryspecialistforU.S.Navy. Operations ManagerforPhoenixInternational,Inc.(Phoenix), investigation? the aircraft,inordertodeterminecauseofcrash. investigator, youneedtoanalyzeallthedata,aswellpiecesof ters offshoreinover2,000metersofseawater.Asanairsafety A commercial747with230passengershasbeenlost100kilome- Introduction orally presented.—Editor] [This paperservedasanalternateselectionfortheseminarbutwasnot operations company,from1989until1998whenhejoinedPhoenix. industry in1983andwaswithEastportInternational,anROV PHOTO NOT AVAILABLE I’ve oftenbeenaskedabouttheequipment,capabilities,and While consideringthesubjectofthispaper,Iwasconfronted Phoenix holdstheU.S.Navy’s5-yearunderseaoperationscon- My nameisStevenSaintAmour,andIamtheCommercial How areyougoingtorecoverthepiecesneedforyour Your problemis:Howareyougoingtofindtheaircraft? • ISA SI 2003 P Steven SaintAmour and inspections.Hebeganworkinthediving four militaryaircraftrecoveries,andfivevesselsearch in onecivilaviationdisaster,threerocketrecoveries, During thelast4years,hehasbeendirectlyinvolved responsible forPhoenix’sROVBuildProgram. Manager forPhoenixInternational,whereheis By StevenSaintAmour,CommercialOperationsManagePhoenixInternational,Inc r oceedings The ArtandScience Search &Recovery: istheCommercialOperations such asthespaceshuttle covery operationstoassistinvestigatorsduringnationaltragedies Flight 990.Wehavealsoconductedunderwatersearchandre- 800, SwissairFlight111,AlaskaAirlines261,andEgyptAir Navy contractandassetsforsuchinvestigationsasTWAFlight cies intimesofemergency. expertise availabletoforeigngovernmentsandinvestigatoryagen- other U.S.governmentagencies.TheNavyalsomakesthis sets duringat-seasearchandrecoveriesfortheU.S.Navy nel operateamixtureofcompany-andgovernment-ownedas- in thebusinessformorethan30years. experience inthisfield,withsomeofourpersonnelhavingbeen dent ofPhoenixtoourseniortechnicians,average16years fessionals whocarriedoutthesedifficultmissions,fromthepresi- covery operationsunderthiscontractformanyyears.Thepro- ees havebeeninvolvedwiththeperformanceofsearch-and-re- the underseaoperationscontract,in2001.Manyofouremploy- pable thantheirearliercounterparts. driven byrecentrapidadvancesintechnology,arefarmoreca- new generationsystemsaremorecomplexyetreliable,and today—side scansonar,ROVs,andmannedsubmersibles.The trade” usedinthe1960sand70saresimilartothosestill external assistancewasrequired.Remarkably,the“toolsof low theU.S.Navytorapidlyaccessthisexpertiseincaseswhere the U.S.Navyestablishedasearch-and-recoverycontracttoal- expertise togovernmentunderwatersearch-and-recoveryneeds, and-recovery missions. indisputable factthatremotetechnologycouldbeusedforsearch- Palomares, Spain,inanaircraftaccident1966,itbecame Navy neededa systemtodetectpingersindeep water. mercially availableforlocating pingersinshallowwater,butthe immersion inwater.Ahandheld pingerlocatorsystemwascom- signals foranextendedperiod oftimeimmediatelyupontheir on allcommercialandmany militaryaircraft.Thepingersemit advantage ofthepresence acousticbeacons(pingers)carried ward atthetimeandwaspart ofthatteam.Theconcepttook Mike Kutzleb,currentlypresidentofPhoenix,workedfor Sea- type systemtoassistinrapidlylocatingaircraftlostover water. recovery contractor,Seaward,Inc.,todesignandbuildaproto- Pinger locators: have abetterunderstandingofwhatisinourtoolbox. equipment thatisreferredtothroughoutthispapersoyou will covery missions,Iwouldliketopresentabriefdescription ofthe Before wedelveintotheoperationalaspectsofsearch-and-re- Tools ofthetrade Forexample,theFAAandNTSBhaveutilizedU.S. Through theunderseaoperationscontract,Phoenixperson- Phoenix wascompetitivelyawardedthecontract,nowcalled In recognitionoftheapplicabilitycommercialunderwater In1976,theU.S.Navytaskeditssearch-and- Columbia and Challenger . disasters. easily transported, andquicklysetup.The depthcapabilityis Search andInspectionROVs: the twoprimarytypes—search/inspection andwork-classvehicles. all shapes,sizesandcapabilities. Iwilllimitmyexplanationto ROV computerandthesurface controlcomputer.ROVscomein tors andfibersforpower computertelemetrybetweenthe to theworkplatformbyanumbilical thatcarriescopperconduc- objects. TheROViscontrolledfromthesurfaceandattached visual range.Theyalsohavearms-manipulatorsforhandling ing, camerasforseeingandsonardetectingobjectsoutside are tele-presencerobotics.TheROVhasthrustersformaneuver- Remotely OperatedVehicle(ROV): lost aircraftormappinganunderwaterdebrisfield. side scansonar,itisthesystemofchoicewhensearching fora very shallowwater.Giventhetowspeedandareacoverage of meters ofwater.Somenewdesignscanoperateupto10knots in four knotsinrelativelyshallowwatertoabout2 6,000 quality acousticimagesoftheseafloor. Digital technologycanallowtheproductionofnear-photographic or more)provideveryhighresolutionimageryattheshortranges. large objectsatgreatrangewhilethehigherfrequencies(200kHz is achieved.Thelowerfrequencies(30-100kHz)areusedtodetect cies, onelowandtheotherhigh,optimumsonarperformance much shorterranges.Byoperatingwithtwosimultaneousfrequen- Using ahigherfrequencyresultsinextremelygoodresolutionbut the greaterrangeofsonar,butresolutionisdegraded. range andresolutionofthesonarsystem.Thelowerfrequency, terpretation byaskilledsidescansonaroperator. is received,processed,anddisplayedinafashionthatallowsin- reflected backtothesonarsystemwherereturnsignalorecho other objects(naturalorman-made),aportionoftheenergyis the medium(watercolumn).Uponstrikingoceanflooror they emitadiscretepulseofenergy(sound)thatradiatesoutinto deployed andthespeedoftowship. The depthofthetowfishisdeterminedbylengthcable board sidesandispulledthroughthewaterusingtowcable. contains twoormoretransducersmountedonitsportandstar- and topsidesonarcontrol/dataprocessingsystem.Thetowfish depths of6,000meterswithahighdegreeresolution. vices capableofsearchinglargeareastheseafloortowater sign refinementshaveledtopresent-dayacousticimagingde- periments conductedbyDr.HaroldEdgerton.Subsequentde- tems beganinthe1960sasaresultofunderwateracousticex- Side ScanSonar: tive searchtool. detection range,TPLhasevolvedintoareliableandcost-effec- able searcherstoquicklypinpointthelocation.Withitsextended hydrophone andtheotherhasadirectionaltoen- There aretwodifferentsystems—onehasanomni-directional versions oftheTPLcanbeusedinwaterdepthsto6,000meters. search teamtorespondquicklyinanemergency.Thecurrent The systemwasdesignedtofitintotwosuitcases,enablingthe reliably detectanacousticpingeratrangesofupto2,000meters. Locator (TPL)system.FieldtestsofTPLshowedthatitcould Tow speedsforsidescansonarsearchoperationsrangefrom The operationalfrequencyofthetransducersdetermines Side scansonarscanbeconsideredanalogoustoradarsinthat A typicalsidescansonarsystemconsistsofatowfish,towcable, The resultingprototypesystemwasnamedtheTowedPinger TheU.S.developmentofsidescansonarsys- TheseROVsaretypicallysmall, Remotelyoperatedvehicles always occurwithout warning,anditisimportant thatanalysis and executionoftheinvestigation. Bynature,accidentsalmost considered uniquewithregard totheplanning,preparation, The