Flight Safety DIGEST MAY 2005

See What’s Sharing Your Airspace Flight Safety Digest Flight Safety Foundation For Everyone Concerned With the Safety of Flight Vol. 24 No. 5 May 2005 www.ﬂ ightsafety.org

OFFICERS AND STAFF Chairman, Board of Governors Amb. Edward W. Stimpson President and CEO Stuart Matthews In This Issue Executive Vice President Robert H. Vandel General Counsel and Secretary Kenneth P. Quinn, Esq. Treasurer David J. Barger See What’s Sharing Your Airspace ADMINISTRATIVE Trans-Pacifi c fl ights by a nearly 26,000-pound gross weight U.S. Manager, Support Services Linda Crowley Horger Air Force Global Hawk unmanned aerial vehicle (UAV) helped drive the current quest for commercial applications. Flying FINANCIAL UAVs in civil airspace demands solutions to problems such as Director of Finance collision avoidance and failure of data/communication links with and Administration Juan G. Gonzalez a ground-based pilot thousands of miles from the aircraft. Accountant Millicent Wheeler 1 MEMBERSHIP Director, Membership 2004 Was ‘Safest Year Ever’ and Development Ann Hill For Air Transport STATS Membership Services Coordinator Ahlam Wahdan Despite an increase in passenger traffi c, there were fewer fatalities Membership Services worldwide in 2004 involving large Western-built commercial jets Coordinator Namratha Apparao than in each of the previous two years. The number of accidents and the rates of accidents declined for air carriers fl ying under U.S. 27 PUBLICATIONS Federal Aviation Regulations Part 121, and there were no fatal Director of Publications Roger Rozelle accidents for scheduled fl ights under Part 135. Senior Editor Mark Lacagnina Senior Editor Wayne Rosenkrans Senior Editor Linda Werfelman ‘Root Causes’ in the System Can Underlie Human Error Associate Editor Rick Darby ARY Web and Print Operational human error in accidents is often only the fi nal Production Coordinator Karen K. Ehrlich LIBR manifestation of ‘latent’ human error in management, design Production Designer Ann L. Mullikin and maintenance. An open organizational culture and user- Production Specialist Susan D. Reed 34 centered design are said to be among the ways to minimize Librarian, Jerry Lederer human error. Aviation Safety Library Patricia Setze TECHNICAL Severe Vibration Accompanies 39 Director of Technical Programs James M. Burin Braking During Landing Rollout Technical Programs Specialist Joanne Anderson After the landing at an airport in Wales, ground personnel Managing Director of found what appeared to be brake parts on the runway.

Internal Evaluation Programs Louis A. Sorrentino III BRIEFS Q-Star Program Administrator Robert Feeler Manager, Data Systems and Analysis Robert Dodd, Ph.D. Manager of Aviation Safety Audits Darol V. Holsman

Founder Jerome Lederer 1902–2004

Flight Safety Foundation is an international membership organization dedicated to the continuous improvement of aviation safety. Nonprofi t and independent, the Foundation was launched offi cially in 1947 in response to the aviation industry’s need for a neutral clearinghouse to Cover photo: The AeroVironment Helios Prototype fl ew to nearly 97,000 feet in August 2001 in fl ight tests by disseminate objective safety information, and for a credible and knowl- the U.S. National Aeronautics and Space Administration (NASA). With a wingspan of 256 feet (78 meters), edgeable body that would identify threats to safety, analyze the problems solar panels, fuel cells and 10 electric motors, it fl ew partly “to develop technologies to enable unmanned and recommend practical solutions to them. Since its beginning, the aerial vehicles (UAVs) to perform a variety of long-duration missions, including environmental monitoring and Foundation has acted in the public interest to produce positive infl uence telecommunications-relay services,” NASA said. This UAV experienced control diffi culties at about 3,000 feet on aviation safety. Today, the Foundation provides leadership to more during its 10th fl ight in a restricted area near the Hawaiian island of Kauai, struck the Pacifi c Ocean and was than 900 member organizations in more than 150 countries. destroyed on June 26, 2003. A similar UAV, Pathfi nder-Plus (see page 22), continues to be fl own for NASA research. (Photo: Carla Thomas, NASA) See What’s Sharing Your Airspace

Trans-Pacific flights by a nearly 26,000-pound gross weight U.S. Air Force Global Hawk unmanned aerial vehicle (UAV) helped drive the current quest for commercial applications. Flying UAVs in civil airspace demands solutions to problems such as collision avoidance and failure of data/communication links with a ground-based pilot thousands of miles from the aircraft.

— FSF EDITORIAL STAFF

echnological advances, military- UAVs has intensifi ed as news media cover military The General Atomics industrial initiatives and fast-track operations in Afghanistan and Iraq. Large UAVs Aeronautical Systems regulatory activities are on the verge routinely are fl own over these countries and con- Altair can fly at 52,000 of launching a new era of routine trolled via satellite data link by a ground-based pilot T feet and remain flight by unmanned aerial vehicles (UAVs) in situated in a ground control station in California, airborne for more than civil airspace. Often defi ned as “aircraft designed U.S.3 The typical remote-control system comprises 30 hours. The aircraft’s to operate with no human pilot aboard,”1 UAVs the UAV, ground control station, control data link also are called uninhabited air vehicles, remotely that operates at line-of-sight distances and/or over- uses include nautical operated aircraft and other terms. For example, the-horizon distances, and data/voice radio relay. charting, fisheries the U.S. Federal Aviation Administration (FAA) assessment and climate and the U.S. Department of Defense (DOD) have “UAVs will range in size from several ounces to research. (Photo: Tom begun to replace “UAV” with “unmanned aircraft thousands of pounds,” said a 2004 MITRE Corp. Tschida, NASA) system” as their preferred term.2 Public interest in report for FAA. “Many will fl y slowly and lack

FLIGHT SAFETY FOUNDATION • FLIGHT SAFETY DIGEST • MAY 2005 1 S EE WHAT’ S SHARING YOUR AIRSPACE

maneuverability, whereas others anything our imagination and entrepreneurs can will operate at very high speeds serve up. Who would have dreamt 20 years ago with great agility. … Further, the that hailing a taxi could mean calling a very light types of missions being planned jet? Or that the product you purchased over the for UAVs of the future are rarely Internet would arrive one day on a UAV? Twenty point-to-point but typically in- years ago, we called people like that dreamers. volve some form of patterned Today, they’re called investors.”6 fl ight or tracking activity that may include intermittent short-[term As of early 2004, more than 40 countries oper- orbits] or long-term orbits.”4 ated more than 80 types of UAVs, said the report of a European UAV task force. The performance The ground-based pilot’s location of current UAVs varies widely in speed, altitude, in a ground control station causes mission duration and payload capability.7 signifi cant limitations compared with pilots of manned aircraft. “To date, approximately 90-plus percent of all funding for UAV systems is a direct result of “Rather than receiving direct sen- national government requirements channeled sory input from the environment in through their military and defense program ele- which his/her vehicle is operating, a ments,” the European task force report said. “The [ground-based pilot] receives only primary mission profi les are quite similar both on that sensory information provided the military [side] and civil side; [they] are mainly by on-board sensors via data link,” earth observation … and communications.” said aerospace researchers at the University of Illinois, U.S. “Currently, this consists primarily of Most large UAVs currently being fl own in U.S. visual imagery covering a restricted fi eld of view. airspace carry radio-relay equipment that enables Sensory cues that are lost therefore include ambi- ground-based pilot–air traffi c control (ATC) voice ent visual information, kinesthetic/vestibular input communication to be conducted much as if the and sound. … [A ground-based pilot] can be said pilot were aboard an aircraft. This simplifi es com- to perform in relative ‘sensory isolation’ from the munication and provides pilots of aircraft near vehicle under his/her control. … To the [ground- a UAV with a method of situational awareness. based pilot] of a UAV with a conventional display, Equipping small UAVs with radio-relay equipment turbulence is indicated solely by perturbations of for ATC communication often is not possible be- the camera image provided by the UAV sensors.”5 cause of payload limitations.8

Integration of military/government UAVs and “UAVs offer a unique range of features, most commercial UAVs into civil airspace may occur notably ultra-long endurance and high-risk sooner than many aviation professionals an- mission acceptance, which cannot be reasonably ticipate, based on the highly touted suitability of performed by manned aircraft,” said the MITRE UAVs for dull, dirty and/or dangerous missions report. “These features — when coupled with (i.e., fl ights that exceed human physical/mental advances in automation and sensor technologies, stamina or subject pilots to environmental hazards and the potential for cost savings — make a strong or signifi cant safety/security threats). case for the eventual emergence of a robust civil, government and commercial UAV market.” Moreover, FAA Administrator Marion Blakey said in March 2005, “[Forecasts of air traffi c] present some enormous challenges and risks. And to a per- Some Divide UAVs son here, we fully recognize that the current system Into Three Basic Categories cannot accommodate such huge new demand. It was never designed to handle the projected new he UAV community and civil aviation authori- mix of air traffi c in our skies. … In 2025, we could Tties envision gradual integration of different be looking at three times more passengers, opera- categories of UAVs into civil airspace. Table 1 (page tions and cargo than we had in 2000 [in the United 3) shows the performance specifi cations of several States]. We need a system that can accommodate Continued on page 5

2 FLIGHT SAFETY FOUNDATION • FLIGHT SAFETY DIGEST • MAY 2005 S EE WHAT’ S SHARING YOUR AIRSPACE

Table 1 PerformancePerformance Specifications Specifications of Unmanned of Unmanned Aerial Aerial Vehicles Vehicles, by Gross by Gross Weight Weight (continued)

UAV Category Manufacturer, First Dimensions and Operating Propulsion, Control and Name Flight and Cost1 Gross Weight Payload2 Parameters2 and Endurance3

RQ-4A Global Northrop Grumman 25,600 pounds Wingspan 116.2 feet 345 knots; Single turbofan engine; Hawk Corp.; 1998; US$20 (11,612 (35.4 meters); length 13,500 nautical autonomous; 36 hours million kilograms) 44.3 feet (13.5 meters); miles; more endurance payload 2,000 pounds than 65,000 feet (907 kilograms)

Proteus Scaled Composites; 12,500 Wingspan 92.0 feet 280 knots at Two turbofan engines; 1998 pounds (5,670 (28.0 meters); length 40,000 feet; autonomous/remotely kilograms) in 56.3 feet (17.2 meters); 65,000 feet at operated/onboard pilot; nonmilitary height 17.6 feet (5.4 7,000 pounds 18 hours endurance uses meters) on landing (3,175.1 gear; 2,000 pounds (907 kilograms); kilograms); payload 58,000 feet at 1,800 pounds to 12,500 pounds; 7,260 pounds (816.5 optionally kilograms to 3,293.1 manned by kilograms) one pilot

X-45A The Boeing Co.; 11,000 pounds length 27 feet (8.2 Mach 0.75 Single turbofan; autonomous 2002 (4,990 kilograms) meters); 3,000 pounds (approximately (1,361 kilograms) 496 knots); 450 nautical miles (833 kilometers) radius of action with 30-minute loiter; more than 35,000 feet

MQ-9 Predator B General Atomics 10,000 pounds Wingspan 64.0 feet (19.5 220 knots; 400 Single turboprop; autonomous; Aeronautical (4,536 kilograms) meters); length 36.2 feet nautical miles more than 24 hours endurance Systems; 2001; (11.0 meters); internal (741 kilometers) $8.7 million payload 750 pounds radius of action; (340 kilograms); external 45,000 feet payload 3,000 pounds (1,361 kilograms)

Altair (variant of General Atomics 7,000 pounds Wingspan 86.0 feet 210 knots at Single turboprop engine MQ-9 Predator B) Aeronautical (3,175 kilograms) (26.2 meters); length 52,000 feet or turbofan engine; triplex Systems; 2003; 36.0 feet (11 meters); redundant ﬂ ight-control $8 million internal payload 660 system; radio link; satellite pounds (300 kilograms); communications link; external payload autonomous; more than 30 3,000 pounds (1,361 hours endurance kilograms)

RQ-6 Fire Scout Northrop Grumman 2,550 pounds Main-rotor diameter 27.5 More than Single turbine engine and main (VTOL) Corp.; 1999 (1,157 kilograms) feet (8.4 meters); length 125 knots; 110 rotor; autonomous; more than 22.9 feet (7.0 meters) nautical miles six hours endurance folded; height 9.4 feet (204 kilometers) (2.9 meters) radius of action; 20,000 feet

MQ-1 Predator General Atomics 2,250 pounds Wingspan 48.7 feet (14.8 70 knots; 400 Single turboprop; autonomous; Aeronautical (1,021 kilograms) meters); length 27 feet nautical miles more than 24 hours endurance Systems; 1994; (8.2 meters); payload 450 (741 kilometers) $2.4 million pounds (204 kilograms) radius of action; 25,000 feet

FLIGHT SAFETY FOUNDATION • FLIGHT SAFETY DIGEST • MAY 2005 3 S EE WHAT’ S SHARING YOUR AIRSPACE

Table 1 Performance Specifications of Unmanned Aerial Vehicles by Gross Weight (continued)

UAV Category Manufacturer, First Dimensions and Operating Propulsion, Control and Name Flight and Cost1 Gross Weight Payload2 Parameters2 and Endurance3

Perseus B Aurora Flight 2,200 pounds Wingspan 71.5 feet 52 knots; Single reciprocating engine Sciences; 1994 (998 kilograms) (21.8 meters); length 25 1,600 nautical and propeller; remotely feet (7.6 meters); height miles (2,963 operated (includes fl ight 12 feet (3.7 meters); kilometers) termination system with payload 264 pounds (120 point-to-point; parachute); 24 hours kilograms) 60 nautical endurance miles (111 kilometers) radius of action; 62,000 feet RQ-5 Hunter Israel Aircraft 1,600 pounds Wingspan 29.2 feet (8.9 100 knots; 144 Two reciprocating gasoline Industries/Malat and (726 kilograms) meters); length 23.0 feet nautical miles engines and propellers; Northrop Grumman (7.0 meters); payload (267 kilometers) autonomous; 11.6 hours Corp.; 1990; $1.2 200.0 pounds (90.7 radius of action; endurance million kilograms) 15,000 feet Pathfi nder-Plus AeroVironment; 1998 700.0 pounds Wingspan 121.0 feet Slow-fl ying Eight solar-powered electric (317.5 kilograms) (36.9 meters); length 11.0 ultralight; motors with propellers; feet (3.4 meters); payload unspecifi ed remotely operated via satellite 700.0 pounds (317.5 range; 82,000 link; unspecifi ed endurance kilograms) feet RQ-2 Pioneer Israel Aircraft 452 pounds Wingspan 17.0 feet (5.2 80 knots; Single reciprocating gasoline Industries/Pioneer (205 kilograms) meters); length 14 feet 100 nautical engine and propeller; UAVs; 1985 (4.3 meters); payload 75 miles (185 autonomous; fi ve hours pounds (34 kilograms) kilometers) endurance radius of action; 15,000 feet RQ-7 Shadow 200 Israel Aircraft 327 pounds Wingspan 12.8 feet (3.9 82 knots; Single reciprocating engine Industries; 1998; (148 kilograms) meters); length 11.2 feet 68 nautical and propeller; autonomous/ $325,000 (3.4 meters); payload miles (126 remotely operated; four hours 60.0 pounds (27.2 kilometers) endurance kilograms) radius of action; 15,000 feet; launched by catapult rail; recovered by arresting equipment RMAX Type IIG Yamaha Motor Co.; 207 pounds (94 Main-rotor diameter 11 knots for Reciprocating water-cooled (VTOL) 2003 kilograms) 10.20 feet (3.12 meters); crop dusting; gasoline engine; autonomous length 11.90 feet (3.63 492 feet (150 fl ight termination/return meters); height 3.54 feet meters) typical to takeoff point using GPS (1.08 meters); payload distance from sensor and gyroscopic 61.7 pounds (28.0 operator; 16.4 sensor; remotely operated kilograms) feet for crop by maneuver command and dusting to 328 hover; one hour endurance feet for aerial imagery Bird Eye 500 Israel Aircraft 77.2 pounds Wingspan 6.6 feet (2.0 60 knots; area Single electric motor and Industries/Malat; (35.0 kilograms) meters); length 5.2 feet of 39 square propeller (hand launch 2004 (1.6 meters); payload 30 miles (10,000 or bungee-cord launch); ounces (850 grams) hectares); 1,000 autonomous with in-fl ight feet waypoint control; more than one hour endurance

4 FLIGHT SAFETY FOUNDATION • FLIGHT SAFETY DIGEST • MAY 2005 S EE WHAT’ S SHARING YOUR AIRSPACE

Table 1 Performance Specifications of Unmanned Aerial Vehicles by Gross Weight (continued)

UAV Category Manufacturer, First Dimensions and Operating Propulsion, Control and Name Flight and Cost1 Gross Weight Payload2 Parameters2 and Endurance3

APV-3 RnR Products; 2003 50.0 pounds Wingspan 12.0 feet (3.7 90 knots Single reciprocating gasoline (22.7 kilograms) meters) engine and propeller; autonomous/remotely operated; preprogrammed navigation and in-fl ight waypoints dynamically confi gurable; eight hours endurance at 45 knots Aerosonde UAV Aerosonde; 1998 33 pounds (15 11 pounds (5 kilograms) 81 knots; Gasoline engine; propeller; kilograms) 23,000 feet autonomous; 32 hours endurance FQM-151 Pointer AeroVironment; 1989 10 pounds (4.5 Wingspan 9.0 feet (2.7 Airspeed not One electric motor, battery and kilograms) meters) specifi ed; 3.0 propeller; remotely operated; nautical miles one hour endurance (5.6 kilometers) radius of action Dragon Eye AeroVironment; 2000; 4.5 pounds (2.0 Wingspan 3.8 feet (1.2 35 knots; 2.5 Specifi cation unavailable $40,000 kilograms) meters); length 2.4 feet nautical miles (launched by bungee cord); (0.7 meter); payload 1.00 (4.6 kilometers) autonomous pound (0.45 kilogram) radius of action UAV = Unmanned aerial vehicle VTOL = Vertical takeoff and landing (rotary wing) GPS = Global positioning system 1. UAV cost, if available, may not include all elements of the complete system, such as sensors and ground control station. 2. Data are not complete for all UAVs. 3. UAVs may be remotely operated (i.e., with fl ight-control inputs/commands by a ground-based pilot using a ground control station) or operated autonomously (i.e., with pre-programmed fl ights conducted by on-board computers and navigation systems); combinations of these fl ight modes also may be used. Data are not complete for all UAVs. Source: U.S. National Aeronautics and Space Administration; U.S. Department of Defense; U.S. Government Accountability Offi ce; manufacturers