investigationofanoverwater accidentshouldalwaysbe Tasking andplanning and theliftcableistransferred tothewinchforrecovery. into theliftcable,spoolerisraisedtosurfaceby ROV, object tobelifted.Oncethehasbeenriggedandhooked that istakendowntothebottombyROVandplacednear the Spooler: and liftcableareminimized. dampen theheavesothatsnaploadingforcestowreckage being lifted,therodtravelssmoothlydownintocylinder to is aheaveandthedragincreasesoverallweightof object will stayfullyextendedatthepredictedweight.Inevent there weight oftheliftdetermineschargepressuresothat rod that ischargedwithaninertgassuchasnitrogen.Theestimated hydraulic cylinder.Thecylinderisplumbedtoagasaccumulator mounted tothebaseofunitandrodendalarge ing system.Therecoverylinerunsthroughaseriesofblocks pensating systemisahydraulicandgas-compensateddampen- Heave motioncompensatingsystem: tra, orDyneema. margin. Liftcablesaretypicallymadeofsteelwire,Kevlar,Spec- the expectedload,dynamicsofsupportvessel,andasafety type oflineusediswhollydependantonthesizeandweight age andisusedtohaulthewreckagesurface.Thesize Lift cable: winch. tained powerunitthatsupplieshydraulictotherecovery Diesel hydraulicpowerunit(HPU): intact aircraft. possible sizeofthewreckage,i.e.inanaircraftrecovery—theentire sure releasedbrakes.Thewinchsizeisdictatedbythelargest typically alow-speed;highlinepullhydraulicwinchwithpres- ing therecoverylineandwreckagetosurface.Thiswinchis Winch: compensating system. power unit,liftcable(steelorsynthetic),andaheavemotion face. Thesystemconsistsofarecoverywinch,dieselhydraulic intended forliftinglargedebrisfromtheseabottomtosur- HLS (heavyliftsystem): surface. the manipulatorsorslungunderROVfortransportto recovering largeitemsofdebris.Smallercanbycarriedin form acompleteinspection,theninstallriggingandliftlinesfor your primarytoolforrecoveryoperations.TheseROVscanper- (2,000 lbs)toverylarge(14,000lbs).Awork-classROVwillbe lators orroboticarms.Theyrangeinsizefromrelativelysmall capable ofconductingsophisticatedoperationsbyusemanipu- Work-Class ROVs: sent down. were usedtoscoutareasfordebrisandstandbywhiledivers the applicationisTWAFlight800,whereNavy’sminirovers this wouldnotbeconsideredpartofitsmission.Anexample the capabilitytopickupsmallitems—10poundsorless—but meters. Itsmissionistolocateandinspect.TheROVmighthave generally nomorethan300metersbutcommonly150 Themaincomponentusedforbothloweringandhaul- Thespoolerisasubmersiblereelfittedwithliftcable Theliftcableisthelinethatattachedtowreck- Astheirnameindicates,thesevehiclesare TheHLSisamobilerecoverysystem Theheavemotioncom- TheHPUisaself-con- ISASI 2003 Proceedings • 169

ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS • • • location relativetothecrashposition • • search. our sonartechniciansinanalyzingtherecordsduring mation thatwillminimizethesizeofsearchareaandassist interviews asappropriate.Weareespeciallyinterestedininfor- gather everyscrapofavailableevidenceandconductpersonal estimate. effort, andplanningthetaskinordertodeveloparealisticcost ering anyandallavailablelossdatatodefinethescopeof events thatissetinplaceusuallyfollowsthesamepattern. office. Nomatterwhoinitiatestheinitialcontact,chainof ther directlybytheresponsibleagencyorthroughSUPSALV Diving (00C)foraction. CNO willturntotheOfficeofSupervisorSalvageand then tothefleet.Iffleetisunableperformtask, This requestwillgototheChiefofNavalOperations(CNO), pected equipmentfailure,orifthereareclassifiedassetsonboard. dent thatisnotcompletelyunderstood,iftherehasbeenasus- recovery taskingiftheyfeelthatthereissomeaspectofanacci- ron (owner)ortheNavalSafetyCenterwillrequestsearch-and- other entitytoassistintheeventualeffort. sumption thatwewilleitherbetaskedbytheU.S.Navyoran- ning amission,wetypicallystartcollectinglossdataontheas- accident. Evenbeforeweareformallyrequestedtostartplan- aware ofallaircraftthathavebeenlost,oftenwithinhoursthe even iftheinvestigatorbelievesitisinconsequential. we shouldknow.Weneedalloftheavailablelossinformation critical. Nevertakeforgrantedwhatyoutheinvestigatorthinks the actuallogisticsphaseofafast-responseeffort,informationis operation. ning whentimeiscriticalandensuresasmoothsuccessful and memoriesfadeorareinfluenced.Thisalsoallowsforplan- and planningeffortsbeginassoonpossiblebeforeevidence 170 • tors thatenterintothedevelopment ofthecostestimateinclude equipment, vessel,andlogisticsupportrequired.Additional fac- tional parameters,whichincludethetypeofsearch/recovery • • • • battery replacement • • • • • During theloss-analysiseffort,ourSearchProjectManagers In theearlystages,wearetypicallyfocusedontwothings:gath- For commercialaviationinvestigations,wearecontactedei- In thecaseofaU.S.militaryaircraftloss,aircraft’ssquad- Either bynewsreportsornotificationtheNavy,wearemade During operationalplanning,whichoftenisracingalongside This informationisusedbyPhoenixtodeterminetheopera- The pr V Radar data P T Size oftheair Survivor orbodyr Floating debrisr Floating debrisoroilslicksightingpositions Eyewitness r Local envir Depth ofwater—dictatestype ofequipmenttobeused R W V • oice communicationsbetweentheair ositions ofanyvesselsinthear essels ofopportunity ype ofair esident logisticalsupport eather ISA SI 2003 esence ofanacousticbeacon,itsfr craft onmental conditions eports ofthecrash P r craft andtype oceedings ecovery positions ecovery pick up positions ea oftheaccidentandtheir craft andA equency , lastdateof TC used duringthe sonarsearchisdetermined, andthesearcharea the rangescalehasbeenselected, thetracklinespacingtobe with thewater,waterdepth,andbottom terrain.Once clude theoverallsizeofaircraft, itspredictedimpactangle size andshapeofthedebris field. Factorstobeconsideredin- made basedontheexpected size ofthetargetandpredicted search willbeinitiatedforthe aircraft.Sonarrangeselectionis not heardduringitsanticipated30-daylife,asidescan sonar ing, themoretimeyousaveduringrecoveryphase. pinpoint thepinger’sposition.Themoretimeyouspendlocaliz- tracks aremadeinreciprocaldirectionstothispeaksignal to location recorded.Multipleparallelandperpendicularpasses/ phone anddirectionalarray.Peaksignalsarenotedtheir time duringthebattery’slife. location assoonpossibleinordertomaximizetheTPLsearch TPL ascomparedtothesidescansonar,itiscriticalarriveon acoustic pingersis30days.Giventhehighersearchrateof essence ingettingtothesearcharea.Typicalbatterylifeon mobilized. scan sonar,reducingtheamountofequipmentthatneedstobe mapping tool.TheTPLoperatesonthesamecableasside the primarysearchtool,withsidescansonarasabackupand outfitted withapinger,theTPLsystemisnormallyselectedas completed andanoperationsplanisfinalized.Iftheaircraft Once adecisionismadetogoaftertheaircraft,lossanalysis The mission entire aircraft. recovering acriticalpieceofwreckagetotherecoveryan sion objectivesandprioritieswillbe.Thesecanrangefrom tives. Ultimatelythesafetyinvestigatorsstatewhatmis- potentially theoutcomeofinvestigation. steps willhaveaqualitativeeffectontheoveralloperationand the leadinprovidinglocalonshore/offshorelogisticalsupport. tion, Phoenixisoftentaskedtocoordinatetheseeffortsandtake as wasthecaseonspaceshuttleColumbiaaccidentinvestiga- as aresultofthebrief. the civilagencymakingrequest.Ago/nogodecisionismade tomer representativetobriefingswiththefleetor times, aPhoenixProjectManagerwillthenaccompanythecus- cial customerrepresentativewhilereviewingtheestimate.Many • • • • • • • • • • • If apingerwasnotinstalledontheaircraftor was We searchuntilwehearthepingeronTPL’somnihydro- If anemergencypingerispresentontheaircraft,timeof The firstkeystep,however,istoestablishthemissionobjec- Fundamental stepsshouldbetakenpriortoanoperation.These In theeventthatamajorrequiresmulti-agencyeffort, The resultsoftheseeffortsarebriefedtotheNavyorcommer- Human r T Hazar Customs andvesselagents Air transportation T Cranes P V Deck equipment V ersonnel essel duration essel pr rucking ime criticalissues:car dous materialhandling eparation anduse emains go and techniquesthatwillpossibly aidtheinvestigators. tions basedonpastexperience inregardtooperationalscenarios a safeandorganizedmanner. Phoenixmaymakerecommenda- recovery, itisalwaysamission objectivetoconductarecoveryin cific tothetypeofrecovery.Outside oftheobviousobjective ment ofrecoverytoolsshould beincludedintheloadout,spe- available aboutthewreckageorsalvageeffort,afullcomple- ing optimalweatherwindows.Unlessveryspecificinformation is ation—larger moredifficultrecoveriesshouldbeconducted dur- for identificationofdebrisonthebottom. identify. Trainedaviationtechniciansareanimportantresource dent, especiallyahigh-energyimpact,targetmaybedifficult to als tomanufacturer’sinformationanddrawings.Afteran acci- crew. ThesesourcescanrangefromDepartmentofDefense manu- in whichitistobesalvagedandforrecognitionbytherecovery salvaged areexaminedforplanningpurposesboththemethod any andallavailablesourcesofinformationabouttheitemtobe investigation andthelocationofyettobeidentifiedtargets. vide adetailedmapofthedebrisfield,whichcanaidinboth identifiable pieces,i.e.,cockpit,engines,tail,etc.Thiswillpro- for specificwreckage.Navigationalfixesaretakenatallmajor low theROVtoeffectivelylookinhighestprobabilityareas for thewreckage.Thiswilleliminaterandomsearchingandal- boundaries ofthedebrisfieldandattempttoestablishapattern tive” positionsontargetsfortheROVtobeflownto. scanning sonar.Fromthesidescansonarsearch,wehave“rela- saves timeasopposedto“searching”forpieceswiththevehicle’s give thegeographiclocationofpiecesonseafloor.This GPS informationwiththerelativefromUSBLto gation softwareusedonthesearch.Thiscombines to followtheROV’spath.Itsoutputsarefedintosamenavi- an ultrashortbaseline(USBL)acoustictrackingsystemisused sides theexperienceofcrewissubseanavigation.Normally, such aswinches,motioncompensationequipment,andliftlines. phase wouldincludeanROVsystem,riggingandrecoveryassets spection-and-recovery phase.Thetypicalcomponentsofthis timely andsuccessfulrecoverymission. sonar recordswiththemandprovidingthecontinuitytoensurea fer totherecoverytask,bringingpositionalinformationand the searchgoes.Inthiscase,severalofteamwilltrans- team. Itsometimesismoreprudentto“waitandsee”howwell lized onlyafteraprobablecontacthasbeenlocatedbythesearch the piecescannotbepredicted,recoveryassetsmaymobi- ery team.However,ifthelossdataisweakorexpectedsizeof vessel. Ifthisisthecase,oursearchteamnowbecomesrecov- both thesearch-and-recoveryassetsaremobilizedonsame Under certaincircumstanceswhenlossdataisexceptionallygood, Recovery allows anacousticpicturetobemadeoftheentiredebrisfield. sonar isavaluabletoolformappingthedebrisfieldprecisely.This is methodicallycovereduntiltheaircraftlocated. The recoveryof wreckageisaccomplishedin avarietyofways, Any recoveriesshouldbeplannedwithweatherasaconsider- Before anysalvageoperationcommences,itisimportantthat Upon arrivingonscene,thefirststepwillbetoestablish Probably oneofthemostimportanttoolsinyourtoolboxbe- Once thetargetislocated,operationwillmovetoin- In theeventTPLhasdetectedpingers,sidescan who areableto provideapreciseposition,then asearchshould regarding thelossandtools used.Ifthereareeyewitnesses Search timeisdirectlyrelated tothequalityofinformation How longwillittaketofindit? that willbesafeforoperations. factor willbetheabilityof vesseltoholdstationinamanner the vesselandseverityof theseas.Inanycase,ultimate knots. However,theserulesofthumballdependonthesize ery ofalargeheavy,object,thisdropsto1meterandlessthan15 conditions is3-meterseasand29knotsofwindorless.For recov- the cabletovehicle.Theruleofthumbforoperationalweather the vesselbyitsumbilicalandanymotionmaybetranslated down ject torollingandpitchingbytheseas.TheROVisattached to more difficultintherespectthatvesselisstationaryand sub- back ondeck.However,forROV(recovery)operations,itisfar easy, andthemainconcernisgettingfishoverside or tion. TowingtheTPLorsidescansonarbehindvesselisfairly Weather islessofanissuewiththesearchportionanyopera- What typeofweathercanyouworkin? questions are: is asimportantunderstandingtheircapabilities.Sometypical Understanding thelimitationsofsearch-and-recoveryassets Difficulties action. is designedtodampenthesurgeofvesselcreatedbywave winch andrunthroughaheavemotioncompensationunitthat to therigging.Therecoverylinewouldbeattacheda be loweredtothebottom,andliftcablewouldconnected and connectedtothisrigging.Alternatively,thespoolerwould or clamps.Arecoverylinewouldbeloweredfromthesurface be riggedforrecoverybytheuseofchainornylonchokers,slings, separate recoverylineisusedfromthesurface.Thepiecewould estimate cannotbemadesuchasmajorsectionsoffuselage,a be calculatedsafelytoweighlessthan2tons. through framelift.Thismethodwouldbeusedforitemsthatcan items, mayberecoveredbelowtheROVwithamethodcalled fresh watertostopanyeffectsofcorrosion. further damage.Uponrecovery,theyareimmediatelyplacedin special recoveryboxesareoftenusedtominimizeanychanceof dividually andfirst.Extremecareistakenintheirrecovery, critical toanyinvestigationandwouldtypicallyberecoveredin- full orallitemsofinteresthavebeenrecovered. The basketisrecoveredundertheROVorseparatelywhenit can makenumerousshorttripstodeposititemsintothebasket. recovery basket.Thebasketisfarmoreefficient,sincetheROV surface withtheROVmanipulatorsorbyplacingobjectina size—are recoveredbytheROVeitherliftingobjectto tified wreckage. “shopping list,”theROVwillproceedwithrecoveryofiden- sentative hasmadeafinaldeterminationonwhathewantshis ner toconductourrecoveries. eration astowhatwillbethemosteconomicalandefficientman- which aretailoredtotheactualsituation.Wealwaysgiveconsid- For largerstructuresweighingmorethan2tonswherean Larger items,i.e.,structuressuchaswings,cockpits,orlike The oneexceptiontotheruleareblackboxes.These Small items—typically200lbsorlessandrelativelysmallin After thedebrisfieldhasbeenmappedandclientrepre- ISASI 2003 Proceedings • 171

ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS bottom. extent, butthebestsolutionisnottocomeincontactwith cameras. Bottomcurrentwillnegatethisproblemtoacertain bottom, sedimentisstirredup,whichusuallyobscurestheROV’s around alargedebrisfield. ROV downwhentravelingoverlongdistancessuchasflying tance theROVhastotravel,greaterdrag.Thisslows water column.Thedeepertheoperationandlongerdis- and acousticnavigation. clear water.Thisonlyunderscorestheneedforgoodplanning Line ofsightfortheROVcamerasrangesfrom5to25metersin night withapenlightlookingforfoilwrapper.Itisnoteasy. the equivalentofsearchingafootballfieldonfoggy,moonless The bestanalogyIhaveheardisthatsearchingwithanROV What arethelimitationsofROV? riving onbottom. to acquirethetargetquickly,oftenwithinjustafewhoursofar- ing accuratepositionsofsonartargets,theROVshouldbeable of itsownwiththeonboardsonar.Withskillednavigatorsprovid- on thewreck,butwilltraveltobottomandconductasearch calculated position.InmostcasestheROVwillnotdropdirectly that whenthefirstdiveismade,itmadebasedonarelative be asmuch6miles.Thepositionofthefishiscalculatedso sonar fishistrailingbehindthevessel.Indeepwaterthiscould The sidescansonarwillgivegooddata;howeverthe position isgiven? How longwillittaketofindthetargetwithROVafterasearch faster thanasidescansonar. with anacousticpinger,sincetheTPLcansearchmanytimes cases, thesearchtimeswillbelowestifaircraftisoutfitted is available,itimpossibletopredictsearchtime.Inallofthese quick—several daysorless.Inthecasewherelittlenolossdata of land,butwithinreliableradartrackingrange,couldalsobe be relativelyquick—24hoursorless.Anaircraftlostout-of-sight 172 Finally, goodpilotingiscritical.AnytimetheROVsitson The ROVispullingaroundanumbilicalthathasdraginthe • ISA SI 2003 P r oceedings mistakes orresults. for consistentexpectationsandnothavingtorelearntechniques, sons learned”fromeachofyourinvestigations.Thiswillprovide define youroptions. cies. VisitingthePhoenixfacilityinLandovercanhelptobetter specific capabilitiesandexpertisethatareavailabletotheiragen- Island; CH-46,5000fsw,Somalia;SAA747,14350Mauritius. world andbroughthometheaircraft,i.e.,CH-46,17400fsw,Wake real time,onscene!Wehavemobilizedtothefarcornersof (oil field).Wecantellanaircraftdebrisfieldfromoysterbed, for apipelineorcableroute,notplacingtemplatesBOPstacks on aircraftsearchandrecovery—notsimplemapping,surveying serve you.Phoenixpersonnelhavespenttheircareersfocused ROV contractors,butfewhavethespecializedexperiencetobest nology, logisticsandmarineoperations.Therearemanydiving/ ists whohavemanyyearsofexperienceindiving,remotetech- These typesofinvestigationsrequiretheparticipationspecial- Conclusion power tominimizethis. and does.Breakdownsdooccur,butweeverythinginour ment mayberequiredtooperatecontinuouslyfordaysatatime nance hoursanaircraftrequiresperflighthour.Oursubseaequip- wear andtear.Thinkforamomentofthenumbermainte- order; however,likeanycomplexdevice,theROVissubjectto attention. Weworktoensurethatallofourequipmentisingood This equipmentisworkingunderharshconditionsandrequires Equipment breakdowns up piecesthendiscardingthemshortlythereafter. mended. Inordinateamountsoftimehavebeenspentpicking the surveyplanofmappingdebrisfieldishighlyrecom- cover aspecificpieceimmediatelyupondiscovery,holdingto Unless theinvestigatorisabsolutelypositivethathewantstore- Can wepickupthatpiece? I wouldliketochallengeyourcommunityarchivethe“les- It isessentialforemergencyresponseteamstobeawareofthe N the baseof flames.” TheBoarddetermined thattheprobable for theextinguishingagentto beeffective,itmustappliedto tory “hadlittleornoeffecton thefire,”notingthat“[i]norder attendant’s dischargeofafire extinguishingagentintothelava- and diedinthefire.Theairplane wasdestroyed. crewmembers onboard,23passengers wereunabletoevacuate ated anemergencyevacuation.Ofthe41passengersand five in thecabin.Afterairplanelanded,flightattendants initi- emergency descentwasinitiated. then informedthecaptainthatthey“bettergodown,”and an he instructedtheflightattendantsnottoopenit.Thefirstofficer assess thesituation,hefoundthatlavatorydoorwashot, and He thenclosedthedoor.Whenfirstofficercamebackto lavatory, aimingatthepanelingandseamstrash bin. ond flightattendantdischargedaCO tory. Althoughneitherflightattendantsawanyflames,thesec- coming fromtheseamsbetweenwallsandceilinginlava- of thelavatories.Anotherflightattendantsawthatsmokewas noticed astrangesmellandflightattendantsawsmokeinone Covington, Ky.Thefirewasinitiallydetectedwhenapassenger cinnati andNorthernKentuckyInternationalAirport)in the GreaterCincinnatiInternationalAirport(sincerenamedCin- experienced aninflightfireandmadeemergencylandingat Douglas DC-9,C-FTLU,operatedbyAirCanadaasFlight797, On June2,1983,about1920EasternDaylightTime,aMcDonnell Air CanadaFlight797accident are alsodiscussed. issued in2002.TherecommendationsandFAAresponsestothem incidents thatinitiatedfivesafetyrecommendationswere more than20yearsago,andfocusesonrecentaccidents chronicles theSafetyBoard’sactivitiesinthisarea,whichbegan T orally presented.—Editor] paper servedasanalternateselectionfortheseminarbutwasnot Survival FactorsInvestigatorintheOfficeofAviationSafety.[This PHOTO NOT AVAILABLE In itsfinalreport,theSafetyBoard determinedthattheflight During thedescent,smokeincreasedandmovedforward Board RecommendationsRelatingto inflight firesoncommercialairplanes.Thispaper tory ofadvocatingsafetyimprovementspertainingto he NationalTransportationSafetyBoardhasalonghis- National TransportationSafety years. GeorgejoinedtheNTSBin1998asa CAMI’s CabinSafetyWorkshopprogramfor12 projects. Inaddition,heservedascoordinatorfor aircraft evacuationandwatersurvivalresearch 1982. Forthenext16years,heworkedonnumerous FAA CivilAeromedicalInstituteinOklahomaCity Mark George Inflight FireEmergencies beganhiscareerasastatisticianatthe 2 fireextinguisherintothe By MarkGeorge plained that14 CFR121.417requiredcrewmembers tobetrained appropriate handfireextinguisher onanactualfire.” risk toessentialaircraftcomponents, andthedischargeofan of thefirethroughinteriorpanels thatcanbepenetratedwithout ing equipment,theuseof fireaxtogainaccessthesource including hands-on-trainingin thedonningofprotectivebreath- of thepropermethodsaggressivelyattackingacabinfire by for bothflightcrewsandattendantstobeknowledgeable the location,source,andseverityoffireorsmokewithin cabin; attendants torecognizetheurgencyofinformingflightcrews of termined orifimmediateextinctionisnotensured;for flight source andseverityofthefirearenotpositivelyquickly de- to beginanemergencydescentforlandingorditching ifthe determine thesourceandseverityofanyreportedcabinfire and ments forflightcrewstotakeimmediateandaggressiveaction to necessary specifythattheybeamendedtoemphasizerequire- review thetrainingprogramsoftheirrespectivecarriersandif the FAA“requirethataircarrierprincipaloperationsinspectors also issuedRecommendationA-84-76,whichrecommendedthat lent astheextinguishingagent. stalled intheairplane,atleasttwocontainhalon1211orequiva- receptacle; and,oftherequiredhand-heldfireextinguishersin- extinguisher thatdischargesautomaticallyifafireoccursinthe tendant; eachlavatorytrashreceptacleisequippedwithafire passenger cabinthatwouldbereadilydetectedbytheflightat- vides awarninglightinthecockpitoranaudio lavatory andgalleyhasasmokeorfiredetectorsystemthatpro- of FederalRegulations rulemaking torequirethatallairplanesoperatedunder14 able AlternateAction”onJan.15,1986,aftertheFAAcompleted three oftheserecommendationswereclassified“Closed—Accept- planes useatechnologicallyadvancedagent,suchashalon.All erated fireextinguisherscarriedaboardtransport-categoryair- tion A-83-72,whichaskedtheFAAtorequirethathand-op- to andinlavatorywastereceptacles;SafetyRecommenda- require theinstallationofautomaticfireextinguishersadjacent ries; SafetyRecommendationA-83-71,whichaskedtheFAAto the FAAtoexpediteactionsrequiresmokedetectorsinlavato- (FAA), includingSafetyRecommendationA-83-70,whichasked several recommendationstotheFederalAviationAdministration emergency descent.” accident wastheflightcrew’sdelayeddecisiontoinstitutean tion providedtothecaptain.Contributingseverityof derestimate offireseverity,andconflictingprogressinforma- cause oftheaccidentwas“afireundeterminedorigin,anun- In itsNov.2,1984,responsetotheSafety Board,theFAAex- In itsfinalreportontheAirCanadaaccident,SafetyBoard As aresultoftheAirCanadaaccident,SafetyBoardissued (CFR) Part121beequippedasfollows:each 1 ISASI 2003 Proceedings • Code 173

ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS instructed her not tousethehalonextinguisher, indicatinghe halon intotheventwhereshe hadseentheglow.Thecaptain crew oftheseobservationsand askedthecaptainwhethertospray beneath theventatthatlocation. attendants alsoreportedseeing anorangeorred,flickeringglow the rightsidewallandabove the floorvent,wasscorched.Flight carry-on bag,whichhadbeenonthefloorbesidehim,next to other rowwhenhestatedthathisfeetwerehot.Thisindividual’s cabin. FlightattendantsreseatedapassengerinRow11 to an- flight attendantsnoticedsmokeintheforwardsectionof coach off-duty flightattendantsretrievedhalonfireextinguishers when lavatories butwereunabletolocatethecauseofsmell. Two Following thecaptain’sinstructions,flightattendantschecked the phurous or“litmatch”smellandreportedittotheflight crew. injured. dants, threeoff-dutyflightattendants,or113passengerswere age, andnoneofthetwoflightcrewmembers,threeatten- evacuation wasperformed.Theairplanesustainedminordam- national AirportinCovington,Ky.Afterlanding,anemergency gency landingattheCincinnatiandNorthernKentuckyInter- as Flight2030,experiencedaninflightfireandmadeemer- McDonnell DouglasMD-88,N947DL,operatedbyDeltaAirLines On Sept.17,1999,about2230EasternDaylightTime,a Delta AirLinesFlight2030 training stillexisted. airplanes. Subsequenteventsillustratedthattheneedforsuch ing crewmemberstorecognize,locate,andfighthiddenfireson additional advisorymaterialemphasizingtheimportanceoftrain- locating andsuppressingfires.However,theFAAdidnotissue tinguishers haveprovidedcrewmemberswithbettermethodsof ticular, requirementsforsmokedetectorsandhalon-typefireex- sulted insomechangesthatimprovedaircraftfiresafety;par- continue tovoiceourconcerninfutureaccidentinvestigations.” cern forthesafetyissueinvolvedhasnotdiminished,andwewill that, “[a]lthoughwehaveclosedthisrecommendation,ourcon- mendation A-84-76as“Closed—UnacceptableAction,”stating Board disagreedandonMay12,1986,classifiedSafetyRecom- book,” and theguidanceinAirCarrierOperationsInspector’sHand- 121.417, thevariousAirCarrierOperationsBulletins(ACOBs), letter, theFAArespondedthat“duetorequirementsof14CFR the aircraftshouldbelandedimmediately.InitsMarch7,1986, immediately identifiedorcannotbeextinguishedimmediately, programs shouldemphasizethatifthesourceofafirecannotbe den fires.”TheBoardalsoreiterateditsbeliefthatcrewtraining prepare crewstoassesseffectivelythehazardoforfighthid- posed’ fires,whicharerelativelyeasytocontrol.Thisdoesnot “current firefightingtrainingisdirectedprimarilytoward‘ex- In itsApril12,1985,letter,theBoarddisagreed,statingthat riers isatestimonytotheadequacyofcurrentregulations.” lations wereadequate,statingthat“thesafetyrecordofU.S.car- the crewmemberistoserve.”TheFAAconcludedthatregu- ing initialandrecurrenttrainingforeachtypeaircraftinwhich gency drillsand“actuallyoperatetheemergencyequipmentdur- for fireemergencies,andfurtherrequiredthemtoperformemer- 174 Flight attendantNo.1wentto thecockpittoinformflight Shortly aftertakeoff,severalflightattendantsdetecteda sul- The FAA’sresponsetotheAirCanadarecommendationsre- • ISA 2 furtheractionbytheFAAwasunwarranted.TheSafety SI 2003 P r oceedings experienced an inflightfirethatbeganshortly aftertakeofffrom American AirlinesasFlight 1683, wasstruckbylightningand McDonnell DouglasDC-9-82 (MD-80),N3507A,operatedby On Nov.29,2000,about 1753 EasternStandardTime,a American AirlinesFlight1683 cue andfirefightingpersonnel. evacuated. Thefirewaseventuallyextinguishedbyairport res- ued burningaftertheairplanelandedandpassengers were captain’s seat.Thearcingignitedinteriorpanels,whichcontin- cabin wascausedbyelectricalarcinginthebulkheadbehind the oxygen.” cause hewasconcernedthatthehalon“wouldtakeaway more ered usingahalonfireextinguisher,butdecidedagainst itbe- to fight.” to usethehalonfireextinguisherbecauseshe“didnotsee afire fire extinguisherbutwasunsurewheretoaimit.Shedecidednot told investigatorsthatshe“debatedwhethertousethehalon” ing” andheard“poppingnoises”atthefrontofcabin.She attendant No.1sawalargeamountofelectrical“arcingandspark- use anyofthefirefightingequipmentavailabletothem.Flight attendant madeanyefforttolocatethesourceofsmokeor she couldnolongerseetheforwardgalley.However,neitherflight ley areaaroundtheirjumpseats. ness classbecauseoftherapidlyaccumulatingsmokeingal- flight attendantNo.2reseatedthemselvesinemptyseatsbusi- She closedthecockpitdoorandreturnedtocabin. The captaintoldherthattheywerereturningtoGreensboro. where” andnoticedthatthecrewhaddonnedtheiroxygenmasks. attendant No.1wenttothecockpit,whereshesawsmoke“every- were seatedontheforwardjumpseat,bothsmelledsmoke.Flight passengers receivedminorinjuriesduringtheevacuation. received minorinjuriesfromsmokeinhalation,andeightother crewmembers onboard,threeandtwopassengers the effectsoffire,heat,andsmoke.Of57passengersfive was performed.Theairplanesubstantiallydamagedfrom national AirportinGreensboro,N.C.Anemergencyevacuation an emergencylandingattheGreensboroPiedmont-TriadInter- (AirTran) asFlight913,experiencedaninflightfireandmade Douglas DC-9-32,N838AT,operatedbyAirranAirways On Aug.8,2000,about1544EasternDaylightTime,aMcDonnell AirTran Flight913 fire thatwasgrowinginsize. the blanket,andsmolderinghadbecomeaself-sustaining static portheater.Electricalarcingfromtheheaterhadignited ing insulationblanketinthecargocompartmentadjacenttoa covered thatthesourceofsmokeincabinwasasmolder- charged becausethecaptainhadinstructedhernottodoso. she becamealarmedthatahalonfireextinguisherhadbeendis- turn. WhenflightattendantNo.1returnedfromthedeck, visible. Thereafter,thesmokebegantodissipateanddidnotre- tinguisher intotheventandobservedthatglowwasnolonger another flightattendanthadalreadydischargedahalonfireex- was concernedaboutsprayinghaloninthecabin.Meanwhile, Preliminary findingsindicatedthatthesmokeinforward Flight attendantNo.1reportedthatthesmokebecamesodense Shortly aftertakeoff,flightattendantsNo.1and2,who During itsinvestigationofthisincident,SafetyBoardstaffdis- 4 Anoff-dutyAirTranpilotseatedinfirstclassconsid- 3 clared anemergency