UAVs, including those that might be among the • Medium UAVs comprise those capable of ﬁ rst to be integrated into civil airspace. Universally routinely conducting special-purpose flight accepted categories have not emerged, but many operations with explicit flight restrictions studies distinguish UAVs with the following char- (such as the military RQ-2 Pioneer UAV acteristics, whether they have a military application and RQ-7 Shadow 200 UAV); the maximum or government/commercial application: altitudes, airspeeds and payloads are similar to a small manned aircraft. DOD and FAA • Large UAVs are capable of operating at the have proposed to characterize medium UAVs highest altitudes, possibly with the heaviest as restricted in access to civil airspace and re- payloads, longest endurance and/or high- quiring acceptable evidence of airworthiness est airspeeds (such as the military RQ-4A and ground-based pilot qualification other Global Hawk UAV and MQ-1 Predator than a pilot certificate; UAV). For example, DOD and FAA have proposed to characterize large UAVs as • Small/light UAVs — similar to radio-con- capable of beyond-line-of-sight flight trolled model aircraft — comprise those operations throughout all categories of designed for operation at altitudes up to a civil airspace; compliant with U.S. Federal few hundred feet at airspeeds much slower Aviation Regulations (FARs) Part 91 general than small manned aircraft while carrying operating requirements for manned aircraft a payload such as a video camera, environ- (including sense-and-avoid capability); and mental sensors and transmitter (such as the requiring airworthiness certification and military FQM-151 Pointer UAV and Dragon ground-based pilot certification; Eye UAV). DOD and FAA have proposed to

FLIGHT SAFETY FOUNDATION • FLIGHT SAFETY DIGEST • MAY 2005 5 S EE WHAT’ S SHARING YOUR AIRSPACE

characterize small/light UAVs as generally approached the Proteus from various angles. limited to line-of-sight operations, with the The cooperative aircraft maneuvered indi- operator providing acceptable evidence of vidually and in converging groups of two, airworthiness and ground-based pilot/user and included a U.S. National Aeronautics and qualification other than a pilot certificate. Space Administration (NASA) F/A-18 Hornet National regulations typically differentiate jet. In 18 simulated-conflict scenarios, sensors between a small/light UAV and a model air- — combining a radio-based traffic-advisory craft based on criteria such as aircraft weight system and infrared/radar technology — de- and/or its use for commercial purposes.9 tected the collision threats and transmitted traffic advisories to the ground-based pilot, Recent examples of the state of UAV technology who altered the flight path in time for colli- include the following: sion avoidance in all scenarios;11

• One ground-based pilot simultaneously com- • In April 2003, the Proteus was flown by a manded flights by two turbofan-powered ground-based pilot while several “noncoop- Boeing X-45A UAVs in flight formations in erative” aircraft without operating transpon- 2004, and accurately controlled the time of ders — ranging from a glider to the NASA arrival for multiple-UAV flights over specified F/A-18 — approached from various angles geographic locations;10 for 20 simulated-conflict scenarios. “For this series, the [Proteus was] equipped with a 35- • During March 2002 flight tests near Las gigahertz radar system [designed] to detect Cruces, New Mexico, U.S., a Proteus aircraft any approaching aircraft on a potential with a safety pilot aboard was flown by a collision course, regardless of whether the ground-based pilot while several “coopera- intruder [was] equipped with an operating tive” aircraft with operating transponders transponder,” NASA said. The ground-based

The military Boeing X-45A unmanned aerial vehicle is designed to fly at Mach 0.75 (approximately 496 knots) at 35,000 feet. (Illustration: The Boeing Co.)

6 FLIGHT SAFETY FOUNDATION • FLIGHT SAFETY DIGEST • MAY 2005 S EE WHAT’ S SHARING YOUR AIRSPACE

pilot altered the UAV’s flight path in response operations during the country’s response to the The Scaled Composites to data received from the on-board radar December 2004 Indian Ocean tsunami.14 Proteus aircraft, right, system, maintaining a minimum 500-foot and a NASA F/A-18 (152-meter) distance from the intruder air- • Beginning in 2004, U.S. Army RQ-5 Hunter Hornet flew near Las craft in all scenarios. “The detection ranges UAVs were used for reconnaissance flights Cruces, New Mexico, were a little less than we expected but varied by the U.S. Department of Homeland U.S., in 2002 during greatly from about 2.5 [nautical miles] to 6.5 Security along the Arizona border with nautical miles [4.6 kilometers to 12.0 kilo- Mexico 90 miles (145 kilometers) southeast testing of collision- meters], based on the structure and radar of Tucson, Arizona, U.S. The UAV manufac- avoidance systems cross-section of the target aircraft,” NASA turer, Northrop Grumman Corp., said that for unmanned aerial said;12 the capabilities of these UAVs include “sus- vehicles. (Photo: Tom tained autonomous flight, high-resolution Tschida, NASA) • In March 2005, two APV-3 UAVs flown day and nighttime visual and infrared sen- along computer-generated flight paths dem- sors, integrated [global positioning system onstrated the ability, without ground-based (GPS)]-location systems, and the ability pilot intervention, to search in a rectangular to relay communication signals to border- grid pattern above a simulated forest fire patrol agents. … Individuals on the ground while simultaneously conducting synchro- may be unaware of this law-enforcement nized flight maneuvers that avoided obstacles, activity because of the [UAV’s inconspicu- NASA said;13 ous] visual profile at altitude and its quiet engine.”15 • News media reports said that UAVs were used by the defense ministry of India to help • A civilian UAV flight in the Netherlands was locate survivors in distress, direct helicopter conducted by Israel Aircraft Industries/Malat rescue operations and manage recovery in June 2004. The Bird Eye 500 UAV was flown

FLIGHT SAFETY FOUNDATION • FLIGHT SAFETY DIGEST • MAY 2005 7 S EE WHAT’ S SHARING YOUR AIRSPACE

through Class D airspace, then through restricted airspace of a U.S. Navy facility to Flight Level (FL) 210 (approximately 21,000 feet), then entered airspace controlled by FAA air traffic controllers in Honolulu to operate over the plantation for four hours, then returned to the special use airspace and Class D airspace for descent and landing. “FL 210 was considered the optimal altitude from an ATC perspective, since the vast majority of [air carrier] aircraft operating in this same airspace are either at lower or much higher altitudes,” said a report on the project. “Preplanned flight tracks were combined with spontane- ous, controlled maneuvers to guide the UAV to cloud-free areas.”

• The U.S. Air Force in August 2003 became the first U.S. organization to receive a national certificate of authorization (COA) from FAA for operation of the Global Hawk A ground-based in the civil airspace of urban Amsterdam “to in U.S. airspace (called the National Airspace pilot flies the Scaled demonstrate the system’s silent operation, System), enabling FAA approval of flights to Composites Proteus ability to operate in high winds and mini- be granted in as few as five days compared aircraft during collision- mal [equipment/personnel] for operation.” with the normal minimum of 60 days. This avoidance testing by The manufacturer said, “The [ground-based process enabled this UAV to be operated in pilot] conducted rail-track monitoring, NASA. (Photo: nearly all FAA ATC regions by conducting vehicle tracking, waterway monitoring and takeoffs, climbs, descents and landings in Tom Tschida, NASA) other missions using a high-resolution color restricted areas and warning areas while camera and flying autonomous dedicated conducting en route operations above the 16 flight patterns.” highest cruise altitudes used by air carrier aircraft. Some specialists call this flying • During a Canadian Forces Experimentation through a “keyhole to the sky.”19 A U.S. Air Centre test flight of a UAV over the Atlantic Force Global Hawk also completed the first Ocean in July 2003, personnel in the ground trans-Pacific flight by a UAV and the longest control station observed a “dark slick” trailing nonstop UAV flight in April 2001 during a behind a commercial cargo ship. The UAV was deployment from California, U.S., to Australia flown closer to obtain images of the ship, its and completed a return flight to the United name and the apparent pollution. The infor- States two months later;20 and, mation was provided to Transport Canada for further investigation, said National Defence • Small helicopter-type UAVs were used in the 17 of Canada. United States for security-related surveillance of public events such as the 77th Academy Awards • In Hawaii, U.S., NASA conducted two scien- of the U.S. Academy of Motion Picture Arts and tific-research UAV flights to obtain images of Sciences in Los Angeles, California.21 a coffee plantation in September 2002. The purpose of the images was to measure field ripeness, map weeds and detect problems Japanese UAVs Stimulate in fertilization and irrigation. The solar- Commercial Applications powered Pathfinder-Plus UAV flew most of the mission in civil airspace.18 During each he more than 2,000 helicopter-type UAVs used flight, the UAV— equipped with an altitude- Tin growing rice, wheat and soybeans in Japan reporting transponder — initially climbed have set an important commercial example, many

8 FLIGHT SAFETY FOUNDATION • FLIGHT SAFETY DIGEST • MAY 2005 S EE WHAT’ S SHARING YOUR AIRSPACE specialists said. One of the most widely used types multi-sensor station-keeping [maintaining a — the Yamaha Motor Co. RMAX series — has position in the sky]; news media support; aerial been fl own for observation of volcanic eruptions advertising; fi sh spotting; surveying and mapping; with an on-board video camera and video link. commercial imaging; cargo; and commercial se- This type of UAV also has been fl own for plant- curity.”26 Although UAV commercial cargo fl ights growth observations, airborne-radiation detection often are envisioned, predicting when such fl ights and bridge inspections. might begin depends on many variables.

These agricultural UAVs typically are operated One U.S. air cargo company said, “Since its in- at altitudes of 10 feet to 16 feet and an airspeed ception in 1986, UPS Airlines has maintained a of 11 knots. The portable ground control station primary goal: provide quality service to custom- allows the ground-based pilot to fly the UAV ers while operating in a safe and effi cient manner. The Yamaha RMAX within a 200-meter (656-foot) line-of-sight dis- While this includes taking advantage of the latest Type IIG — weighing tance. Most Japanese agricultural UAVs operate technology, we don’t see UAVs as part of our fl eet 207 pounds (94 27 by line-of-sight data link at altitudes of less than in the near future.” kilograms) with a 164 feet. For observation missions, some of these main-rotor diameter of helicopter-type UAVs can operate autonomously Scientifi c-research applications are creating many 10.2-feet (3.1 meters) — i.e., with programmed fl ights conducted by on- new markets, and NASA leases payload space/ — applies insecticides board computers and navigation systems — using time on UAVs as large as the Altair to qualifi ed beyond-the-horizon data links at altitudes up research organizations, and has considered the from 16 feet above to 492 feet. Changes to Japanese civil aviation Global Hawk. In one consultant’s 2004 business crops in Japan while regulations will be required to routinely fl y UAVs model for NASA, the purchase price of an Altair flying at 11 knots. at higher altitudes in civil airspace.22 Under the was estimated to be US$8 million and associated (Photo: Yamaha Motor Corp.) current system, more than 5,500 licensed ground- based pilots are available to operate the agricultural UAVs for application of liquid/granular crop insecticide across 500,000 acres [202,343 hectares] of agricultural land per year.23

In one agricultural test of UAVs in the United States, a NASA APV-3 was used in August 2003 and spring 2005 to obtain digital imagery of a 5,000- acre (2,023-hectare) vineyard near Monterey, California, to map differences in the vigor of grape plants and to test airborne frost-detection sensors. In another multi-year project begun in 2001, a NASA Altus II UAV with infrared sensors has been used to demonstrate methods of wildfi re detection and tactical fi re fi ghting.24

The MITRE report said that near-term uses of nonmilitary government UAVs also may include “emergency response; … search and rescue; forest-fi re monitoring; communications relay; fl ood mapping; high-altitude imaging; nuclear- biological-chemical sensing/tracking; traffi c monitoring; humanitarian aid; land-use mapping; and chemical and petroleum spill monitoring.”25

Expected commercial applications for UAVs include “crop monitoring; … motion picture [production]; communications relay; utility inspection [pipelines/power lines];

FLIGHT SAFETY FOUNDATION • FLIGHT SAFETY DIGEST • MAY 2005 9 S EE WHAT’ S SHARING YOUR AIRSPACE

The RnR Products ground equipment $6 million. The report said that Jr., undersecretary of commerce for oceans and at- APV-3 has an eight-hour U.S. Department of Homeland Security costs for mosphere, and NOAA administrator. endurance at 45 knots Predator demonstration fl ights were $2,358 per and a maximum cruise fl ight hour for 106 fl ight hours in one 17-day series “NASA is glad to see that UAVs are being used and $5,469 per fl ight hour for 128 fl ight hours in airspeed of 90 knots. for more and more diverse and important opera- one fi ve-day series, all in late 2003.28 Worldwide, tions, and we’re looking forward to routine access Shannon Kolensky, a the nonmilitary uses of UAVs primarily have to [U.S. airspace] that will allow UAVs to play an cooperative-program been in public-sector services, scientifi c research expanding role in earth science and other types of student, holds the or aerospace research and development. missions,” said Terrence Hertz, deputy associate aircraft during engine administrator for technology, NASA Aeronautics runup. (Photo: Tom Tschida, As an example of such scientifi c research, NASA Research Mission Directorate. NASA) and the U.S. National Oceanic and Atmospheric Administration (NOAA) in April and May 2005 Other objectives include atmospheric sampling conducted a series of Altair UAV fl ights off the coast and imaging of the Channel Islands National of Southern California near the Channel Islands.29 Marine Sanctuary “to examine shorelines and evaluate the potential for marine enforcement “The goal … is to demonstrate the operational surveillance.” capabilities of [UAVs] for science missions related to oceanic and atmospheric research, climate research, marine-sanctuary mapping Complex Impediments Drive and enforcement, nautical charting, and fi sher- Research on Many Fronts ies assessment and enforcement,” said Thomas J. Cassidy Jr., president and CEO of General Atomics ntegration of UAVs into civil airspace faces a Aeronautical Systems. Inumber of challenges. The European UAV task force, for example, said that operators of UAVs “UAVs will allow us to see weather before it hap- may have diffi culty complying with international pens, detect toxins before we breathe them and dis- rules for avoiding collisions; for detecting visual cover harmful and costly algal blooms before the signals from ATC; for preventing unlawful inter- fi sh do — and there is an urgency to more effectively ference with a UAV, data link or ground-control address these issues,” said Conrad C. Lautenbacher station; and for observation by the ground-based

10 FLIGHT SAFETY FOUNDATION • FLIGHT SAFETY DIGEST • MAY 2005 S EE WHAT’ S SHARING YOUR AIRSPACE pilot of visual signals from the pilot of an inter- the UAV from an air traffic controller — if cepting aircraft. a failure occurs in communication between ATC and the ground-based pilot; Overall, recent studies point to the following requirements for large UAVs and medium UAVs to • A backup system for approach and landing if operate in civil airspace: UAV navigation based on GPS fails;

• On-board sense-and-avoid capability — in- • Mitigation of human factors risks in UAV dependent of the data link to the ground- operations, such as flight handovers between control station — to help prevent in-flight ground-based pilots, procedural errors or collision; fatigue/boredom during flight monitoring; and, • On-board capability to autonomously prevent a collision with terrain or obstacles; • Mitigation of environmental risks to UAVs such as weather, bird strikes and turbulence. Long, narrow wings • Adequate UAV airworthiness and reliability; of the General Atomics Moreover, security of UAVs and ground control Aeronautical Systems • A method for autonomous flight continuation stations is essential because of the possibility that (GA-ASI) Altair enable and termination by the UAV if failure occurs in they could be used for hostile purposes, U.S. gov- the data link to the ground control station; ernment reports said. NASA to conduct 210- knot flights at 52,000 • A method for the UAV to comply with ATC “UAVs represent an inexpensive means of launch- feet. (Photo: NASA – instructions — such as direct commands to ing chemical and biological attacks against the Alan Waide, GA-ASI)