anddivertedtoSLC. cockpit toletthemknowwhat washappening.Thecaptainde- see fireagain.Athirdflightattendant calledthepurserand halon untiltheextinguisher was empty.Afterthat,shedidnot into theareaaroundwater heaterandcontinueddischarging handed ittoher.Sheplacedthenozzleof fireextinguisher ment. Anotherflightattendant retrievedafireextinguisherand ing thewaterheater,butdidnotextendoutsidecompart- heater nearthesinkandsawflames.Theflamesweresurround- “bubbling.” Theflightattendantopenedthedoorto water plastic,” butsawnosmoke.Shepartoftheplastictoiletshroud dant wentintothelavatoryandsmelledanodorof“burning dant thatit“smelledlikesmoke”inthelavatory.Theflight atten- went intotheaftlavatory,anduponexiting,toldaflight atten- two flightcrew,fiveattendants,or133passengers. Airport (SLC),SaltLakeCity,Utah.Therewerenoinjuries tothe eventful landingwasmadeattheSaltLakeCityInternational The firewasextinguished,anemergencydeclared,andun- gan Airport(BOS),Mass.,experiencedafireintheaftlavatory. San FranciscoInternationalAirport(SFO),Calif.toBostonLo- 757-222, operatingasUnitedAirlinesFlight32enroutefrom On Jan.11,2003,at0045MountainStandardTime,aBoeing United AirlinesFlight32 els, whichignitedadjacentmaterials. caused arcingintheairplanewiringabovecabinceilingpan- “use itifwasneeded.”However,thesmokedidnotrecur. halon fireextinguisher,instructedhimonitsuse,andtoldto ing, anotherflightattendantgavethepassengerinSeat9Ea be diminishing.Beforetakingherseatfortheemergencyland- hole, assessedtheresults,andfoundthatsmokeappearedto No. 1thenfullydischargedahalonfireextinguisherintothe hole intheblisteredareaofceilingpanel.Flightattendant ceiling panel.Apassengerproducedaknifeandcutcircular passengers ifanyonehadaknifethatcouldbeusedtocutthe ward theblisteredceilingpanel.FlightattendantNo.1asked ing panelaboveRow9begantoblisterandturnyellow. sounded. Thesmokeworsenedinthemidcabinarea,andaceil- near Rows7and8.Thesmokedetectorsintheaftlavatories trieved halonfireextinguishersandbroughtthemtothearea notified theflightcrewwhileothertwoattendantsre- ceiling panelsaboveRows7and8.Shewenttothecockpit gers, sheobserved“dark,dense,black”smokecomingfromthe the circuitbreaker,andsmokestoppedcomingoutoffixture. told herto“pullthebreaker”forfluorescentlight.Shepulled area. Sheshutthelightoffandcalledcockpit.Thecaptain smoke comingfromafluorescentlightfixtureintheforwardentry dant No.1,whowasseatedontheforwardjumpseat,sawwhite and heardaboomontherightsideofairplane.Flightatten- and 61passengerswereinjured. minor damage.Noneofthetwopilots,threeflightattendants, ation atDullesInternationalAirport.Theairplanesustained performed anemergencylandingandorderedapassengerevacu- Reagan NationalAirportinWashington,D.C.Theflightcrew About anhourand20minutesintotheflight,apassenger The SafetyBoardinvestigationconcludedthatalightningstrike A flightattendantbegandischargingahalonextinguisherto- When flightattendantNo.1wentafttocheckonthepassen- After takeoff,thethreeflightattendantssawaflashoflight in acabin-level compartmentusingfirefighting equipmentiden- trained crewmembersattempted toextinguishcargofireslocated adequate. indicated thattheFAA’scurrent trainingrequirementswerein- extinguish cargofiresinsmall ClassBcargocompartmentsalso Technical Centertoevaluate the abilityofflightattendantsto the panelsbeforetheycanusefireextinguisherseffectivel y. moving interiorpanelsorotherwiseaccessingtheareasbehind to quicklyidentifythelocationoffire,whichmayrequire re- in itsrecommendationletterthatcrewmembersshouldbetrained heat fromhiddenlocations.Therefore,theSafetyBoard noted ible, localizedflames,butinlessobviousways,suchassmoke or fires oncommercialairplanescanpresentthemselvesnot as vis- and incidentdescriptionsinthispaperdemonstratethatinflight an actualfireconsistingofavisible,openflame.Theaccident a firefightingdrillinwhichstudentsarerequiredtoextinguish tial trainingprogramforflightattendants,example,includes inherent infightingahiddenfireonanairplane.AirTran’sini- effort toextinguishandthatdoesnotdemonstratetheproblems ing initialtrainingistypicallyanopenflamethatrequireslittle carriers andfoundthattheactual“fire”crewmembersfightdur- tors evaluatedthefirefightingtrainingprogramsofseveralair access totheareabehindinteriorpanels.SafetyBoardinvestiga- to identifythelocationofahiddenfireorknowhowgain fires, theydonotexplicitlyrequirethatcrewmembersbetrained fied in14CFR121.417requireinstructionfightinginflight fight anactualorsimulatedfireduringinitialtrainingonly. tual orsimulatedfire.Asaresult,crewmembersarerequiredto regulation doesnotrequirerecurrenttraininginfightinganac- hand-operated fireextinguisherfoundontheirairplanes;the ing onethatdemonstratestheirabilitytooperateeachtypeof form certaindrillsbiannuallyduringrecurrenttraining,includ- drill that isappropriateforthetypeoffirebeingdemonstratedin while fightinganactualfire each crewmembermustaccomplishaone-timeemergencydrill occur inflightorontheground.Aspartoftheirinitialtraining, instruction onhandlingemergencysituationsincludingfiresthat type ofextinguishertobeusedfordifferentclassesfires,and operation ofportablefireextinguishers,withemphasisonthe instruction on,amongotherthings,thelocation,function,and fires. Theregulationspecifiesthatairlinesmustprovideindividual ing onfirefightingequipmentandproceduresforfightinginflight Title 14CFR121.417requiresthatcrewmembersreceivetrain- Crewmember training extinguishers. cess toareasbehindinteriorpanels,andthemeritsofhalon mendations totheFAApertainingcrewmembertraining,ac- accidents andincidents, charred andmelted. aft wallofthetoiletshroud,inareaoversensor,wasalso tory toiletwaterlevelsensorwascharredandmelted.Theinside ing. Preliminaryexaminationoftheairplanerevealedlava- The resultsofaseriesexperiments conductedbytheFAA Further, althoughtheemergencytrainingrequirementsspeci- Although 14CFR121.417alsorequirescrewmemberstoper- Based oninformationgainedthroughinvestigationofthese As ofthedatethispaper,investigationeventisongo- 7 TechnicalCenterstaffconducted 13testsinwhich 5 theSafetyBoardissuedseveralrecom- 6 usingthetypeoffireextinguisher ISASI 2003 Proceedings • 175

ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS In twooftheoccurrences describedinthispaper, crewmembers in fightinginflightfires Properties ofhalonandthe meritsofhalonextinguishers ing theproblemofgainingaccess toareasbehindinteriorpanels. through thespacearound waterheater. visible inthearea,andhaloncouldbeindirectlyappliedto fire of thefiresourcetowaterheaterwassuchthatflames were attempt tolocatethesourceofodor.Fortunately,proximity flight attendantopenedtheaccessdoortowaterheater inan for cuttingortocuttheholeitself.InUnitedincident, the on theactionsofapassenger,eithertoprovidesharpinstrument cess totheareabehindpanelshouldnothavebeendependent the flightattendant’sactionsuccessfullyextinguishedfire, ac- that mayhavebeencoveredbythepaneling.Inaddition,although the ceiling)riskeddamagetoelectricalwiringandother cables panel, butthemethodused(i.e.,havingapassengercutholein dent, theflightattendantsdidaccessareabehindceiling was notappliedtothebaseofflames.InAmericaninci- action hadlittleeffectonthefirebecauseextinguishingagent the wallsandceilingwheresmokehadbeenobserved.This a CO fires. IntheAirCanadaaccident,oneflightattendantdischarged are abletoeasilyandquicklylocateextinguishhiddeninflight Interior panelsofairplanesarenotdesignedsothatcrewmembers Access toareasbehindinteriorpanels charging halonintotheareasurroundingwaterheater. behind anaccessdoor.Shethenextinguishedthefirebydis- began searchingforthesourceofodor,anddiscoveredflames agent. IntheUnited32incident,flightattendantquickly cutting aholeintheoverheadpanelandapplyingextinguishing working withapassenger,successfullyextinguishedthefireby trast, intheAmericanFlight1683incident,aflightattendant, use anyofthefirefightingequipmentavailabletothem.Incon- attendants madenoefforttolocatethesourceofsmokeor sible lossoftheairplane.InAirTranFlight913accident,flight and couldhaveresultedindeathorseriousinjury,aswellpos- guisher hadbeencarriedout,thefirewouldlikelyhaveprogressed response. Further,ifthecaptain’sordernottousefireextin- tinguisher. Thiscourseofactiondelayedanimmediatefirefighting asked forthecaptain’spermissionbeforedischargingafireex- panels. IntheDeltaFlight2030incident,flightattendant or becausetheycouldnotaccesstheareabehindinterior to theflames,eitherbecausetheyhadnotbeentraineddoso dent, flightattendantsdidnotapplyextinguishingagentdirectly dent andothereventscitedinthispaper.IntheAirCanadaacci- fighting inflightfires,asdemonstratedintheAirCanadaacci- may failtotakeimmediateandaggressiveactioninlocating on aconcernthat,asresultoflimitedtraining,crewmembers effectively fightinflightfires.” training procedureswouldbetterprepareflightattendantstomore fires. Thereportconcludedthat“improvedandmorerealistic act quicklyoraggressivelyenoughtosuccessfullyextinguishthe ing propertechniques,inmostcasesthecrewmembersdidnot noted that,althoughthefirescouldhavebeenextinguishedus- tical tothetypesonwhichtheyhadbeentrained.Thereport 176 The SafetyBoardissuedonerecommendation aimedataddress- The SafetyBoard’srecommendationsinthisareawerebased • 2 ISA extinguisherintothelavatory,aimingatseamsbetween SI 2003 P r oceedings Amend 14 corporated intocrewmember trainingprograms.(A-01-84) tents oftheadvisorycircular(recommendedinA-01-83)are in- Require principaloperationsinspectorstoensurethatthe con- agents tothesourceoffire.(A-01-83) hind interiorpanelsinordertoeffectivelyapplyextinguishing crewmember toremoveorotherwisegainaccessthearea be- are hiddenbehindinteriorpanelsandthereforemayrequire a to signsofaninflightfire.TheACshouldstressthatfires often crewmembers totakeimmediateandaggressiveactioninresponse Issue anadvisorycircular(AC)thatdescribestheneed for the SafetyBoardissuedfollowingrecommendationstoFAA: On thebasisofconcernsdiscussedabove,onJanuary4,2002, Recommendations ing crewmembers’understandingofthebenefitshalongas. graph warningagainstexposuretohalongas. fire. Indeed,thestatementquotedaboveisburiedinpara- when comparedtothepotentiallydevastatingeffectsofaninflight posed byhalongasareoveremphasizedintheAC,especially mal decompositionproductsfromhalon,”thepotentialhazards heat, andoxygendepletion,createagreaterhazardthanthether- products fromthefireitself,especiallycarbonmonoxide,smoke, tinguisher inatransport-sizedcabin. levels cannotbeachievedbydischargingasinglehand-heldex- craft. However,itisnotobviousfromtheACthatmaximum ventilated andnon-ventilatedpassengercompartmentsonair- agents undervariousconditionsthatshouldnotbeexceededin agent ortothedecompositionproductsshouldbeavoided.” However, unnecessaryexposureofpersonneltoeitherthenatural jected toflameorhotsurfacesatapproximately900ºF(482ºC). agents willalsodecomposeintomoretoxicproductswhensub- depending uponexposureconcentrationandtime.Halogenated genated agentsmayproducevariedcentralnervoussystemeffects reduced mentalsharpness..Exposuretoundecomposedhalo- halon vaporsmayresultindizziness,impairedcoordination,and following: “Testsindicatethathumanexposuretohighlevelsof do notdegradevisualacuity.However,AC20-42Calsostatesthe surfaces, havefeweradverseeffectsonelectronicequipment,and no chemicalresiduetocontaminateorcorrodeaircraftparts gaseous dischargeandthereforeamorelimitedthrowrange,leave extinguishers withthesameweightofextinguishingagent,havea that halon-typeextinguishersarethreetimesaseffectiveCO that it“wouldtakeawaymoreoxygen”fromthecabin. chose nottousethehalonextinguisherbecauseofhisconcern in thecabin.InAirTranaccident,anoff-dutycrewmember extinguisher becausehewasconcernedabouthalonbeingsprayed captain specificallyorderedaflightattendantnottousethehalon hesitated tousehalonextinguishers.IntheDeltaincident, of, locating,and fightinghiddenfires.(A-01-85) drills includerealisticscenarios onrecognizingpotentialsigns during crewmemberrecurrent trainingandtorequirethatthose pation infirefightingdrillsthat involveactualorsimulatedfires The SafetyBoardissuedonerecommendationaimedatimprov- Even thoughtheACalsostates,“generally,decomposition The ACalsospecifiesmaximumconcentrationlevelsforhalon FAA AC20-42C, Code ofFederalRegulations Hand FireExtinguishersforUseinAircraft 121.417 torequirepartici- , states 2 ceptable Response.” dum, SafetyRecommendation A-01-84wasclassified“OpenAc- 28, 2002,statingthat,pending theissuanceofmemoran- 8400.10, AirTransportationOperationsInspector’s Handbook.” carriers andwillbeincorporated intothenextversionofOrder tracking codesforthepurpose ofdeterminingnotificationtoair rier trainingprogram.Thismemorandumwillinclude PTRS the informationcontainedinACintotheirapprovedair car- rected tostresstheiroperatorstheimportanceofincluding carriers oftheavailabilityAC.ThePOIswillalso be di- inspectors (POI)directingthemtoinformtheirrespective air -87, theFAAwillsendamemorandumtoitsprincipaloperating issued inresponsetoSafetyRecommendationsA-01-83,-85, and ing (A-01-84),theFAArespondedbystating:“Once ACis ceptable Response.”Todate,theFAAhasnotissuedAC. ance oftheAC,recommendationswereclassified“OpenAc- June 28,2002,statingthat,pendingthedevelopmentandissu- quickly intoaircarriertrainingprograms.” proach willallowthesafetyinformationtobeimplementedmore hind interiorpanels.TheFAAbelievesthatthisalternateap- nizing potentialsignsofcabinfireandlocatingfireshiddenbe- ing fires,includingsimulatedfiredrillswithemphasisonrecog- conducting recurrenttrainingofflightcrewmembersincombat- February 2003.TheACwillalsoaddresstheeffectivemeansfor extinguishing fires.ItisanticipatedthatthisACwillbeissuedby on passengersversusitsoverridingbenefitofcombatingand properties ofhalonandemphasizeitsnegligibleharmfuleffects application ofextinguishingagents.TheACwillalsoaddressthe hidden behindinteriorpanelsandthetechniquesforeffective an inflightfireandaddresstheimportanceofinvestigatingfires The ACwillemphasizetheneedforanimmediateresponseto efforts outlinedinresponsetoSafetyRecommendationA-01-86. ommendations. TheACwillincludeguidancefromtheresearch lar (AC)toaddressthesafetyissuesidentifiedintheserec- intent oftheserecommendationsandwillissueanadvisorycircu- (A-01-83 andA-01-87)theFAAstated:“Theagreeswith mendations. On March8,2002,theFAArespondedtoeachoftheserecom- FAA responsestotherecommendations sively. (A-01-87) to thesafetybenefitsachievedbyfightinginflightfiresaggres- harmful effectsonpassengersandcrewarenegligiblecompared plain thepropertiesofhalonandemphasizethatpotential tions inspectorstoensurethataircarriertrainingprogramsex- Issue aflightstandardshandbookbulletintoprincipalopera- behind interiorpanels.