FLIGHT SAFETY FOUNDATION • FLIGHT SAFETY DIGEST • MAY 2005 11 S EE WHAT’ S SHARING YOUR AIRSPACE

A satellite antenna, United States and allied forces and territory,” said a traffi c that may be a confl ict, evaluate fl ight paths, electro-optical/infrared 2004 report by the U.S. General Accounting Offi ce determine traffi c right-of-way, maneuver well clear sensors and ocean- (now Government Accountability Offi ce). “The according to the rules in FARs Part 91.113 [“Right- color sensors fill acting deputy assistant secretary of state for non- of-way Rules; Except Water Operations”] or ma- the payload bay of proliferation testifi ed in June 2002 that UAVs are neuver as required in accordance with Part 91.111 potential delivery systems for [weapons of mass [“Operating Near Other Aircraft].”32 A confl ict is a General Atomics destruction (WMD)], and are ideally suited for the defi ned as another aircraft that will pass less than Aeronautical Systems delivery of chemical and biological weapons given 500 feet [152 meters], horizontally or vertically, Altair. The equipment their ability to disseminate aerosols in appropri- from the UAV.33 was used to measure ate locations at appropriate altitudes. He added Pacific Ocean color, that, although the primary concern has been that Related FARs on aircraft airworthiness and atmospheric conditions nation-states would use UAVs to launch WMD scientifi c research suggest that sense-and-avoid and temperatures attacks, there is potential for terrorist groups to capability should provide the ground-based pilot for Al Gasiewski, a produce or acquire small UAVs and use them for a minimum traffi c-detection fi eld of view of plus chemical or biological weapons delivery.”30 or minus 110 degrees in azimuth measured from scientist with the U.S. the UAV’s longitudinal axis and plus or minus 15 National Oceanic degrees in elevation from the [longitudinal axis and Atmospheric Sense-and-avoid Capability during cruise fl ight]. (This elevation range would Administration. Considered Indispensable be consistent with research on timely detection of (Photo: NOAA) threats from climbing/descending aircraft.)34 ense-and-avoid capability refers to any technol- Sogy or method that compensates for the absence In civil airspace where aircraft already are re- of an on-board pilot who, regardless of the class quired to carry altitude-reporting transponders, of airspace or whether ATC provides separation UAV sense-and-avoid possibilities include TCAS services, would be required to see and avoid other II (which has not been certifi ed for use on UAVs); aircraft whenever weather conditions permit.31 ground-based secondary surveillance radars (such as the traffi c information service [TIS] in DOD defines sense-and-avoid capability as the United States or FAA’s proposed surveillance “the on-board, self-contained ability to detect data network [SDN]); and automatic dependent

12 FLIGHT SAFETY FOUNDATION • FLIGHT SAFETY DIGEST • MAY 2005 S EE WHAT’ S SHARING YOUR AIRSPACE surveillance–broadcast (ADS-B, assuming that “Global Hawk, for example, flies at relatively the aircraft and UAVs carry this equipment in low airspeed and high climb rate, resulting in a the future). Much of the research into sense-and- steeper climb profi le than typically occurs with jet avoid methods for airspace where threat aircraft transports,” said another MIT Lincoln Laboratory may not carry altitude-reporting transponders report. “As a result, encounters with Global Hawk has focused on airborne primary radar, airborne may involve a higher rate of high-vertical-rate passive infrared sensors, airborne passive visual situations than is refl ected in the existing encoun- sensors, ground-based primary radar (i.e., with ter models [i.e., software models that can generate horizontal threat data provided to ground-based millions of air-traffi c-encounter situations based pilots by ATC via landline communication) or on close encounters in actual air traffi c radar data combinations of these methods.35 in specifi c airspace].”37

In theory, routine authorization of UAV opera- Nevertheless, other specialists have voiced con- tions in civil airspace could be accomplished more cerns about the use of TCAS II for UAVs, primar- readily in some airspace than other airspace, said ily because UAVs do not fi t original assumptions a report by Massachusetts (U.S.) Institute of about how TCAS would be used. Technology (MIT) Lincoln Laboratory. MIT researchers currently are studying how TCAS II “First, the current surveillance, display and algo- with resolution advisories could be employed on rithm designs of TCAS were developed and vali- the Global Hawk.36 dated for aircraft with on-board pilots,” one MIT Lincoln Laboratory report said. “[International “A see-and-avoid capability is seldom needed in Civil Aviation Organization (ICAO)] panels … Class A airspace and, because of the high speeds have advised against using existing TCAS on typically found at these altitudes, the effectiveness [UAVs], citing in particular interactions with of see-and-avoid is minimal,” the report said. “The other aircraft carrying [airborne collision avoid- most challenging requirements for [UAV] see-and- ance systems (ACAS)]. Additionally, [a working avoid in Class E airspace occur below 10,000 feet, group of the ICAO panel] has stated that the ICAO where VFR traffic is permitted to fly without mandate requiring TCAS on large piloted aircraft transponders.” does not apply to [UAVs]. … A second concern is that TCAS was never intended to replace see-and- The researchers have considered the possibility that avoid. … Although pilots routinely monitor their different methods of providing the required UAV TCAS traffi c displays when fl ying in high-density safety level may be required in different airspace airspace and depend upon them rather than the same methods for all airspace. to verify the operating integrity of their TCAS equipment, the “[A UAV] that fl ies exclusively in Class A airspace displayed information by itself would encounter only other IFR [instrument fl ight is not adequate to support avoid- rules] traffi c and would receive positive separation ance maneuvers.”38 control by ATC,” the report said. “It is possible that the necessary improvement in safety could be The TCAS II design also assumes achieved merely by equipping such [a UAV] with a specific pilot-response times 25-foot altitude-reporting Mode S transponder, a — five seconds in the ICAO reliable low-latency communication link between TCAS standard — to a resolu- ATC and the [ground-based pilot], and a means tion advisory, but ground-based of handling lost link and emergency (e.g., engine pilots would operate in an en- failure) operations. In this case, safety analysis vironment involving different need not await the defi nition of new surveillance response times. and avoidance algorithms.” “Response times for a [ground- Problems in using current aircraft collision- based pilot] could differ both due avoidance technology for UAVs, however, include to communication latency and to the differences in typical fl ight paths within a alternate control methods such given type of airspace. as the use of a computer mouse

FLIGHT SAFETY FOUNDATION • FLIGHT SAFETY DIGEST • MAY 2005 13 S EE WHAT’ S SHARING YOUR AIRSPACE

U.S. Air Force rather than a control yoke,” said one report. U.K. CAA said, “The CAA policy on UAV sense- crewmembers at “Remote control delays could decrease maneuver and-avoid criteria is as follows: The overriding Beale Air Force Base, effectiveness and retard or negate TCAS–TCAS principle when assessing if a proposed UAV California, U.S., check coordination. Lack of visual cues to check TCAS sense-and-avoid criterion is acceptable is that their squadron’s first integrity or maneuver reasonableness could also it should not introduce a greater hazard than increase collision risk.” (Latency of the voice link currently exists. … From these discussions [with Northrop Grumman between a ground-based pilot and ATC may be one UAV-community representatives] and its own de- RQ-4A Global Hawk, second or more, and may increase the probability liberations, the CAA has concluded that the full which can cruise at 345 of mistimed responses to transmissions, possibly range of parameters which may have to be taken knots at 65,000 feet. requiring a dedicated communication link for into account in any solution of the sense-and- (Photo: John Schwab, UAV ﬂ ights in relatively congested airspace, the avoid problem has yet to be established. … Any 39 U.S. Air Force) report said.) agreed sense-and-avoid criteria must be acceptable to other existing airspace users.”41 ICAO working groups have recommended that for the near future, UAVs carry 25-foot altitude- Similarly, a 2004 U.S. Air Force report said, “At this reporting Mode S transponders and not be point in the development of [a sense-and-avoid] equipped with TCAS II. “Mode S equipage system, we do not have all the information neces- would allow ground controllers [ATC] and sary to establish a defensible and tangible value for TCAS-equipped aircraft to track the [UAV] [equivalent level of safety comparable to manned with precision, and TCAS[-equipped aircraft] aircraft].”42 and other controlled aircraft could maneuver to avoid the [UAV],” a MIT Lincoln Laboratory Regarding UAV navigation, a DOD report said that report said.40 current GPS-based systems generally meet standards for military operations without redundancy and The U.K. Civil Aviation Authority (CAA) in 2004 without reliance on ground-based navigation aids. said that current technology and research have not yet enabled U.K. regulators to specify an acceptable “However, UAVs have a diminished prospect sense-and-avoid capability for UAVs. for relief since, unlike [the pilot of a] manned

14 FLIGHT SAFETY FOUNDATION • FLIGHT SAFETY DIGEST • MAY 2005 S EE WHAT’ S SHARING YOUR AIRSPACE aircraft, a [ground-based pilot] cannot readily fall reestablished in a given time, the UAV can be pre- back on dead reckoning, contact navigation and programmed to retrace its outbound route home, map reading in the same sense that a manned- fl y direct to home or continue its mission. With re- aircraft [pilot can use these backup methods],” spect to lost communications between the ground the report said.43 control stations and the UAV, or the UAV and ATC, however, there is no procedure for a communica- Therefore, the absence of an on-board pilot chang- tions-out recovery. … Remarkably, most lost-link es the nature of the risk of the UAV colliding in situations bear a striking resemblance to [no-radio fl ight with another aircraft, another UAV or birds, IFR operations], and UAVs would enhance their or striking terrain, water, obstacles or people on predictability by autonomously following [this] the ground. Both scenarios have been computer- guidance. … [Nevertheless,] UAVs, even with an simulated for UAVs in various studies. adequate sense-and-avoid system (autonomous) would enhance overall safety by continuing to fl y IFR [after encountering visual meteorological Visual Methods Aim to conditions].”44 Prevent Midair Collisions “Radio-frequency jamming and unintentional aking UAVs visually conspicuous to the pi- interference primarily increase the workload of Mlots of manned aircraft and to other UAVs [ground-based] pilots and air traffi c controllers will require a variety of methods to complement when aircraft carry backup systems,” said the 2002 sense-and-avoid capability. The European UAV DOD UAV roadmap of broad plans. “For systems task force said that, subject to further research, solely dependent on GPS, loss of service leaves this may be a reason to require UAVs to display UAVs to rely on [inertial-navigation] systems, anti-collision lights and navigation lights 24 hours none of which, in today’s [military] UAVs, have a day. drift rates allowing the successful completion of a sortie through to landing,” DOD said. “For [crews of] manned aircraft, avoiding collisions with UAVs may yield additional complica- tions,” the task force said. “If a crew separates by RTCA Accepts Challenge of see-and-avoid, it may misread the distance to a Setting UAV Standards UAV if its size differs signifi cantly from that of a similarly shaped manned aircraft. [This may occur mong major efforts in the United States only] if the crew can distinguish that it is a UAV Afor accelerated integration of UAVs into at all. Such UAVs shall be visually distinguishable civil airspace is RTCA Special Committee 203, from manned aircraft, from any aspect angle. It Minimum Performance Standards may not be possible to achieve this by distinctive for Unmanned Aircraft Systems and color schemes [which] may not be visible from all Unmanned Aircraft, which conduct- angles) or [by] distinctive lighting (the UAV may ed its fi rst meeting in December 2004. be too small to carry additional battery power). Members include representatives of A UAV may need a method of indicating to a government, military services, aca- manned aircraft close by that the UAV is aware demia, manufacturers, an air traffi c of the presence of the other aircraft (and taking controller association, pilot unions appropriate action).” and airlines. The committee initially is scheduled to develop minimum Autonomous flight continuation/termination aviation system performance stan- after failure of a UAV’s data link to the ground dards for UAVs by December 2005; control station could involve several methods, command, control and communi- based on military UAV practices. cation systems for UAVs by June 2006; and sense-and-avoid systems “In the event of lost command and control, mili- for UAVs by December 2007. These tary UAVs are typically programmed to climb to products “will help assure the safe, a predefi ned altitude to attempt to reestablish efficient and compatible opera- contact,” said a DOD report. “If contact is not tion of [UAVs] with other vehicles

FLIGHT SAFETY FOUNDATION • FLIGHT SAFETY DIGEST • MAY 2005 15 S EE WHAT’ S SHARING YOUR AIRSPACE

operating within [U.S. air- status of an aircraft transparent to both ﬂ yers space],” RTCA said. and to the FAA.”47

FAA said that two critical questions for this committee Simulations Check Conformity will be how to handle com- To Target Safety Levels mand and control of UAVs and how sense-and-avoid study of how to achieve acceptable target levels capability will be provided. Aof safety for UAVs — measured as ground fatalities and fatalities from midair collisions — found “With the standards we can that combinations of methods would be required make real progress … on the for operators of some UAVs to bridge the gap safe integration of [UAVs] between their estimated levels of safety and those into [U.S. airspace],” said already achieved by manned aircraft.48 Nick Sabatini, FAA associate administrator for regulation To reduce the risk of ground fatalities caused by a and certifi cation. “We know UAV accident, various methods theoretically could demand is growing. There achieve an acceptable level of safety and lessen the are many applications for severity of a UAV ground impact, such as fl ight- [UAVs] and many who want termination systems and emergency parachute to use them. This urgency recovery systems, the report said. is why we started the rulemaking [process].” An FAA notice of proposed When a UAV’s location was assigned randomly rulemaking (NPRM) during 2005 will suggest in civil airspace with other air traffi c, simulations a regulatory framework for how UAVs initially showed that the level of safety would not meet will be integrated; the notice primarily will ad- the FAA target level of safety for midair collisions dress “[UAV] operation at lower altitudes using (one collision in 10 million hours of operation), line-of-sight traffi c deconfl iction,” he said.45 The the report said. To theoretically achieve this target fi rst FAR for UAVs also is expected to establish level of safety in a specifi ed air traffi c density, vari- procedures for ground-based pilot letters of ous methods could be used, such as procedural authorization and medical qualifi cations, and separation of UAVs from high-density air routes, airworthiness certifi cation of small/light UAVs, use of frangible UAVs (limiting impact damage to FAA said.46 the scale of a bird strike), avoidance maneuvers by the UAV and operating the UAV above FL 450. During the next 10 years, some specialists predict, the same infrastructure required for UAV integra- “It might be possible to signifi cantly reduce the tion into civil airspace — airborne traffi c man- ambient risk of UAV operation by requiring the agement to replace current ground-radar-based UAV to be operated away from airways and major air traffi c separation — will have to be developed fl ight levels,” the report said. “[In the vicinity of because of the traffi c-volume increases of manned jet routes] without mitigating action, there are aircraft. few regions in the vicinity of airways with an expected level of safety below the target level of “Sense-and-avoid [capability] will become an safety. … Signifi cant mitigation would be required integrated, automated part of routine position to operate higher-mass UAVs in [U.S. airspace]. … reporting and navigation functions by rely- Under the assumptions of the analysis performed, ing on a combination of automatic dependent high-altitude [large] UAVs pose a signifi cantly surveillance–broadcast (ADS-B) and [GPS],” greater risk than other [UAV] classes, and tacti- said a joint report on UAV-integration plans by cal [medium] UAVs represent an intermediate DOD and FAA. “In effect, it will create a virtual risk level. While the threat posed by higher-mass bubble of airspace around each aircraft so that UAVs may be greater, they can also incorporate when bubbles contact, avoidance is initiated. All more sophisticated mitigation measures to prevent aircraft will be required to be equipped to the midair collisions, such as sense-and-avoid systems, same level, making the unmanned or manned or active air traffi c control separation.”49

16 FLIGHT SAFETY FOUNDATION • FLIGHT SAFETY DIGEST • MAY 2005 S EE WHAT’ S SHARING YOUR AIRSPACE

Industry First Seeks and off-board (control station) guidance. Mission- High-altitude Access specifi c orbits or other deviations will be coordinated with ATC, and [either] preprogrammed, or nitiatives to integrate large UAVs — military, directed from the control station. Any abnormal Igovernment and commercial — into civil or ‘emergency’ operations [would] follow pre- airspace are in progress worldwide. For example, established (with ATC) procedures, or coordinated Access 5 — a long-term initiative involving high- with ATC in real-time, to divert to [a UAV-capable] altitude, long-endurance (HALE) UAVs in the airport or otherwise execute actions to minimize United States — focuses on civilian UAVs capable the risk to other [U.S. airspace] users, or people/ of operation at 40,000 feet or higher for 24 con- property on the ground. Approach and landing tinuous hours or more.50 Access 5 is a project led [would] be to [a UAV-designated] airport using by NASA to achieve routine operations by HALE appropriate instrument arrival and approach/ remotely operated aircraft (UAVs) within U.S. landing procedures.” airspace as soon as practical. The Access 5 concept of operations said that such a Because a series of ground-based pilots and/ UAV might maintain position above a metropoli- or support personnel can assume UAV flight tan area to provide telecommunications services control/monitoring duties without interrupting similar to a satellite, with takeoff and landing a multi-day UAV fl ight, one advantage of UAVs is conducted from the same airport and a line- that human performance theoretically could be of-sight data link to the ground control station. In consistent from the beginning of the fl ight to the another scenario, the UAV would transport cargo end of the fl ight.51 from a UAV-designated airport on one side of the United States to a UAV-designated airport on the NASA and the U.S. Department of Energy have opposite side. demonstrated the use of sensors aboard HALE UAVs for remote agriculture monitoring, weather Proponents of civilian HALE UAVs said that the research and disaster monitoring (such as fi res, pilot-certifi cation requirements for operating a fl oods, earthquakes/tsunamis and pollution events). manned aircraft above FL 180 should apply to Members of the Access 5 project cited several ex- the ground-based pilots of these UAVs in Class A amples of anticipated commercial fl ights. airspace and any other category of airspace.