(A-01-86) access panels,orsomeothermeanstoapplyextinguishingagent ping interiorpanelsofnewandexistingairplaneswithports, part ofthiseffort,theFAAshouldevaluatefeasibilityequip- the purposeofapplyingextinguishingagenttohiddenfires.As crewmembers togainaccessareasbehindinteriorpanelsfor modifications thatwillprovidethemosteffectivemeansfor Develop andrequireimplementationofproceduresorairplane The SafetyBoardresponded totheletterfromFAAonJune Regarding therecommendationtoimprovecrewmembertrain- The SafetyBoardrespondedtotheletterfromFAAon Regarding therecommendationsforissuanceofguidance tinguishing ininaccessible areas.TheFAAhas alsodevelopednew consider theapplicabilityofthese methodsoffirefightingandex- Once theresultsofresearch havebeenassessed,theFAAwill as ameansofeffectivelyfighting inflightfiresininaccessibleareas. handheld fireextinguishersin combination withaccesspanels/ports expanding thecurrentresearch toincludethestudyofusing safety recommendation,theFAA andtheTechnicalCenterwillbe for dealingwithinflightfiresindifferentareas.Inresponse to this explore thevariousareasofaircraftandanalyzetechniques detection andinflightfirefightingtechniques.Theresearch will the feasibilityofdevelopingmethodstoimprovemeans of Systems FireProtectionWorkingGroup,iscurrentlyresearching ter). TheTechnicalCenter,inassociationwiththeInternational through theWilliamJ.HughesTechnicalCenter(Cen- suppression issues.TheFAAhasinitiatedresearchprograms initiated effortstoaddressinflightfireaccessibility,detection, and feasibility offightingfiresbehindinteriorpanels.TheFAA has stating: “TheFAAagreesthatthereshouldbeanevaluationofthe to areasbehindinteriorpanels(A-01-86),theFAArespondedby classified ‘OpenUnacceptableResponse.’” change toSection121.417,SafetyRecommendationA-01-85is reconsider itsdecisionnottoreviseSection121.417.Pendingthe this goalwillnotbeaccomplished.TheBoardaskstheFAAto craft. TheBoardbelievesthatwithoutachangetoSection121.417, rent trainingusinghandheldfireextinguisherscarriedonair- crewmembers tofightanactualorsimulatedfireduringrecur- The intentofSafetyRecommendationA-01-85istorequireall quired toperformafirefightingdrillduringrecurrenttraining. during recurrenttraining.However,crewmembersarenotre- that heorshecanoperateeachtypeofinstalledfireextinguisher Section 121.417(2)(B)requireseachcrewmembertodemonstrate using atleastonetypeofinstalledhandheldfireextinguisher. firefighting drillinwhichthecrewmembercombatsanactualfire 121.417(1)(ii) statesthatinitialtrainingmustincludeanapproved ommendation A-01-85.TheBoardnotesthat14CFRSection the proposedACwilladequatelyaddressintentofSafetyRec- 28, 2002,statingthat“TheSafetyBoarddoesnotbelieve more quicklyintoaircarriertrainingprograms.” approach willallowthesafetyinformationtobeimplemented den behindinteriorpanels.TheFAAbelievesthatthisalternate on recognizingpotentialsignsofcabinfireandlocatingfireshid- in combatingfires,includingsimulatedfiredrillswithemphasis means forconductingrecurrenttrainingofflightcrewmembers issued byFebruary2003.TheACwillalsoaddresstheeffective ing andextinguishingfires.ItisanticipatedthatthisACwillbe ful effectsonpassengersversusitsoverridingbenefitofcombat- address thepropertiesofhalonandemphasizeitsnegligibleharm- for effectiveapplicationofextinguishingagents.TheACwillalso gating fireshiddenbehindinteriorpanelsandthetechniques sponse toaninflightfireandaddresstheimportanceofinvesti- A-01-86. TheACwillemphasizetheneedforanimmediatere- search effortsoutlinedinresponsetoSafetyRecommendation issues identified….TheACwillincludeguidancefromthere- … andwillissueanadvisorycircular(AC)toaddressthesafety fires (A-01-85),theFAAstated:“Theagreeswithintent participation infirefightingdrillsthatinvolveactualorsimulated Finally, withregardtotherecommendationforimprovedaccess The SafetyBoardrespondedtotheletterfromFAAonJune Regarding therecommendationaskingFAAtorequire ISASI 2003 Proceedings • 177

ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS intended toremedytheseproblems. sibility tofiresources.Therecentsafetyrecommendationswere gency training,unfamiliaritywithsafetyequipment,andinacces- comings increwresponsewereattributedtoinadequateemer- by additionalinflightfireevents.Inmostofthesecases,short- Flight 797inflightfire,andhasbeenstrengthenedinrecentyears Board in2002beganwiththe1983investigationofAirCanada The impetusforthesafetyrecommendationsissuedbySafety Conclusion tion A-01-86wasclassified“OpenAcceptableResponse.” dures orairplanemodificationsidentified,SafetyRecommenda- Technical Centerandtherequiredimplementationofproce- 28, 2002,statingthat,pendingcompletionoftheresearchat extinguishing agents.” using thenewacceptancecriteriainqualificationofthese aircraft. UnderwritersLaboratories’FireSafetySectioniscurrently agent andextinguishertofightfiresininaccessibleareasofthe ers. Thenewcriteriaincludeatesttodeterminetheabilityof will replacehalon1211inhandheldandlavatoryfireextinguish- acceptance criteriaforevaluatingfire-extinguishingagentsthat 178 The SafetyBoardrespondedtotheletterfromFAAonJune • ISA SI 2003 P r oceedings 7 6 5 4 3 2 1 Footnotes to thesesafetyrecommendations. ing. SafetyBoardstaffwillcontinuetomonitortheFAA’sresponses require achangetoCFRPart121.417,hasnotbeenforthcom- erally positive;however,implementationofA-01-85,whichwould See 14CFR121.417providesadefinitionofanactualfire:“An TheUALFlight32incidentoccurredaftertherecommendationswere TheAirTraninitialflightattendanttrainingprogramincludesafirefighting Asaresultofthisincident,onFebruary6,2001,theSafetyBoardissuedto SeeFAAOrder8400.10. NationalTransportationSafetyBoard.1984. sible CargoCompartment magnitude anddurationtoaccomplishthetrainingobjectives.…” an ignitedcombustiblematerial,incontrolledconditions,ofsufficient issued. used inthedrillisavisible,openflame. drill inwhichstudentsarerequiredtoextinguishanactualfire.Thefire less likelytopropagateafire. replacement ofexistinginsulationblanketswithanalternatethatwouldbe garding theinspectionanddesignofstaticportheaterspossible the FAAthreerecommendations(A-01-003,A-01-004,andA-01-005)re- AAR-84/09. Washington,D.C. port, Covington,Kentucky,June2,1983. McDonnell DouglasDC-9-32,C-FTLU,GreaterCincinnatiInternationalAir- The FAA’sresponsestotherecommendationshavebeengen- Effectiveness ofFlightAttendantsAttemptingtoExtinguishFiresinanAcces- , DOT/FAA/AR-TN99/29,April1999. N Aircraft AccidentReport.NTSB/ Air CanadaFlight797, actual fire means ISASI 2003PictorialReview Photos byEsperisonMartinez ISASI 2003 Proceedings • 179

ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS 180 • ISA SI 2003 P r oceedings ISASI 2003 Proceedings • 181

ISASI 2003 PROCEEDINGS ISASI 2003 PROCEEDINGS 182 • ISA SI 2003 P r oceedings