“In a nominal mission, the [UAV] would take Access 5 said that the follow- off from [a UAV-]designated airport on an [IFR] ing similarities and differences fl ight clearance and climb to its cruising or mis- — compared with other aircraft sion altitude,” the Access 5 report said. “The [UAV — are anticipated: would] remain within or above controlled airspace for all operations. Throughout the fl ight, the link • “[Ground-based pilots] will between the [UAV] and [ground] control station, comply with airport proce- and the link between the [ground-based pilot] dures and follow standard and [ATC would] be maintained, both while operating procedures to the the [UAV] is within line of sight and over the maximum extent possible horizon from the control station. Also, since the or obtain waivers for any [UAV always would] be under the command of a exceptions. … When [these human [ground-based pilot], the links must be UAVs] are unable to taxi to maintained regardless of the level of autonomy the active runway and have of the UAV. … The [UAV would] avoid adverse to be towed, procedures weather conditions throughout the fl ight. also must be publicized to the other airport users; “The [UAV also would] avoid conflicts with other air traffi c through a combination of fl ight • “[In the en route phase, planning, ATC control and collision-avoidance ground-based pilots] will technology. [UAV] navigation during the fl ight maintain contact with the [would] be through a combination of on-board air route traffic control

FLIGHT SAFETY FOUNDATION • FLIGHT SAFETY DIGEST • MAY 2005 17 S EE WHAT’ S SHARING YOUR AIRSPACE

center [and] will use existing route structure to UAV fl ights in [U.S. airspace],” a DOD report said. the maximum extent possible. If deviations are “Rather, the limitations for military UAV fl ight are required, they will be coordinated with ATC; imposed by the [military] services due to the lack of appropriate equipage of these aircraft.”53 • “[Ground-based pilots] will follow existing IFR departure, arrival and approach procedures [and radar approach control facilities/ Current UAV Operations Set services] to the maximum extent possible. Stage for Airspace Access … If [the ground-based pilots] are unable to navigate along the IFR departure and arrival he requirements of military missions, gov- routes designed for manned aircraft, specific Ternment services, scientific research and routes may need to be developed. … In some commercial ventures intermittently have intro- situations, the time of takeoff and arrival may duced UAVs to civil airspace in some countries. be restricted to specific times when manned- Typically, this temporary access is restricted, aircraft operations are minimal; [and,] keeping UAVs separated much as aircraft are separated from a space vehicle during launch or • “Due to the nature of the [UAV] mission, it landing/recovery. may be critical that the mission portion of the flight [be] given some priority for remain- Typically, regional FAA offi cials approve/deny COA ing on the [flight] profile unless safety [of] applications on a case-by-case basis, and they use another aircraft is a greater concern. In this the notice to airmen (NOTAM) system to designate event, safety considerations must take [pre- temporarily areas of civil airspace where the UAV cedence] over the mission objectives.” will be authorized to fl y or otherwise to separate UAVs and aircraft. The COA sets unique requirements for each fl ight — such as requiring airborne U.S. Military Pursues monitoring of the UAV from a chase aircraft. ‘File-and-fl y’ Simplicity Other methods of UAV–aircraft traffi c separa- n the United States, an intense effort is underway to tion also are used in the United States. In Alaska, Ienable military ground-based pilots of large UAVs NOTAMs are published and fl ight corridors are to fi le IFR fl ight plans with FAA and then operate depicted on aeronautical charts showing low- routinely in civil airspace — to “fi le and fl y” as soon altitude routes used routinely for line-of-sight VFR as 2005 — with access to U.S. airspace equivalent to fl ights between military bases by ground-based their counterparts fl ying manned aircraft. pilots of small military UAVs.

“Military UAVs have historically been fl own in NOTAMs are used by civil aviation authorities restricted airspace (over test and training ranges) worldwide to advise civilian pilots how to avoid or war zones, and have thus largely avoided coming airspace confl icts with UAVs. into confl ict with manned civilian aircraft,” said a 2004 DOD report. “This is changing. … Since the The following example of increasingly common Sept. 11, 2001, terrorist attacks, airspace security UAV NOTAMs was published by Airservices has become an equal priority with safety, and the Australia on April 25, 2005: “A small, unmanned operation of military UAVs for homeland defense aerial vehicle (UAV) will be operating within 1,500 in [U.S. airspace] outside of restricted airspace in- meters [4,921 feet] of the aerodrome reference creasingly is being considered.”52 point. The operator will advise Brisbane Centre on 121.2 10 minutes prior to launch and on comple- Under the current system, U.S. military UAV mis- tion of each sortie. During operations, a listening sions involving fl ights outside restricted areas and watch will be maintained on the common traffi c warning areas are accommodated by FAA on a advisory frequency and Brisbane Centre 121.2 fre- case-by-case basis. quency. Broadcasts will be made on the common air traffi c advisory frequency every 15 minutes. “Statutory language within the [U.S.] Code of The unmanned aerial vehicle is a fixed-wing Federal Regulations does not preclude military aircraft with a 4.1-meter [13.5-foot] wingspan

18 FLIGHT SAFETY FOUNDATION • FLIGHT SAFETY DIGEST • MAY 2005 S EE WHAT’ S SHARING YOUR AIRSPACE and may be either gray or red in color. A ground In general, mishaps involving military UAVs pro- control station will be established northeast of vide the only insights into safety lessons relevant the intersection of Runway 05/23 and Runway to all UAVs. Studies of such data by DOD, FAA 16/34. … The unmanned aerial vehicle control- and the Government Accountability Offi ce have ler [ground-based pilot] will maintain separation included the following fi ndings: with other aircraft from the surface to 1,000 feet above ground level … in visual meteorological • One study calculated that from 1986 through conditions for periods of up to 30 minutes.”54 2001, rates of Class A56 mishaps per 100,000 flight hours for DOD’s predominant UAV systems were 32 for the Predator UAV, 334 Safety Requires Consistent for the Pioneer UAV and 55 for the Hunter Accident/Incident Reporting UAV. These three types of DOD UAVs accumulated a combined total of 100,000 flight viation safety researchers in Europe and the hours between 1986 and 2002 (i.e., mishap AUnited States have found UAV accident/ rates were interpolated because none of incident data and UAV-reliability data to be sparse these types had accumulated 100,000 flight The Hunter II is — except for data supplied by military sources. hours). “In comparison to manned [civil- designed for 30-hour Some of these data are withheld from public use ian] aviation mishap rates, general aviation missions up to 28,000 because they contain classifi ed military informa- aircraft suffer about one Class A mishap per feet and is based on tion. Moreover, DOD reports on UAV accidents, 100,000 hours, regional/commuter airliners Northrop Grumman’s incidents and reliability said that military data- about a tenth of that rate, and larger air- earlier RQ-5 UAV collection methods have been inconsistent com- liners about a hundredth of that rate,” the Hunter aircraft, which pared with methods of investigating, analyzing report said.57 and documenting manned-aircraft mishaps.55 have logged more than Some DOD reports have called for improve- • In 2002, a DOD goal was: “Decrease the an- 32,000 flight hours. ments in safety/reliability data consistency and nual mishap rate of larger-model UAVs to less (Photo: Northrop Grumman data analysis to improve UAV safety. than 20 per 100,000 flight hours by fiscal year Integrated Systems)

FLIGHT SAFETY FOUNDATION • FLIGHT SAFETY DIGEST • MAY 2005 19 S EE WHAT’ S SHARING YOUR AIRSPACE

2009 and [to] less than 15 per 100,000 flight occurred could have been anticipated through hours by fiscal year 2015.”58 an analysis of the user interfaces employed and procedures implemented for their use. For most • Data for the Global Hawk from February of the [UAV] systems, electromechanical failure 1998 through August 2001 showed that was more of a causal factor than human error. … four mishaps occurred with three of the Mishaps attributed at least partially to aircraft fail- UAVs destroyed. DOD said that the cause ures range from 33 percent [U.S. Air Force Global of a March 1999 mishap was an inadvertent Hawk] to 67 percent [U.S. Army Shadow] in the radio transmission on the UAV’s flight- data reported here.” termination frequency while it was airborne; the transmission was attributed to human er- After studying the operation of various military ror. The cause of a December 1999 mishap UAVs, CAMI researchers identiﬁ ed the following was a flight-control software error, which human factors issues: caused the UAV to accelerate off the end of a taxiway, damaging its nose and sensors. The • Control difficulty while ground-based pilots cause of a December 2001 mishap was the operated UAVs by visual contact — especially failure of an incorrectly installed bolt in the during takeoff and landing — which varied ruddervator. The cause of a July 2002 mishap based on whether the UAV was flying toward was a fuel-nozzle problem.59 the pilot or away from the pilot (i.e., spatial disorientation); “The proportions of human error-induced mishaps are nearly • Problems involving the authority of the reversed between UAVs and the ground-based pilot being superseded by the aggregate of manned aircraft, i.e., authority of other personnel involved in the human error is the primary cause flight, or problems in control handover; of roughly 85 percent of manned mishaps, but only 17 percent of • Failure of other personnel to communicate [UAV mishaps],” said a DOD re- non-normal flight conditions and UAV con- port about UAV reliability.”60 ditions to the ground-based pilot;

Among efforts to understand hu- • Absence of visual confirmation of autopilot man factors in military UAV mis- status to the ground-based pilot except for a haps and to apply lessons learned toggle-switch position; to civilian UAV operations, the FAA Civil Aerospace Medical Institute • Absence of visual confirmation to the (CAMI) in 2004 found human fac- ground-based pilot of transmission of an tors issues involved in 21 percent to engine-shutdown command, leading to en- 68 percent of military UAV mishaps gine shutdown on the wrong UAV; based on limited data provided by the U.S. military services.61 The • Errors in conducting checklists, resulting in causal factors other than human factors issues also inadvertent engine shutdown and in-flight were categorized by CAMI researchers as mainte- shutdown of a flight-stability-augmentation nance-related, UAV-related or unknown without system; other details; for some types, failures of a tactical automated landing system (which controls the • Errors in following standard operating pro- UAV during approach and landing, usually without cedures for UAV operation and navigation; ground-based pilot intervention) also were cited. • Weather-related decision-making errors dur- “One critical fi nding … is that each of the fi elded ing UAV flights; [DOD UAV] systems is very different, leading to different kinds of accidents and different human • Inadvertent erasure by the ground-based factors issues,” the CAMI report said. “A second pilot of the contents of random-access mem- fi nding is that many of the accidents that have ory inside systems aboard the UAV; and,

20 FLIGHT SAFETY FOUNDATION • FLIGHT SAFETY DIGEST • MAY 2005 S EE WHAT’ S SHARING YOUR AIRSPACE

• Excessive complexity in an automated mission- that [some UAVs] are not ‘fl own’ in the traditional planning process, reducing situation awareness sense of the word. Only one of the [UAVs] reviewed and ability to interpret status reports. This — Predator — has a [ground-based pilot] interface resulted in one incident, for example, of in- that could be considered similar to a manned air- advertent programming of a UAV taxi speed craft. For the other [UAVs], control of the aircraft of 155 knots without the knowledge of the by the [ground-based pilot] is accomplished indi- ground-based pilot. rectly through the use of menu selections, dedicated knobs or preprogrammed routes. These aircraft are Overall, CAMI researchers found that the inter- not fl own but ‘commanded.’” face between the military ground-based pilot and A Northrop Grumman hardware/software of the ground control station One predictable source of human factors errors RQ-4A Global Hawk often was inconsistent with accepted aviation during UAV fl ights might be ATC. can fly 3,000 nautical principles. miles (5,556 kilometers), “Because so few UAVs have interacted with the air remain on station for “The design of the user interfaces of [military traffi c system to date, it is diffi cult to predict their 24 hours and return UAV] systems are, for the most part, not based on impacts,” said the MITRE report. “Controller roles previously established aviation display concepts,” may also be affected by UAV operations, though, to its departure point the CAMI report said. “Part of the cause for this is in instances where controllers have handled UAVs during one flight. that the developers of these system interfaces are not to date, the procedures and communications were (Photo: Northrop Grumman primarily aircraft manufacturers. Another reason is transparent; most [controllers were] not aware Integrated Systems)

FLIGHT SAFETY FOUNDATION • FLIGHT SAFETY DIGEST • MAY 2005 21 S EE WHAT’ S SHARING YOUR AIRSPACE

The AeroVironment they were controlling a UAV. This has at least surface/terrain-induced (boundary layer) winds, Pathfinder-Plus, been the case with the larger, sophisticated UAVs turbulence, icing, extreme cold and precipitation. which preceded that operate within manned-aircraft performance Small UAVs and those having a light wing load the AeroVironment parameters; however, this may not be the case for are especially sensitive. Even with the larger UAVs, Helios Prototype, other UAVs that are typically slower, cannot per- weather conditions, such as turbulence, have form standard-rate turns at altitude, and may be caused lost [data] links (signal dropout) and even has been flown to unable to climb or descend at rates familiar to loss of control where conditions exceeded the au- 82,000 feet. These controllers.” topilot’s ability to recover. … In most UAV weather UAVs have been used accidents … the [UAVs] were not equipped with to demonstrate the Among ATC changes under consideration in sensors and the [ground-based pilot] was not fully use of solar-powered the United States are methods of distinguishing aware of the hazardous meteorological conditions aircraft as high-altitude UAVs in fl ight plans and radar displays, possibly that existed at the time.” components in a indicating to ATC whether the ground-based pilot telecommunications is manually controlling the UAV or monitoring Overall, the European UAV task force said that autonomous operation by the UAV, and backup civil aviation authorities should “establish a com- system. (Photo: Nick methods of ATC–ground-based pilot communica- mon and harmonized reporting system of occur- Galante, NASA) tion (such as dedicated telephone lines). rences for both UAV operators and [air traffi c] service providers in order to be able to assess and One specialist said that air traffi c controllers also categorize all possible occurrences, determine ap- may require special procedures for UAVs oper- propriate levels of safety and identify key risk areas ating in reduced vertical separation minimums when UAV operations will be integrated outside (RVSM) airspace to prevent upsets caused by wake restricted airspace.” turbulence.62

Moreover, effects of weather should be studied for Work Advances to Update insights into any relevant issues for commercial Civil Aviation Regulations UAVs. ome specialists in the UAV community and in “Many UAVs have conﬁ gurations and charac- Scivil aviation authorities said that they have a teristics that make them more vulnerable to common intention to amend regulations rather weather than most manned aircraft,” the MITRE than to develop new regulations solely for UAVs. report said. “Generally speaking, today’s UAVs are Steps have been taken to establish such regula- lighter, slower and more fragile than their manned tions in several countries. Minimizing UAV-related counterparts and consequently are more uniquely changes in ATC also is a goal of air trafﬁ c services sensitive to certain meteorological events such as providers.

22 FLIGHT SAFETY FOUNDATION • FLIGHT SAFETY DIGEST • MAY 2005 S EE WHAT’ S SHARING YOUR AIRSPACE

“Regulation of UAVs [by FAA] is important be- outside an approved area if they are operated cause it will provide a legal basis for them to oper- clear of populous areas and clear of controlled ate in [U.S. airspace] for the fi rst time,” said the airspace in Australia; approval by CASA is required DOD roadmap for military UAVs. “This, in turn, for fl ight at altitudes more than 400 feet above should lead to their acceptance by international ground level.64 and [non-U.S.] civil aviation authorities.” Requirements for fl ights by large UAVs are the The basic principle for international operations most extensive, requiring fl ight under the con- by UAVs comes from Article 8 of the Convention trol of a certifi ed ground-based pilot (offi cially on International Civil Aviation, which in 1944 called a “UAV controller”) and CASA approval of said, “No aircraft capable of being fl own with- the operation or series of operations along with out a pilot shall be fl own over the territory of a certifi cation of the UAV operator, and compli- contracting state without special authorization ance with rules for maintenance; radiotelephony; by that state and in accordance with the terms ATC procedures, clearances and instructions; and of such an authorization. Each contracting state operational and equipment specifi cations for the undertakes to insure that the fl ight of such an class of airspace. aircraft without a pilot in regions open to civil aircraft shall be so controlled as to obviate danger In 2004, Italy passed legislation to civil aircraft.”63 authorizing the National Civil Aviation Authority, in concert with The Civil Aviation Safety Authority (CASA) in Air Navigation Services Co., to de- Australia said that it implemented in 2002 the velop regulations enabling Italian world’s fi rst comprehensive civil aviation regula- military UAVs to operate routinely tions for UAVs. In April 2005, CASA launched a in restricted areas and corridors of project to refi ne the regulations implemented in civil airspace under civilian air traffi c 2002 to address airworthiness requirements for control, news reports said.65 Australian UAVs, including design, manufacture and continuing airworthiness. Other civil avia- U.K. CAA said, “Currently, an tion authorities also have ordered development equivalent level of [UAV] safety to of required UAV operational concepts, policies, that required for ‘manned fl ying’ standards, regulations and guidance. is achieved by both appropriate regulation and restricting peace- Current Australian regulations exempt the small- time military UAV operations to est UAVs (called “micro UAVs,” with gross weight segregated airspace (i.e., danger of 100.0 grams [3.5 ounces] or less) from most areas). Further, there are currently of the requirements applicable to large UAVs but no national procedures which operators must comply with rules on hazardous permit either civil [UAVs] or military [UAVs] operation, fl ight in controlled airspace and fl ight to routinely fl y in nonsegregated airspace. … exceeding an altitude of 400 feet above ground A signifi cant increase in both civil and military level. Large UAVs in Australia include unmanned [UAV] fl ying is anticipated, most of which will airships with an envelope capacity greater than require access to all classes of airspace if it is to 100 cubic meters (3,531 cubic feet), unmanned be both operationally effective and/or commer- powered parachutes or UAVs with a gross weight cially viable. To achieve this, [UAVs] will have greater than 150 kilograms (331 pounds), and un- to be able to meet all existing safety standards, manned rotorcraft or powered lift devices with applicable to equivalent manned aircraft types, a gross weight greater than 100 kilograms (220 appropriate to the class (or classes) of airspace pounds). The large UAVs generally are treated as within which they are intended to be operated. conventional aircraft in Australian operating rules, … To ensure that air traffi c controllers are aware requiring a special certifi cate of airworthiness (re- that a fl ight is a UAV fl ight, all UAV call signs stricted category) or an experimental certifi cate. shall include the word ‘unmanned.’ … There are currently no regulations governing the qualifi ca- Small UAVs are those that do not fi t into the tions required to operate a civil registered UAV other categories; they can be authorized to fl y in U.K. airspace.”66

FLIGHT SAFETY FOUNDATION • FLIGHT SAFETY DIGEST • MAY 2005 23 S EE WHAT’ S SHARING YOUR AIRSPACE

U.K. CAA guidance includes the following by DOD and the Defense Management Innovations Could general principles for UAV operation out- Contract Agency. UAVs may be operated Benefi t All Civil side of danger areas — which are depicted only by DOD’s civilian contractors out- Airspace Users on CAA charts as permanent/temporary side of restricted areas or warning areas airspace to be avoided by pilots based on when a written agreement with FAA au- ow other airspace users will infl uence current NOTAMs when the area is acti- thorizes the operations. Ground-based Hproposals for integrating UAVs into vated — in U.K. airspace: pilots conducting such operations must civil airspace is not clear, although many have a least a private pilot certifi cate, already participate in the dozens of work- • The ground-based pilot continuously an instrument rating, a current annual ing groups established by the worldwide must monitor UAV performance and instrument review, 300 fl ight hours as pi- UAV community. Other airspace users ATC communications, must comply lot-in-command or mission commander — and society as a whole — are expected with ATC instructions and must be (of UAVs or aircraft, including 100 fl ight to look at the broad context of security re- capable of taking immediate active hours in a manned aircraft), hold current quirements and the indirect benefi ts pro- control of the UAV at all times; ground-based pilot certifi cations (when vided by UAVs to the rest of the aviation FAA begins to require ground-based pilot community. Concerns could be allayed • The UAV must be equipped, as a mini- certifi cation) and comply with guidance not only by addressing safety issues, but mum, with the same types of equip- from the military service about ground- by introducing communication methods, ment required for manned aircraft in based pilot qualifi cation and currency if collision-avoidance systems, worldwide the class of airspace to be used and these requirements are more restrictive data networks, unconventional fuels and an approved method of preventing than the Defense Contract Management autonomous-control capabilities that 68 midair collisions (the method might Agency’s requirements. have collateral applications to manned include “a combination of radar cover- aircraft.70 age and a chase aircraft or an approved Other countries, including the Netherlands on-board system”); and Sweden, have emphasized in their Accuracy, objectivity and data-driven UAV airworthiness-certifi cation processes methods will be required to demonstrate • The UAV must be equipped with an that UAVs should be considered as systems to aviation safety professionals that safety approved method of assuring terrain rather than as isolated aircraft, so that ad- issues are addressed adequately in the clearance; and, equate consideration can be given to the context of UAV applications and busi- ground control station, data link systems, ness opportunities. Flying UAVs in civil • The ground-based pilot must com- systems for autonomous fl ight and colli- airspace of the 21st century recalls risks of ply with the visual/instrument flight sion avoidance, and related software, the aviation in the early 20th century. ■ rules applicable to manned aircraft. MITRE report said. “Thus, UAVs fitted with nonvisual collision-avoidance systems must “Approvals for [UAV] fl ight testing and Notes operation have already been granted for still comply with the IFR when [in in- 1. European Organization for the Safety of strument meteorological conditions some civil projects,” said guidance ma- Air Navigation (Eurocontrol); European (IMC)] (i.e., they may not fly VFR terial prepared by the Swedish Aviation Joint Aviation Authorities (JAA). Joint in IMC just because they can ‘sense’ Safety Authority. “Applications for Initiative on UAVs. A Concept for European and avoid; quadrantal/semi-circular military UAVs to be permitted to fl y in Regulations for Civil Unmanned Aerial rules [for selecting altitudes based on airspace where civil aviation occurs are Vehicles (UAVs). UAV Task-Force Final course] will continue to apply.” expected shortly. … Special (occasional) Report. May 11, 2004. Lt. Col. Trond Bakken, chairman of the Eurocontrol solutions to compensate for faults or un- UAV–Operational Air Traffi c (OAT) Task In Europe, the 2002 European certainty in a UAV system may in some Force, said that the current task force was Commission regulation and implement- cases be necessary. A manned escort plane created in April 2004 for an 18-month ing rules that established the European can be such a solution when the need to initiative to develop voluntary Eurocontrol Aviation Safety Agency (EASA) will see other aircraft is essential and cannot specifi cations for the use of military UAVs require some UAVs to have an EASA be fulfi lled. … Until suffi cient positive as operational air traffi c outside segregated airworthiness certifi cate (i.e., those that experience of a UAV system has been ac- airspace. “In the longer term, the UAV– OAT Task Force also will examine the are neither experimental nor military and quired, special air traffi c controllers could 67 harmonization of UAV operations in seg- weigh more than 150 kilograms). be [provided to separate] UAVs from regated airspace in Europe,” Bakken said. other traffi c within controlled airspace “The task force is now engaged in drafting In the United States, the qualifi cations and to coordinate UAV fl ight paths with the air traffi c management specifi cations for ground-based pilots have been set other air traffi c services units.”69 for initial submission to the Eurocontrol

24 FLIGHT SAFETY FOUNDATION • FLIGHT SAFETY DIGEST • MAY 2005 S EE WHAT’ S SHARING YOUR AIRSPACE

Military Team in 2005. In parallel, consul- feasibility, military utility and operational Roadmap 2002–2027. December 2002. This tations between different internal divisions value of an unmanned air combat system is the second DOD roadmap; a third revi- are taking place discussing regulatory, for the Air Force and the Navy. sion titled OSD Unmanned Aircraft Systems safety, legal, airspace and navigational is- Roadmap was scheduled to be published in 11. U.S. National Aeronautics and Space sues.” May 2005. Administration (NASA). “NASA Flight 2. Pickup, Sharon; Sullivan, Michael J. Tests Validate UAV Collision-avoidance 21. Tactical Aerospace Group. “UAV “Unmanned Aerial Vehicles: Improved Technologies.” News release 02-15. March Helicopter Provides Surveillance Over the Strategic Acquisition Planning Can Help 19, 2002. The tests were conducted under the Oscars.” News release. March 2, 2005. The Address Emerging Challenges.” Testimony Environmental Research Aircraft and Sensor company said that security offi cials used before the Subcommittee on Tactical Air Technology (ERAST) program of NASA. a UAV to provide a roving/stationary air- and Land Forces, Committee on Armed borne video camera with real-time video 12. NASA. “Researchers Encouraged by Services, U.S. House of Representatives. downlink to a wide-area video distribu- Collision-avoidance Test Results.” News re- U.S. Government Accountability Offi ce tion network watched by local, state and lease no. 03-21. April 8, 2003. The tests were (GAO). GAO–05–395T. March 9, 2005. federal commanders. Offi cials on the scene conducted near Mojave, California, U.S. The document said, “DOD defi nes a UAV were able to monitor the video feed from as a powered aerial vehicle that does not 13. NASA. “New Software Allows UAVs to the UAV using portable handheld comput- carry a human operator; can be land-, Team Up for Virtual Experiments.” News ers. air- or ship-launched; uses aerodynamic release no. 05-18AR. March 18, 2005. The 22. Sato, Akira. “Civil UAV Applications forces to provide lift; can be autonomously UAVs were fl own over a remote area of in Japan and Related Safety and or remotely piloted; can be expendable Edwards Air Force Base, California. Certifi cation.” Aeronautic Operations, or recoverable; and can carry a lethal or 14. O’Sullivan, Arieh. “Israeli-made UAVs Yamaha Motor Co. A paper presented to nonlethal payload.” Helping in India.” The Jerusalem Post online the European Unmanned Vehicle Systems 3. Ibid. edition. Jan. 5, 2005. Association’s UAVs: Concerted Actions for Regulations (UCARE) Symposium in 4. DeGarmo, Matthew; Nelson, Gregory 15. Northrop Grumman Corp. “Northrop Paris, France. June 2002. T. Prospective Unmanned Aerial Vehicle Grumman Unmanned Systems Employed Operations in the Future National Airspace by Department of Homeland Security for 23. Offi ce of the Secretary of Defense. System. The MITRE Corp. Published by Border Patrol Missions.” News release. 24. NASA. UAV Applications Center. American Institute of Aeronautics and Nov. 18, 2004. Astronautics (AIAA). September 2004. 16. Israel Aircraft Industries. “IAI/Malat’s 25. DeGarmo. “Late in 2003, an ASTM com- 5. McCarley, Jason S.; Wickens, Christopher Bird Eye 500 Mini UAV Makes Successful mittee was formed to address certifi cation D. “Human Factors Concerns in UAV Historic Maiden Voyage Over Amsterdam, issues concerning UAV integration,” the Flight.” Institute of Aviation, Aviation Netherlands.” News release. June 20, 2004. report said. “Sense-and-avoid systems were Human Factors Division, University of Approval for the fl ight was received from fi rst on the agenda. In August 2004, the Illinois at Urbana-Champaign. Dutch aviation authorities after the UAV’s ASTM F38 committee on UAV standards 6. Blakey, Marion C. Speech to RTCA Annual safety and reliability were checked during developed its fi rst draft standard titled Forum, Washington, D.C., U.S. March 15, demonstration fl ights over the southern Standard Specifi cation for Design and 2005. part of the country. Performance of an Airborne Sense-and- avoid System (designated F2411-04).” 7. Eurocontrol and JAA. 17. Baker, Shanna. “Not Your Ordinary Remote-controlled Airplane.” The Lookout. 26. DeGarmo. 8. DeGarmo, Matthew T. Issues Concerning Canadian Forces Experimentation Centre. Integration of Unmanned Aerial Vehicles July 2003. 27. Spalding, Travis. Public relations supervi- in Civil Airspace. Center for Advanced sor, UPS Airlines. E-mail communication Aviation System Development, 18. Herwitz, Stanley R.; Johnson, Lee F.; with Rosenkrans, Wayne. Alexandria, The MITRE Corp. Report no. MP Dunagan, Stephen E.; Brass, James Virginia, U.S. May 17, 2005. Flight Safety 04W0000323. November 2004. A.; Witt, Glen. “Orchestrating a Near- Foundation, Alexandria, Virginia, U.S. real-time Imaging Mission in National 9. Offi ce of the Secretary of Defense, U.S. 28. Moiré. Cost and Business Model Analysis Airspace Using a Solar-powered UAV.” Department of Defense (DOD). Airspace for Civilian UAV Missions: Final Report. Paper presented at the Second AIAA UAV Integration Plan for Unmanned Aviation Report prepared for the Suborbital Science Conference, Sept. 15–18, 2003. San Diego, — November 2004. November 2004. Offi ce, Earth Science Enterprise, NASA. California, U.S. June 8, 2004. 10. The Boeing Co. “Two Boeing X-45A 19. Baker, Sue. “FAA Approves Global Hawk Unmanned Jets Continue Coordinated 29. General Atomics Aeronautical Systems. Flights in National Airspace.” U.S. Air Flights.” News release Dec. 10, 2004. The “General Atomics Aeronautical Systems Force Materiel Command news release no. Joint Unmanned Combat Air Systems Joins With NOAA and NASA to Help 0839. Aug. 21, 2003. X-45 program is a program of the U.S. Bridge the Gap Between Earth and Defense Advanced Research Projects 20. Offi ce of the Secretary of Defense, DOD. Science; Altair Demonstration Marks Agency, the U.S. Air Force, the U.S. Navy Unmanned Aerial Vehicle Reliability Study. Launch of First NOAA-funded UAV Earth and Boeing to demonstrate the technical February 2003. Unmanned Aerial Vehicles Science Mission.” News release. April 20,

FLIGHT SAFETY FOUNDATION • FLIGHT SAFETY DIGEST • MAY 2005 25 S EE WHAT’ S SHARING YOUR AIRSPACE

2005. “UAV Flight Demonstration Project 40. Drumm et al. 53. Ibid. — Spring 2005.” 41. Directorate of Airspace Policy, U.K. Civil 54. Airservices Australia. Notice to airmen. 30. Christoff, Joseph A. “Nonproliferation: Aviation Authority. Unmanned Aerial Briefi ng reference no. 133448. April 13, Improvements Needed for Controls on Vehicle Operations in UK Airspace — 2005. Exports of Cruise Missile and Unmanned Guidance. Civil Aeronautics Publication 55. Williams, Kevin W. A Summary of Aerial Vehicle Technology.” U.S. General (CAP) 722. Second edition, Nov. 12, 2004. Unmanned Aircraft Accident/Incident Accounting Offi ce. Testimony before 42. Ebdon and Regan. Data: Human Factors Implications. the Subcommittee on National Security, Civil Aerospace Medical Institute, FAA. Emerging Threats and International 43. Offi ce of the Secretary of Defense. Final Report no. DOT/FAA/AM-04/24. Relations, Committee on Government 44. Ibid. December 2004. Reform, U.S. House of Representatives. March 9, 2004. 45. Sabatini, Nick. Prepared remarks for RTCA 56. In DOD terminology, Class A mishaps are Special Committee 203. Nov. 30, 2004. aircraft accidents resulting in loss of the 31. Offi ce of the Secretary of Defense. 46. Swartz, Steve. “Did You See the UAV?” FAA aircraft or loss of human life, or causing 32. Ibid. Aviation News. March–April 2005. more than US$1 million in damage. 33. Ebdon, Derek; Regan, John. Sense-and- 47. Offi ce of the Secretary of Defense. Airspace 57. Offi ce of the Secretary of Defense. avoid Requirement for Remotely Operated Integration Plan for Unmanned Aviation Unmanned Aerial Vehicles Roadmap Aircraft (ROA). U.S. Air Force Air Combat — December 2004. 2002–2027. Command White Paper. June 25, 2004. 48. Weibel, Roland E.; Hansman Jr., R. John. 58. Ibid. 34. Offi ce of the Secretary of Defense. Safety Considerations for Operation of 59. Offi ce of the Secretary of Defense. 35. Drumm, A.C.; Andrews, J.W.; Hall, T.D.; Different Classes of UAVs in the NAS. Sept. Unmanned Aerial Vehicle Reliability Study. Heinz, V.M.; Kuchar, J.K.; Thompson, S.D.; 20, 2004. February 2003. Welch, J.D. “Remotely Piloted Vehicles in 49. Weibel and Hansman. 60. Ibid. Civil Airspace: Requirements and Analysis Methods for the Traffi c Alert and Collision 50. Systems Engineering and Integration 61. Williams. Team, Access 5. In the United States, high- Avoidance System (TCAS) and See-and- 62. DeGarmo. avoid Systems.” Paper presented to the altitude, long-endurance UAVs include 23rd Digital Avionics Systems Conference, AeroVironment’s Pathfi nder-Plus and 63. U.K. CAA. CAP 722. Salt Lake City, Utah, U.S. Oct. 24–28, Helios, General Atomics Aeronautical 64. Civil Aviation Safety Authority, Australia. 2004. The authors are conducting research Systems’ Predator B and Altair, Northrop “Certifi cation Requirements Related to the at the Massachusetts (U.S.) Institute of Grumman’s Global Hawk and Aurora Flight Design, Manufacturing and Airworthiness Technology (MIT) Lincoln Laboratory Sciences’ Perseus B. “Pathfi nder-Plus and of UAVs.” April 11, 2005. Staff writers; under contract to the U.S. Air Force. Helios are solar-powered [UAVs] that have Blackman, Sherridan. “Attack of the very limited maneuvering capability,” said 36. Ibid. Drones.” Flight Safety Australia. December the Access 5 report. “Both climb and turn 2002. 37. Kuchar, James; Andrews, John; Drumm, slowly, and their cruise speed generally does Ann; Hall, Tim; Heinz, Val; Thompson, not exceed [43 knots]. The Perseus B cruises 65. Kington, Tom. “Italy Could Fly Steven; Welch, Jerry. “A Safety Analysis at 65 knots up to an altitude of 60,000 feet. UAVs in Civil Air by Year End.” Process for the Traffi c Alert and Collision The Predator B and Altair HALE [UAVs’] June 14, Avoidance System (TCAS) and See-and- maneuvering capabilities are similar to a 2004. UVOnline. “Italian Parliament avoid Systems on Remotely Piloted Vehicles.” manned aircraft, yet their climb speed and Approves Law on UAV Deployment.” Paper presented to the AIAA Unmanned cruise speed (200-knot range) still are con- July 2004. Unlimited Technical Conference, Chicago, siderably slower than the majority of other DeGarmo and Nelson. Illinois, U.S. Sept. 20–23, 2004. aircraft that operate at their altitude. Global 66. U.K. CAA. CAP 722. Hawk’s climb and cruise speeds are more 38. Drumm et al. Working Group 2 of the 67. Ibid. similar to those of commercial jet aircraft ICAO Secondary Surveillance Radar that operate in the same airspace.” 68. Offi ce of the Secretary of Defense. Airspace Improvements and Collision Avoidance Integration Plan for Unmanned Aviation Systems Panel (SICASP) in 2002 published 51. Systems Engineering and Integration — November 2004. a report about TCAS on UAVs; this work Team, Access 5. HALE ROA [High-alti- currently is being used by MIT Lincoln tude Long-endurance Remotely Operated 69. Wiklund, Eskil. Swedish Aviation Safety Laboratory researchers. Working Group Aircraft] Concept of Operations. Version Authority. Flying With Unmanned Aircraft A of the same panel, renamed ICAO 2.0, March 2005. (UAVs) in Airspace Involving Civil Aviation Surveillance and Confl ict Resolution Activity — Air Safety and the Approvals 52. Offi ce of the Secretary of Defense. Airspace Systems Panel (SCRSP), met in 2004. Procedure. March 25, 2003. Integration Plan for Unmanned Aviation 39. Kuchar et al. — 2004. 70. DeGarmo.

26 FLIGHT SAFETY FOUNDATION • FLIGHT SAFETY DIGEST • MAY 2005 AVIATION STATISTICS

2004 Was ‘Safest Year Ever’ For Air Transport

Despite an increase in passenger traffic, there were fewer fatalities worldwide in 2004 involving large Western-built commercial jets than in each of the previous two years. The number of accidents and the rates of accidents declined for air carriers flying under U.S. Federal Aviation Regulations Part 121, and there were no fatal accidents for scheduled flights under Part 135.

— FSF EDITORIAL STAFF

lthough the number of accidents in- 42 total fatalities. (One CFIT accident was a posi- volving large Western-built commercial tioning fl ight for an on-demand [air taxi] opera- jets operated by International Air tion.) That compared with seven CFIT accidents in A Transport Association (IATA)-member 2003, resulting in 203 total fatalities, and fi ve CFIT airlines and non-IATA-member airlines increased accidents in 2002, with 357 total fatalities.5 from 99 in 2003 to 103 in 2004,1 the number of fatalities in 2004 declined to 428, compared with In 2004, more than 1.8 billion passengers traveled 663 in 2003. In 2002, 974 people were killed during in commercial aircraft worldwide.6 fl ight operations.2 Other preliminary data for 2004, published by the The industrywide hull-loss accident3 rate declined U.S. National Transportation Safety Board (NTSB), 10 percent in 2004, to 0.78 hull losses per million showed that the number of accidents7 and the ac- departures, IATA said (Figure 1, page 28). cident rates for U.S. air carriers operating under U.S. Federal Aviation Regulations (FARs) Part 121, Giovanni Bisignani, IATA director general and Domestic, fl ag and supplemental operations, declined CEO, said, “2004 was the safest year ever for air in 2004 compared with 2003 (Table 1, page 29). The transport.” number of fatal accidents8 and the rates of fatal accidents per 100,000 fl ight hours and fatal accidents Preliminary reports of the number of fatal airline per 100,000 departures were identical to those of accidents worldwide in 2004, including accidents the previous year. involving non-Western aircraft, ranged from 26 to 28.4 There were no fatal accidents in 2004 involving scheduled (commuter) operations under FARs Commercial jets were involved in three controlled- Part 135, Operating requirements: Commuter fl ight-into-terrain (CFIT) accidents in 2004, with and on demand operations and rules governing

FLIGHT SAFETY FOUNDATION • FLIGHT SAFETY DIGEST • MAY 2005 27 S TATISTICS

Figure 1 Western-built Jet Air-transport Traffic and Hull-loss Rates, 1995–2004 1.8 24 Hull Losses per Million Departures Millions of Departures Flown 1.5

1.32 1.34 of Millions 20 1.2

1.11 Departures 0.9 16 0.94 0.78

0.6 Flown 12

Hull Losses per Million Departures 0.3

0.0 8 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 Year Note: A hull loss is defined as an accident in which an aircraft is substantially damaged and is not subsequently repaired for whatever reason, including a financial decision of the owner.

Source: International Air Transport Association

persons on board such aircraft. The fatal- with one fatality. The Aug. 13 accident was There were 11 passenger fatalities and accident rate for Part 135 on-demand a nonscheduled Part 121 cargo fl ight. Both three passenger serious injuries9 in Part operations per 100,000 fl ight hours was accidents remain under investigation.] 121 accidents in 2004 — equivalent to higher than the rate for scheduled Part 56.5 million enplanements per passenger 121 operations. Part 121 scheduled operations in 2004 fatality and 207 million enplanements had a lower accident rate, with 0.124 per passenger serious injury, NTSB said. NTSB said that 28 total accidents and accidents per 100,000 flight hours and Including aircraft crewmembers, there two fatal accidents occurred involving 0.199 accidents per 100,000 departures, were 14 fatalities. Part 121 air carriers, compared with 54 than scheduled operations under Part total accidents and two fatal accidents 135, which had 1.515 accidents per For Part 121 air carriers, the number in 2003. 100,000 flight hours and 0.843 acci- and rate of major accidents10 increased dents per 100,000 departures, NTSB said in 2004 compared with the previous two [The two fatal Part 121 accidents in (Table 2, page 30). The fatal-accident years, but the numbers and rates of acci- 2004 included a British Aerospace (BAe) rate was 0.006 accidents per 100,000 dents classifi ed as serious accidents,11 in- Jetstream 32 that struck terrain during flight hours and 0.009 accidents per jury accidents12 and damage accidents13 an instrument approach to Kirksville 100,000 departures for Part 121 sched- declined (Table 3, page 30). (Missouri, U.S.) Regional Airport on uled operations. Part 135 on-demand Oct. 19, with 13 fatalities, and a Convair operations had a fatal- accident rate of Part 121 air carriers experienced three 580 that struck terrain during a visual ap- 0.780 per 100,000 flight hours; the fatal- hull losses in 2004, for a rate of 0.171 per proach to Cincinnati/Northern Kentucky accident rate based on departures was million fl ight hours (Table 4, page 31). (U.S.) International Airport on Aug. 13, not available. Continued on page 31

28 FLIGHT SAFETY FOUNDATION • FLIGHT SAFETY DIGEST • MAY 2005 S TATISTICS Accidents Accidents Departures per 100,000 Accidents Accidents per 1,000,000 Miles Flown . , such as suicide, sabotage and terrorism, are included in the totals for for included in the totals are sabotage and terrorism, such as suicide, , s resulting from the Sept. 11, 2001, terrorist act are excluded from this from act excluded terrorist are 2001, 11, the Sept. from s resulting Accidents Accidents per 100,000 Flight Hours per 100,000 Flight Table 1 Table Scheduled and Nonscheduled Service (Airlines), 1985–2004 Scheduled and Nonscheduled Service (Airlines), Flight Hours Flight Miles Flown Departures Accidents, Fatalities and Accident Rates, U.S. Air Carriers Operating Under FARs Part 121, Part 121, Under FARs Operating Carriers Air U.S. Rates, and Accident Fatalities Accidents, Fatalities Accidents All Fatal Total Aboard Aboard Total Fatal All Fatal All Fatal All Fatal All Year Year 1985 21 7 526 525 8,709,894 3,631,017,000 6,306,759 0.241 0.080 0.0058 0.0019 0.333 0.241 0.080 0.0058 0.111 6,306,759 3,631,017,000 0.0009 0.434 0.310 0.038 0.0076 0.053 7,601,373 0.0005 0.319 0.231 0.020 0.0057 0.028 8,709,894 7,202,027 0.0004 0.376 0.260 525 0.018 0.0064 0.026 4,360,521,000 7,716,061 4,017,626,000 526 0.0024 0.366 0.248 0.098 0.0061 0.144 10,645,192 4,503,426,000 7,645,494 7 9,976,104 230 7 8 4,605,083,000 11,140,548 21 3 232 274 0.0012 0.297 0.198 24 0.049 0.0049 0.074 1985 11,274,543 8,092,306 5 285 276 0.0008 0.333 0.221 34 0.034 0.0054 0.051 4,947,832,000 1986* 7,814,875 3 278 0.0008 0.228 0.146 30 0.032 0.0036 0.051 12,150,116 4,824,824,000 1987* 11 7,880,707 12 39 0.0007 0.267 0.168 11,780,610 5,039,435,000 1988* 0.030 0.0040 0.049 28 0.0002 0.285 0.181 8,238,306 6 0.008 0.0044 0.012 49 8,073,173 62 12,359,715 1989 5,478,118,000 24 0.0005 0.426 0.267 0.022 0.0064 0.035 4 5,249,469,000 31 8,457,465 33 1990 13,124,315 26 0.0009 0.450 0.269 0.036 0.0063 0.061 5,654,069,000 4 12,706,206 8,228,810 237 1991 18 0 1 13,505,257 5,873,108,000 239 1 162 1992 0.0006 0.475 0.309 23 0.025 0.0073 0.039 4 13,746,112 10,318,383 168 23 350 1993 0.0001 0.455 0.297 0.006 0.0074 0.009 0.0003 0.451 0.291 3 10,979,762 6,696,638,000 380 0.011 0.0072 0.018 11,308,762 1994* 36 0.0004 0.488 0.306 5 15,838,109 6,736,543,000 0.0003 0.383 0.236 0.016 0.0074 0.026 0.011 0.0058 0.018 11,468,229 7,101,314,000 1995 10,954,832 37 6 8 4 16,816,555 17,555,208 7,524,027,000 1996 7,294,191,000 49 0 1 1 11 12 18,299,257 1997 17,814,191 50 0.0003 0.518 0.310 2 92 0.011 0.0074 0.019 525 92 10,422,862 1998 51 531 0.0003 0.260 0.159 3 0.011 0.0038 0.019 10,785,000 7,280,383,000 1999 56 6 46 17,433,964 7,378,300,000 2000 21 22 17,575,000 2001* 2 14 14 2002 41 54 0 2 0 2003 28 10,508,473 0 7,192,501,000 17,290,198 0.237 2004 — 0.0057 — 0.390 — FARs = U.S. Federal Aviation Regulations Aviation Federal = U.S. FARs preliminary. 2004 data are Note: Administration. Aviation Federal the U.S. miles and departures by compiled are hours, Flight 121 Part under FARs been operated seats used in scheduled passenger service with 10 or more aircraft have 1997, March 20, Since actsThese in this category. an occurrence those in which an illegal act the symbol * are for was responsible by followed Years fatalitie killed, Other who were aircraft than the people aboard accident rates. from excluded accidents and fatalities but are table. Board Safety Transportation National U.S. Source:

FLIGHT SAFETY FOUNDATION • FLIGHT SAFETY DIGEST • MAY 2005 29 S TATISTICS

Table 2 Accidents, Fatalities and Accident Rates, U.S. Aviation, 2004 Accidents per Accidents per Accidents Fatalities 100,000 Flight Hours 100,000 Departures All Fatal Total AboardFlight Hours Departures All Fatal All Fatal U.S. air carriers operating under FARs Part 121 Scheduled 21 1 13 13 17,000,000 10,547,000 0.124 0.006 0.199 0.009 Nonscheduled 7 1 1 1 575,000 238,000 1.217 0.174 2.941 0.420 U.S. air carriers operating under FARs Part 135 Scheduled 5 — — — 330,000 593,000 1.515 — 0.843 — Nonscheduled 68 24 65 64 3,072,000 — 2.210 0.780 — — FARs = U.S. Federal Aviation Regulations Note: All data are preliminary. Flight hours and departures are compiled and estimated by the U.S. Federal Aviation Administration. Departure information for nonscheduled FARs Part 135 operations is not available. Source: U.S. National Transportation Safety Board

Table 3 Accidents and Accident Rates by NTSB Classification, U.S. Air Carriers Operating Under FARs Part 121, 1985–2004

Accidents Accidents per Million Flight Hours Flight Hours Year Major Serious Injury Damage (millions) Major Serious Injury Damage 1985 8 2 5 6 8.710 0.918 0.230 0.574 0.689 1986 4 0 14 6 9.976 0.401 0.000 1.403 0.601 1987 5 1 12 16 10.645 0.470 0.094 1.127 1.503 1988 4 2 13 11 11.141 0.359 0.180 1.167 0.987 1989 8 4 6 10 11.275 0.710 0.355 0.532 0.887 1990 4 3 10 7 12.150 0.329 0.247 0.823 0.576 1991 5 2 10 9 11.781 0.424 0.170 0.849 0.764 1992 3 3 10 2 12.360 0.243 0.243 0.809 0.162 1993 1 2 12 8 12.706 0.079 0.157 0.944 0.630 1994 4 0 12 7 13.124 0.305 0.000 0.914 0.533 1995 3 2 14 17 13.505 0.222 0.148 1.037 1.259 1996 6 0 18 13 13.746 0.436 0.000 1.309 0.946 1997 2 4 24 19 15.838 0.126 0.253 1.515 1.200 1998 0 3 21 26 16.817 0.000 0.178 1.249 1.546 1999 2 2 20 27 17.555 0.114 0.114 1.139 1.538 2000 3 3 20 30 18.299 0.109 0.109 1.093 1.475 2001 5 1 19 21 17.814 0.281 0.056 1.067 1.179 2002 1 1 14 25 17.290 0.058 0.058 0.810 1.446 2003 2 3 24 25 17.434 0.115 0.172 1.377 1.434 2004 3 0 13 12 17.575 0.171 0.000 0.740 0.683

FARs = U.S. Federal Aviation Regulations NTSB = U.S. National Transportation Safety Board Note: Since March 20, 1997, aircraft with 10 or more seats used in scheduled passenger service have been operated under FARs Part 121. A major accident is defi ned as one in which a Part 121 aircraft was destroyed, or there were multiple fatalities or there was one fatality and a Part 121 aircraft was substantially damaged. A serious accident is defi ned as one in which there was one fatality without substantial damage to a Part 121 aircraft or there was at least one serious injury and a Part 121 aircraft was substantially damaged. An injury accident is defi ned as one with at least one serious injury and without substantial damage to a Part 121 aircraft. A damage accident is defi ned as one in which no person was killed or seriously injured but in which any aircraft was substantially damaged.

Source: U.S. National Transportation Safety Board

30 FLIGHT SAFETY FOUNDATION • FLIGHT SAFETY DIGEST • MAY 2005 S TATISTICS

Both the number and rate were the highest of any year since 2001. Table 4 Number and Rate of Destroyed Aircraft, U.S. Air Carriers Accidents involving commuter flights under Operating Under FARs Part 121, 1985–2004 Part 135 increased to five in 2004, compared Flight Hours Hull Losses per with two the previous year (Table 5, page 32). Year Hull Losses (millions) Million Flight Hours The rates of 1.515 accidents per 100,000 flight 1985 8 8.710 0.918 hours and 0.843 accidents per 100,000 departures also represented increases compared 1986 2 9.976 0.200 with 2003. There were no fatal accidents in the 1987 5 10.645 0.470 category. 1988 3 11.141 0.269 1989 7 11.275 0.621 In 2004, on-demand operations under Part 135 resulted in 68 accidents, compared with 75 in 1990 3 12.150 0.247 2003. The 24 fatal accidents in 2004 compared 1991 5 11.781 0.424 with 18 in 2003 (Table 6, page 33). The 2004 1992 3 12.360 0.243 accident rate per 100,000 ﬂ ight hours in Part 1993 1 12.706 0.079 135 on-demand operations declined from the 2003 rate, and the fatal-accident rate increased 1994 3 13.124 0.229 from the 2003 rate. 1995 3 13.505 0.222 1996 5 13.746 0.364 The U.S. data are available on the Internet at 1997 2 15.838 0.126 . ■ 1998 0 16.817 0.000 1999 2 17.555 0.114 Notes 2000 3 18.299 0.164

1. International Air Transport Association (IATA) news 2001 5 17.814 0.281 release. March 7, 2005. . 2002 1 17.290 0.058

2. International Air Transport Association (IATA) data 2003 2 17.434 0.115 for 2002 and 2003 were cited by Airwise News, April 2004 3 17.575 0.171 27, 2005. FARs = U.S. Federal Aviation Regulations 3. A hull loss is deﬁ ned by IATA as “an accident in Note: Since March 20, 1997, aircraft with 10 or more seats used in scheduled passenger which an aircraft is substantially damaged and is not service have been operated under FARs Part 121. subsequently repaired for whatever reason, including Source: U.S. National Transportation Safety Board a ﬁ nancial decision of the owner.”

4. Data were included in Flight International Volume the aircraft with the intention of fl ight and all 67 (Jan. 25–31, 2005); Airclaims, cited in Air such persons have disembarked, and in which any Safety Week, Jan. 10, 2005; and Aviation Safety person suffers death or serious injury, or in which Network, Jan. 1, 2005 . 8. NTSB defi nes a fatal accident as one that results in 5. Bateman, C. Don, chief engineer for avionics death within 30 days of the accident. safety systems, Honeywell. E-mail communications with Darby, Rick. Alexandria, Virginia, U.S., 9. NTSB defines a serious injury as one that requires May 9, 2005, and May 10, 2005. Flight Safety hospitalization for more than 48 hours, commenc- Foundation, Alexandria, Virginia, U.S. ing within seven days from the date the injury was received; results in a fracture of any bone (except 6. IATA, op. cit. simple fractures of fingers, toes or nose); causes severe hemorrhages, nerve, muscle or tendon 7. The U.S. National Transportation Safety Board damage; involves any internal organ; or involves (NTSB) defi nes an accident as “an occurrence as- second-degree burns or third-degree burns, or any sociated with the operation of an aircraft which burns affecting more than 5 percent of the body takes place between the time any person boards surface.

FLIGHT SAFETY FOUNDATION • FLIGHT SAFETY DIGEST • MAY 2005 31 S TATISTICS Accidents per Accidents 100,000 Departures ively to 36 accidents, nine of them fatal, nine of them fatal, 36 accidents, to ively Accidents Accidents Miles Flown Miles Flown per 1,000,000 . rt 135 operation conducted with no revenue passengers aboard is rt passengers aboard conducted with no revenue 135 operation Accidents Accidents per 100,000 Flight Hours Flight , such as suicide, sabotage and terrorism, are included in the totals for for included in the totals are sabotage and terrorism, such as suicide, , Table 5 Table 353,670369,535 50,773,000273,559 44,943,000 707,071319,206330,000 41,633,000 2.262 603,659 47,404,000 51,100,000 — 3.247 513,452 580,805 0.271 0.1576 2.559 593,000 0.627 0.2670 — 1.515 — 0.0223 0.313 1.131 0.1681 — 1.988 0.0422 — — 0.0211 0.0978 0.166 0.344 1.363 — 0.172 — 0.843 — 1,724,586 307,393,000 2,798,8112,341,760 0.812 450,133,000 0.116 3,160,089 0.0455 0.641 0.0065 0.128 0.500 0.0333 0.071 0.0067 0.475 0.095 Flight Hours Flight Miles Flown Departures Fatalities ight operation. This interpretation has been applied to accidents beginning in the year 2002. It has not been applied retroact 2002. accidents beginning in the year has been applied to interpretation This ight operation. Accidents, Fatalities and Accident Rates, U.S. Air Carriers Operating Under FARs Part 135, Scheduled Service, 1985–2004 Scheduled Service, Part 135, Under FARs Operating Carriers Air U.S. Rates, and Accident Fatalities Accidents, Accidents 2155 7000 2122 5000 4244 5364 8000 All All Fatal Total Aboard All Fatal All Fatal All Fatal FARs = U.S. Federal Aviation Regulations Aviation Federal = U.S. FARs preliminary. 2004 data are Note: (FAA). Administration Aviation Federal the U.S. miles and departures by compiled are hours, Flight 121 Part under FARs been operated seats used in scheduled passenger service with 10 or more aircraft have 1997, March 20, Since Board Safety Transportation National U.S. Source: Years followed by the symbol * are those in which an illegal act was responsible for an occurrence in this category. These actsThese in this category. an occurrence those in which an illegal act the symbol * are for was responsible by followed Years accident rates. from excluded accidents and fatalities but are Pa FARs any (NTSB), Board Safety Transportation National the U.S. to provided legal interpretation 2002 FAA Based on a February a nonscheduled ﬂ considered that occurred during the period 1983–2001. that occurred 199920002001 132002 1 2003 7 52004 2 12 13 12 13 342,731 52,403,000 300,432 672,278 43,099,000 3.793 558,052 1.459 2.330 0.2481 0.666 0.0954 0.1624 1.934 0.0464 0.744 1.254 0.358 1992*1993 2319941995 161996 10 71997 12 41998 11 3 21 16 2 24 1 25 21 5 9 23 14 25 2,335,349 46 2,638,347 9 507,985,000 12 2,784,129 46 554,549,000 3,114,932 2,627,866 594,134,000 2,756,755 3,601,902 0.942 550,377,000 982,764 3,581,189 590,727,000 0.606 0.300 3,220,262 246,029,000 0.359 3,515,040 0.152 0.0433 1,394,096 0.457 0.108 0.399 0.0289 0.0138 0.0168 0.076 1.628 0.0072 0.036 0.706 0.0050 0.0218 0.509 0.444 0.0186 0.225 0.279 0.0036 0.0017 0.0650 0.111 0.373 0.084 0.0203 0.313 0.062 1.148 0.028 0.359 19851986 18 1 7 37 36 1,737,106 300,817,000 2,561,463 1.036 0.403 0.0598 0.0233 0.703 0.273 198719881989 331990 181991 19 10 1 2 23 5 59 21 8 31 57 21 99 1,946,349 31 2,092,689 350,879,000 77 2,240,555 380,237,000 2,809,918 393,619,000 2,291,581 2,909,005 1.695 2,818,520 433,900,000 0.860 0.514 0.848 2,820,440 0.096 0.0940 0.223 1.004 0.0473 0.0285 0.0483 0.0053 0.349 1.174 0.0127 0.619 0.0530 0.356 0.674 0.0184 0.069 0.177 0.815 0.284 Year Year

32 FLIGHT SAFETY FOUNDATION • FLIGHT SAFETY DIGEST • MAY 2005 S TATISTICS

Table 6 Accidents, Fatalities and Accident Rates, U.S. Air Carriers Operating Under FARs Part 135, Nonscheduled Service, 1985–2004

Accidents per 100,000 Accidents Fatalities Flight Hours Year All Fatal Total Aboard Flight Hours All Fatal 1985 157 35 76 75 2,570,000 6.11 1.36 1986 118 31 65 61 2,690,000 4.39 1.15 1987 96 30 65 63 2,657,000 3.61 1.13 1988 102 28 59 55 2,632,000 3.88 1.06 1989 110 25 83 81 3,020,000 3.64 0.83 1990 107 29 51 49 2,249,000 4.76 1.29 1991 88 28 78 74 2,241,000 3.93 1.25 1992 76 24 68 65 2,844,000 2.67 0.84 1993 69 19 42 42 2,324,000 2.97 0.82 1994 85 26 63 62 2,465,000 3.45 1.05 1995 75 24 52 52 2,486,000 3.02 0.97 1996 90 29 63 63 3,220,000 2.80 0.90 1997 82 15 39 39 3,098,000 2.65 0.48 1998 77 17 45 41 3,802,000 2.03 0.45 1999 74 12 38 38 3,204,000 2.31 0.37 2000 80 22 71 68 3,930,000 2.04 0.56 2001 72 18 60 59 2,997,000 2.40 0.60 2002 60 18 35 35 2,911,000 2.06 0.62 2003 75 18 42 40 2,927,000 2.56 0.61 2004 68 24 65 64 3,072,000 2.21 0.78

FARs = U.S. Federal Aviation Regulations Note: 2004 data are preliminary. Flight hours are estimated by the U.S. Federal Aviation Administration (FAA). Miles ﬂ own and departure information for nonscheduled Part 135 operations are not available. In 2002, FAA changed its estimate of nonscheduled (on-demand) activity. The revision was applied retroactively, beginning with the year 1992. In 2003, FAA again revised ﬂ ight-activity estimates for 1999 to 2002.

Source: U.S. National Transportation Safety Board

10. NTSB classiﬁ es as a major accident one in which a 12. NTSB classifies as an injury accident a nonfatal ac- U.S. Federal Aviation Regulations (FARs) Part 121 cident with at least one serious injury and without aircraft was destroyed, or there were multiple fatali- substantial damage to a Part 121 aircraft. ties or there was one fatality and a Part 121 aircraft 13. NTSB classifies as a damage accident an accident was substantially damaged. in which no person was killed or seriously injured, 11. NTSB classiﬁ es as a serious accident one in which there but in which any aircraft was substantially dam- was one fatality without substantial damage to a Part aged. organ; or involves second-degree burns or 121 aircraft or there was at least one serious injury third-degree burns, or any burns affecting more and a Part 121 aircraft was substantially damaged. than 5 percent of the body surface.

FLIGHT SAFETY FOUNDATION • FLIGHT SAFETY DIGEST • MAY 2005 33 PUBLICATIONS RECEIVED AT FSF JERRY LEDERER AVIATION SAFETY LIBRARY

‘Root Causes’ in the System Can Underlie Human Error

Operational human error in accidents is often only the final manifestation of ‘latent’ human error in management, design and maintenance. An open organizational culture and user-centered design are said to be among the ways to minimize human error.

— FSF LIBRARY STAFF

Books “The direct cause [of the accidents discussed] is usually a human error, while the root cause tends to The Blame Machine: Why Human Error Causes be the main underlying system fault that made the Accidents. Whittingham, R.B. Oxford, England: error possible or more likely,” says the author. “The Elsevier Butterworth-Heinemann, 2004. 288 pp. contributory causes are less signifi cant factors or Figures, tables, appendix, references, index. systems that also infl uenced the probability of error. The system of interest may be an item of equipment, n the immediate aftermath of a serious ac- a procedure or an organizational system.” “Icident, there is a natural tendency, especially by the media, quickly to suggest a cause and maybe Although some instances of human error are attribute blame,” says the author. “Quite often the “active” — occurring in operations — many are words ‘it is believed the accident was a result of “latent,” the author says. Types of latent errors that human error’ are heard. There is an air of fi nality he discusses in connection with pertinent acci- about this statement, it being implicit that no more dents include organizational and management can be said or done. The inevitable has happened. errors, design errors and maintenance errors. [It is hoped that] the accident investigation which is held later does not accept human error as inevi- Among the factors that can minimize human error table, but goes on to reveal the underlying reasons that is encouraged by systemic failure, the author why the error occurred.” says, are the following:

The book focuses on system faults that can make • An open organizational culture that ac- errors more likely, the author says. Part I classiﬁ es knowledges mistakes and learns from them, types of errors and suggests ways of estimating emphasizing underlying causes more than their frequency, so that their frequency can be individual actions; and, minimized. Part II comprises studies of serious accidents in various industries, including aviation, • User-centered design, in which “the design of to show how systemic errors were implicated. the system is matched as closely as possible

34 FLIGHT SAFETY FOUNDATION • FLIGHT SAFETY DIGEST • MAY 2005 R ESOURCES

to human capabilities and limitations.” This of their fl ying experiments. … Between 1900 and is contrasted with system-centered design, 1906, they made hundreds of pictures of gliders in which the system rather than the human and airplanes in fl ight, including their famous im- operating it is the focus of attention. The age, taken Dec. 17, 1903, of the fi rst moment that author says, “There are numerous examples man left the ground in a heavier-than-air, powered of complex technological systems that have fl ying machine.” been designed mainly with system func- tionality in mind, ignoring the capabilities John Daniels, one of the crewmembers from a and limitations of the user. Such systems nearby life-saving station who had volunteered as invariably result in degraded levels of hu- assistants and witnesses, took the Dec. 17 photo- man performance, with grave consequences graph. “Daniels forgot about the camera, but when for productivity, equipment availability and Orville inspected it [after the fl ight], the shutter safety.” had been released,” the authors say. “Daniels had automatically squeezed his hand in the excitement. The Wright Brothers Legacy: Orville and It was the only photograph he ever took. And it was Wilbur Wright and Their Aeroplanes. Burton, not until days later, when the brothers went into Walt; Findsen, Owen. New York, New York, U.S.: their darkroom in the shed behind their Dayton Harry N. Abrams, 2003. 224 pp. Photographs. home, that they held the glass plate up to the safe light and saw that they had the picture, one of the n important aspect of the Wright Brothers’ most famous photographs ever taken.” “A quest to fl y was their use of photography, because it provided irrefutable evidence that on The photograph, which shows the Flyer, as the that cold and barren stretch of beach [at Kitty airplane was called, after it had risen from the Hawk, North Carolina, U.S.] they had succeeded,” launching track, with Orville stretched out at the says Alexander Lee Nyerges, director and CEO controls and Wilbur running alongside, is repro- of the Dayton [Ohio, U.S.] Art Institute, in his duced in a double-page spread. foreword. “It is quite fi tting that the Wrights had the foresight to employ photography in Both the text and the photographs follow the ca- their endeavors. … Not only did they use pho- reers of the brothers as they continued to develop tography as a documentary tool, but they took later airplane models and worked to create a suc- great pains to ensure that it was dramatic and cessful business from their invention. Although aesthetically powerful. The beauty and majesty they never fully capitalized on their work or cre- of the ‘fi rst fl ight’ photograph is not the result ated a long-lasting aviation manufacturing com- of good fortune alone. The Wrights planned to pany, the Wright brothers are recognized today as capture that fl eeting moment on fi lm. And they outstanding inventors and craftsmen who turned were successful.” a page of history and changed the ways that the world pursues business, pleasure, national defense The book was published in connection with and many other aspects of life. an exhibition at the Dayton Art Institute, of which Orville Wright was a founding trustee. It includes photographs — black-and-white, sepia Reports and hand-tinted — by, and of, the Wright brothers. Wright-related material, including postcards, The Airport System Planning Process. U.S. news photographs, magazine covers and souvenirs, Federal Aviation Administration (FAA) also is represented. Advisory Circular (AC) 150/5070-7. Nov. 10, 2004. Figures, appendixes, glossary, references. “Starting about 1898, Wilbur and Orville pur- 83 pp. Available from FAA via the Internet at chased a four[-inch] by fi ve[-inch] [10-centimeter or by mail.* by 13-centimeter] glass-plate camera and built a darkroom in a shed in their backyard,” say the au- AA says that the primary purpose of airport thors. “It began as another of their many hobbies, Fsystem planning is to study the performance but when they began fl ying gliders at Kitty Hawk and interaction of an entire aviation system to in 1900, their photographs served as a visual record understand the interrelationship of member

FLIGHT SAFETY FOUNDATION • FLIGHT SAFETY DIGEST • MAY 2005 35 R ESOURCES

airports. These may include airports serving a Caring for Precious Cargo, Part II: Behavioral large metropolitan area, a single state or several Techniques for Emergency Aircraft Evacuations bordering states within the United States. The With Infants Through the Type III Overwing airport system may include all types and combi- Exit. U.S. Federal Aviation Administration nations of airports, heliports, spaceports (opera- (FAA) Offi ce of Aerospace Medicine. DOT/ tions involving horizontally launched reusable FAA/AM-05/2. Final report. March 2005. vehicles) and seaplane bases. Corbett, Cynthia L. Figures, tables, appendixes, references. 24 pp. Available on the Internet at A systemwide planning process determines or through NTIS.** the type, extent, location, timing and costs of development needed to establish or maintain a n 1995, FAA estimated that infant enplane- viable system of airports in a specifi c geographic ments represented approximately 1 percent of area. Four main activities of the process are as I all passenger enplanements. FAA has projected follows: 80 million infant enplanements for the 10-year period, 2000–2009. An accurate number of child • Conduct an overall needs assessment; passengers fl ying on U.S.-registered carriers is • Determine cost estimates and funding not easily obtainable by airlines or government sources; organizations because children younger than two years can fl y seated on an adult’s lap and without • Identify standards prescribed by federal, state, a purchased ticket. regional and local governing bodies (e.g., environmental considerations); and, The report says that there are few recommended procedures for management of these “invisible” • Implement studies, surveys and other actions passengers in emergencies, and their effect on to determine specific needs. emergency evacuations is generally unknown. The intent of this study was to identify a set The product of the process is a cost-effective plan of recommended procedures for passengers of action consistent with established goals and with infants and small children to follow when objectives that contributes to local and national evacuating an airplane using Type III overwing transportation systems and serves aviation-user exits. requirements. Simulated overwing evacuations were conducted This AC is a guidance document, organized to with adult passengers, some of whom carried guide planners through the process. The intended anthropomorphic dummies representing infants audience includes members of the public, govern- ranging in age from two months to 24 months. ment agencies and regulatory entities, planning During the fi rst and last trials, no instructions were commissions, airport proprietors, and members given as to how the dummies should be carried. of the aviation community. In intervening trials, passenger carriers were told to follow printed instructions for carrying infant The AC identifies reports, books and other dummies horizontally, carrying them vertically reference materials from academia, special- and passing them to people who already had ex- interest organizations, FAA, the U.S. Department ited the passageway. of Transportation, and state and local organizations. Results showed that dummy size, carrying method [The Airport System Planning Process cancels and maneuvers through an exit with an infant af- AC 150/5050-3B, Planning the State Aviation fected egress speed. Carrying an infant dummy System, dated January 1989, and AC 150/5070- either horizontally or vertically permitted faster 5, Planning the Metropolitan Airport System, egress than passing the dummy through the exit dated May 1970. The Airport System Planning to a waiting adult. Process AC incorporates guidance information from previously canceled AC 150/5050-5, dated Most participants said that carrying an infant ver- November 1975.] tically against the body was the easiest method,

36 FLIGHT SAFETY FOUNDATION • FLIGHT SAFETY DIGEST • MAY 2005 R ESOURCES except for larger, 24-month dummies. Larger LASORS 2005 (LAS: Licensing Administration and dummies were slightly easier to pass through an Standardization; ORS: Operating Requirements exit to a waiting person. “Results confi rm that and Standards) explains the responsibilities and passing an infant to another participant pro- requirements of JAA and U.K. national licenses duces slower egress than carrying the infant,” the and associated ratings and explains administrative report says. The report also compares individual procedures for issuance, revalidation and renewal. egress times for passengers with infants and for LASORS 2005 is designed “to give pilots a one-stop passengers without infants, noting that egress with reference for all aspects of safe airplane operation,” infants did not signifi cantly infl uence egress times the CAA says. of other individuals. The “LAS” section combines into one source all Participants also identifi ed risks of injuries to fl ight crew licensing information from the JAR- infants regardless of the method used, such as an FCL, the U.K. ANO, Aeronautical Information infant’s head or limbs striking the exit frame or Circulars (AICs) and the old U.K. Civil Aviation an infant being dropped as it was passed through Papers (CAPs) 53/54. an exit to another adult. LAS encompasses licensing and standardization An earlier study (DOT/FAA/AM-1/18) had been procedures as applied by the U.K. CAA Safety similarly conducted using Type I fl oor-level exits Regulation Group. It provides guidance on how with infl atable escape slides. Information ob- licenses and ratings are obtained and revalidated. tained from both studies will be used to develop Subparts link information to locations within passenger-education materials, pre-evacuation JAR-FCL documentation. briefi ngs, safety cards and training programs about the safest and most effi cient techniques The following licenses are discussed: fl ight radio for emergency evacuation of all passengers, es- telephony operator, private pilot, commercial pecially infants. pilot, airline transport pilot and fl ight engineer. Ratings discussed include instrument, IMC (instrument meteorological conditions) and night Regulatory Materials qualifi cation; type and class; and instructor.

The “ORS” section focuses on private pilots of LASORS 2005. U.K. Civil Aviation Authority airplanes. The CAA says, however, that the in- (CAA), 2005. Figures, tables, appendixes, formation and advice are relevant to all pilots illustrations, photographs, glossaries, references, whatever their experience or type of aircraft index, cross-references. 608 pp. Available on fl own. Information specifi c to helicopter pilots the Internet at or from The and balloon pilots is included. Stationery Offi ce (TSO).*** LASORS is published annually. Interim changes, he U.K. CAA is empowered by the U.K. additions and updates to regulations and proce- TAir Navigation Order (ANO) to grant dures are transmitted through AICs and also are European Joint Aviation Authorities (JAA) and published on the Internet. U.K. flight crew licenses and associated ratings, as appropriate. Standards for Airport Markings. U.S. Federal Aviation Administration (FAA) Advisory This book says, “A holder of a Joint Aviation Circular (AC) 150/5340-1H. Change 2. Dec. 6, Regulations–Flight Crew License (JAR-FCL) is 2004. References. 12 pp. Available from FAA via entitled to act as a member of fl ight crew in an the Internet at or from aircraft registered in JAA member states within the U.S. Government Printing Offi ce.**** the privileges of the license or rating. A holder of a U.K. national license is entitled to act as a AA says, “By Jan. 1, 2007, airport operators member of fl ight crew in aircraft registered in the Fholding an airport operating certifi cate issued U.K. within the privileges of the license or rating under U.S. Federal Aviation Regulations (FARs) concerned.” Part 139, Certifi cation of Airports, must comply

FLIGHT SAFETY FOUNDATION • FLIGHT SAFETY DIGEST • MAY 2005 37 R ESOURCES

with POFZ/TERPS [precision obstacle-free list of products eligible for U.S. importation is zone/terminal instrument procedures] marking included with notations specifi c to participat- and sign standards.” ing countries — product types (e.g., aircraft, engines and parts), agreement dates and other Change 2 to the AC, Standards for Airport information. Markings, refl ects revisions to AC 150/5300-13, Airport Design. Change 2 incorporates new man- This AC offers guidance on some of the most datory hold markings, primarily in the section common situations encountered in the design- titled “Holding Position Markings for Instrument approval process for FAA type certifi cation or Landing System (ILS)/Precision Obstacle-free for FAA Technical Standard Order (TSO) design Zone (POFZ).” approval. Guidance for obtaining FAA airworthiness certifi cation or approval of civil aeronautical Changes are made to the standards for marking the products to be imported to the United States has boundary of the POFZ and to holding positions been updated to refl ect the latest BASA IPAs. of Category (CAT) II/III operations. References to related FAA documents, contact Revisions to related ACs and FAA manuals, such information for FAA directorates and their as- as new separation standards for some taxiways signed product responsibilities, and Internet ad- and changes in sign standards, are refl ected in dresses for obtaining the text of BAAs and BASAs Change 2. Other affected FAA documents are are included. identifi ed. [This AC cancels AC 21-23A, Airworthiness Airworthiness Certifi cation of Civil Aircraft, Certifi cation of Civil Aircraft, Engines, Propellers, Engines, Propellers, and Related Products and Related Products Imported to the United States, Imported to the United States. U.S. Federal dated Oct. 20, 2000.] Aviation Administration (FAA) Advisory Circular (AC) 21-23B. Nov. 17, 2004. Figures, appendixes, references. 68 pp. Available from Sources FAA via the Internet at * U.S. Federal Aviation Administration or from the U.S. Department of Transportation 800 Independence Ave. SW (USDOT).***** Washington, DC 20591 U.S. ** National Technical Information Service (NTIS) he AC says, “FAA does not issue standard 5285 Port Royal Road Tairworthiness certificates, nor grant air- Springfi eld, VA 22161 U.S. worthiness approvals, for aeronautical products Internet: manufactured in a country with which the *** The Stationery Offi ce [United States] does not have a BAA [Bilateral TSO–London Airworthiness Agreement] or a BASA [Bilateral 51 Nine Elms Lane Airworthiness Safety Agreement] with IPA London SW8 5DR U.K. [Implementation Procedures for Airworthiness] Internet: for the kinds of products concerned. [FAA] must **** U.S. Government Printing Offi ce issue a type certifi cate prior to the issuance of an 732 N. Capitol Street NW FAA standard airworthiness certifi cate.” Washington, DC 20401 U.S. Internet: BAA and BASA agreements are described in gen- ***** U.S. Department of Transportation eral terms, and countries that are exceptions are Subsequent Distribution Offi ce noted. Bilateral agreements related to airworthi- Ardmore East Business Center ness are listed by country, agreement type, and 3341 Q 75th Avenue application or scope of U.S. acceptance. A tabular Landover, MD 20785 U.S.

LIBRARY

38 FLIGHT SAFETY FOUNDATION • FLIGHT SAFETY DIGEST • MAY 2005 ACCIDENT/INCIDENT BRIEFS

Severe Vibration Accompanies Braking During Landing Rollout

After the landing at an airport in Wales, ground personnel found what appeared to be brake parts on the runway.

— FSF EDITORIAL STAFF

he following information provides an “Almost immediately, the crew experienced very awareness of problems through which heavy vibration and felt the aircraft pulling to the such occurrences may be prevented in left,” the report said. “Braking application was re- Tthe future. Accident/incident briefs are duced, and the vibration lessened. Firm braking based on preliminary information from govern- was again applied at an estimated 60 knots, and ment agencies, aviation organizations, press infor- extremely heavy vibration again occurred. Both pi- mation and other sources. This information may lots felt a signifi cant lateral acceleration to the left. not be entirely accurate. … The commander used the tiller in an attempt to regain the runway centerline, and he brought the aircraft to a halt on the runway.” Investigation Reveals Broken Aircraft rescue and fi re fi ghting personnel found Landing-gear Torsion Link no fi re but said that brake parts appeared to be on Boeing 737. Minor damage. No injuries. the runway. The airplane was shut down on the

AIR CARRIER runway, and passengers were disembarked. fter a night fl ight from Spain to Wales, the crew Aconducted an instrument landing system (ILS) One day before the incident, another pilot said approach to Runway 30 with anti-skid “ON,” auto- that after a normal landing, he felt “an unusual brakes “OFF” and 30 degrees of fl ap. Winds were juddering” through the rudder pedals when he from 040 degrees at 20 knots, gusting to 31 knots. applied heavy braking. The vibration stopped as braking was eased. The crew said that the landing was “fi rm but not heavy and without excessive sideways drift.” The Examination of the airplane revealed that the crew selected reverse thrust and applied the wheel lower torsion link on the left main landing gear brakes. had broken “at a point where a cutout in the fl ange

FLIGHT SAFETY FOUNDATION • FLIGHT SAFETY DIGEST • MAY 2005 39 A CCIDENTS/INCIDENTS

of the link formed two ‘T’ section limbs, both of possible to adjust the [horizontal stabilizer] if which had fractured,” the report said. “The frac- PTS 1 was disengaged, and if AP 2 was engaged, tures had resulted from overload, approximately the [horizontal stabilizer] was always set to ‘pitch in the plane of the link, with the left limb having down.’” The irregularities were noted in the tech- failed fi rst and the right limb fracture showing nical logbook. signs of very low-cycle, very high-stress load reversals.” Repairs to the PTS 1 irregularity were postponed, with reference to the minimum equipment lists, The report said that the manufacturer had attrib- because of lack of time. The PTS 2/AP 2 irregular- uted similar failures over a number of years to ity was checked, no fault was identifi ed, and the excessive wear of the torsion link apex joint. problem was signed off as “fi xed.” The airplane was returned to service.

Delayed Repairs Cited in The report said that causes of the incident were: Excess-trim Incident Airbus A300-600. No damage. • “As a result of the deferred elimination of a One minor injury. fault on PTS 1, the AP could be operated with PTS 2 only; uring a climb to cruise altitude with the Dautopilot (AP) engaged, the fl ight crew for • “There was a fault on PTS 2 for which there the domestic fl ight in Germany observed that the was no confirmation or elimination; airplane was about to exceed the maximum al- lowable airspeed (VMO). They reduced the preset • “At a certain airspeed, the signal interruption speed and selected a higher climb rate, then ob- between engaged AP 2 and PTS 2 caused a served that the airspeed was continuing to increase continuous change of the [horizontal stabi- and that the nose was beginning to pitch down. lizer] in the direction of pitch down; They disengaged the AP and, while hand-fl ying the airplane and establishing the proper fl ight at- • “Because of a system deficiency caused by titude, found that excessive nose-down trim had [a software error], the continuous change been applied. of the [horizontal stabilizer] did not result in a warning and the self-deactivation of the “A great amount of control force had to be ap- system; [and,] plied until the wrong trim could be correct[ed] by means of the electrical trim device,” the incident • “The prescribed procedure for abnormal report said. “Vertical acceleration was so great functions of the horizontal stabilizer [trim] during the re-establishment of the original fl ight was not executed in time.” attitude that one [cabin] crewmember fell and injured herself slightly. The fl ight was continued with disengaged AP and no further incidents.” Engine Damaged By High-pressure In the A300-600, the AP moves the elevator, and Compressor-blade Failure the trim system 1 (PTS 1) adjusts the horizontal stabilizer; the report said that these functions are Boeing 767. Minor damage. No injuries. intended to maintain the horizontal stabilizer in the neutral position. In the event that PTS 1 is not fter departure from an airport in Australia on available or that it fails, a backup trim system (PTS Aa fl ight to Brunei, as the crew fl ew the airplane 2) controls the horizontal stabilizer. through 11,000 feet, they heard a loud bang and observed an elevated exhaust gas temperature The airplane’s APs were checked during periodic (EGT), accompanied by a decrease in power from maintenance two days before the incident, and a the right engine. malfunction was corrected by changing components. During a fl ight the day before the incident, They then observed that the right EGT indicated two irregularities were observed: “It was only 662 degrees Celsius (C; 1,224 degrees Fahrenheit

40 FLIGHT SAFETY FOUNDATION • FLIGHT SAFETY DIGEST • MAY 2005 A CCIDENTS/INCIDENTS

[F]), 12 degrees C (22 degrees F) higher than the the ground through a hole in the clouds at about maximum operating temperature, for 22 seconds. 400 feet, descended and landed the airplane at They closed the throttle for the right engine; de- the airstrip. clared mayday, a distress condition; and returned to the departure airport for a single-engine, over- A maintenance inspection revealed a loose fuel weight landing. line on the right engine, which had caused the engine’s fuel starvation. The pilot observed An investigation revealed that the high-pressure that the fl aps were extended about fi ve degrees, section of the engine could not be rotated and although the fl ap selector was in the retracted that there was molten metal debris in the exit position. screens of a bleed-air valve and metal spray in the exhaust duct. The accident report said that “The pilot believed that the aerodynamic drag the engine damage was “consistent with the produced by the fl aps in that position would have failure of a sixth-stage [high-pressure com- contributed to the inability to maintain altitude pressor] blade.” The report said that the blade with one engine inoperative and may also have had failed because of “high cyclic stress … in caused the shuddering during the takeoff,” the turbulent airflow created by [an] off-schedule report said. stator-vane angle.” The airplane owner said that the fl aps had been in that position “for some time”; they were repaired after the incident. Loose Fuel Line Cited in Engine Failure Maintenance personnel found no problem with Piper PA-31-350 Chieftain. No damage. the primary attitude indicator. No injuries.

aytime instrument meteorological condi- Cockpit Fills With Dtions prevailed for the fl ight from an air- Smoke During Approach AIR TAXI/COMMUTER port in Australia. Soon after departure, the pilot Cessna 550 Citation II. Minor damage. observed “light aerodynamic shuddering” through No injuries. the airframe; he considered this “an idiosyncrasy of this particular aircraft” and continued the fl ight, ighttime visual meteorological conditions the report said. Nprevailed as the flight crew conducted a positioning fl ight to England from an airport Later, the right propeller speed fl uctuated and in Scotland. The crew received vectors from air the right engine misfired. The pilot adjusted traffi c control for the approach to the destination the propeller lever for the right engine, checked airport. When the airplane was at 8,000 feet, the the fuel fl ow and initiated a climb. At 9,000 feet, crew smelled burning electrical insulation and the misfi ring by the right engine increased, the observed that the passenger cabin was fi lled with exhaust-gas-temperature gauge indicated “above smoke. the redline and in excess of normal operating parameters,” and the fuel-fl ow indication was The fi rst offi cer donned his oxygen mask while decreasing, the report said. the captain declared mayday, a distress condition. His communication with air traffi c control was The pilot shut down the right engine and feath- hindered by breathing problems caused by the ered the right propeller; declared pan-pan, an thick smoke entering the cockpit. The smoke also urgent condition; and prepared for landing at an obscured the instrument panel and forward vi- en route airstrip. The pilot said that during the sion. The captain began the “SMOKE REMOVAL” diversion, he was unable to prevent the airplane checklist from memory, opening the dump valve from descending below the lowest safe altitude. He to depressurize the airplane and partially clear the believed that the attitude indicator was providing smoke, but did not complete all checklist proce- incorrect information, so he began using the at- dures because he considered landing the airplane titude indicator on the copilot’s panel, observed to be the top priority.

FLIGHT SAFETY FOUNDATION • FLIGHT SAFETY DIGEST • MAY 2005 41 A CCIDENTS/INCIDENTS

The crew landed the airplane at the destination that conformed to specifi cations for road tar. The airport as smoke increased in the cockpit. They pilot said that the ineffective braking action was a stopped the airplane on the runway, shut down result of the treatment. both engines and switched off all electrical power. Aircraft rescue and fi re fi ghting person- During the accident investigation, an inspector nel arrived as the crew exited the airplane. An examined the runway surface and described it as inspection revealed no fi re, but smoke continued “having the consistency of putty.” In a skid test to emerge from the open cabin door for about using highway vehicle test equipment, the friction 20 minutes. rating was categorized as “poor.”

An inspection revealed that the circuit breaker The probable cause of the accident was “the for the motor of the cabin defog fan blower had poor friction value on the runway rendering tripped; the motor was found to be defective. braking action inadequate.” Contributing factors were “excessive approach speed and the The report said that completion of the emer- pilot’s inability to maintain directional control gency checklists might have removed power of the aircraft.” from the defective fan and minimized the smoke. Nevertheless, the report said that the captain’s decision not to complete all relevant emergency Airplane Strikes Terrain procedures “was inﬂ uenced by the time available During Crosswind Landing and the need to concentrate on the approach and Cirrus Design SR-22. landing.” Substantial damage. No injuries.

aytime visual meteorological conditions prevailed for the business fl ight in the United Runway Surface Cited in D States, and no fl ight plan had been fi led. The pilot Loss of Braking Action said that before landing, he received “computer Cessna 310J. Substantial damage. weather information” and information from the No injuries. destination airport’s automated weather observing system. The preliminary accident report said that aytime visual meteorological conditions pre- the pilot then told his passenger that he would Dvailed for the business fl ight in the United “make one attempt at the crosswind landing and States. The pilot said that he was fl ying a “standard if he was unable to maintain the runway heading, approach” to Runway 21 at an approach speed be- they would divert.” CORPORATE/BUSINESS tween 105 knots and 110 knots, with 35 degrees of fl ap. The airplane touched down about 500 feet The maximum demonstrated crosswind compo- to 700 feet (153 meters to 214 meters) past the nent for the accident airplane was 20 knots. threshold of the 3,800-foot (1,159-meter) runway, and the pilot immediately applied the brakes. The The report did not provide wind information pilot described braking action as “ineffective.” The for the destination airport but said at the time airplane departed the right side of the runway and of the accident, winds at the closest weather- struck a hangar, then spun 120 degrees and struck reporting facility, 20 statute miles (32 kilometers) a fuel truck. east-southeast of the accident site, were from 170 degrees at 18 knots, gusting to 26 knots. The aircraft landing performance chart indicates that, with 35 degrees of fl aps and in no-wind con- The pilot said that he “used half fl aps for the ditions, the airplane requires between 766 feet (234 landing approach and was maintaining the run- meters) and 1,114 feet (340 meters) for landing. way heading using full left rudder when a gust (The report did not discuss wind conditions at the of wind moved the airplane left of the runway time of the accident.) centerline.” The pilot applied power, and the left wing struck the ground about 10 feet (three me- The runway surface had been treated eight months ters) left of the runway; the airplane’s nose also before the accident with a coal-tar sealer/rejuvenator struck the ground.

42 FLIGHT SAFETY FOUNDATION • FLIGHT SAFETY DIGEST • MAY 2005 A CCIDENTS/INCIDENTS

The pilot was unable to restore engine power, so Section of Elevator he moved the propeller to “full coarse pitch” and Separates During Flight landed the airplane in a fi eld. He used his cellular Cessna 402C. Substantial damage. telephone to call family members who had been No injuries. watching from the airstrip and air show colleagues, who sent the crew of the air show’s emergency- isual meteorological conditions prevailed just response helicopter to help the pilot. Vafter sunset for a fl ight in the United States. About 70 nautical miles (130 kilometers) north The accident report said, “Circumstantial evi- of the destination airport, the pilot began a 500- dence suggests the engine power loss was caused foot-per-minute descent at 185 knots. by a short period of fuel starvation … probably caused by aerobatic maneuvering, momentarily OTHER GENERAL AVIATION About fi ve nautical miles (nine kilometers) north uncovering the outlet in the left fuel tank, which of the airport, the airplane “pitched sharply nose was selected at the time.” down, with an uncontrollable back-and-forth oscillation of the control yoke,” a preliminary Pilots of Fish-patrol Aircraft report said. “A loud shearing noise was heard from the right rear of the aircraft before pitch Land Safely After Collision control was regained. The oscillation lasted for Cessna 185. Minor damage. No injuries. about fi ve seconds.” Cessna 185. Minor damage. No injuries.

When the pilot looked to the rear of the airplane, aytime visual meteorological conditions pre- he observed “sheet metal fl apping in the wind near Dvailed for the fl ights of two fl oat-equipped the elevator section,” the report said. The pilot airplanes in Canada in support of different fi shing declared an emergency, extended the landing gear operations. The pilot of one airplane was on a private and fl aps, slowed the airspeed and conducted a business fl ight in support of company fi shing vessels landing. and was monitoring two radio frequencies. The pilot of the second airplane was conducting a charter fl ight A preliminary examination revealed that about for a government agency to observe herring spawn 16 inches (41 centimeters) of the right elevator’s and was monitoring a different radio frequency. outboard area was missing. The missing section was found about fi ve nautical miles north of the The pilot of the charter airplane began a left turn airport in a residential area. to land near a boat operated by the government agency; he did not see the other airplane. The pilot The report said, “The remaining outboard of the of the private airplane saw the charter airplane too elevator up to the inboard attaching hinge was peeled late to avoid the collision, which occurred 400 feet up and aft. The bolt connecting the elevator trim tab above the water. Both pilots maintained control of to the elevator trim actuator rod was missing.” their airplanes; they established radio contact with each other, and each assessed the damage to the other airplane. They then landed the airplanes. Engine Fails During Aerobatic Maneuvers The accident report said that any of fi ve radio fre- North American P-51D-20 Mustang. quencies would have been appropriate for pilots Minor damage. No injuries. operating in the area, including the three frequencies being monitored by the two pilots. aytime visual meteorological conditions Dprevailed for the aerobatics display at an air “Risks associated with visual fl ight are well docu- show in England and for the fl ight to return the mented,” the report said. “These risks are known airplane to its home airstrip. As the pilot fl ew the to be mitigated by radio position reporting; such airplane over the airstrip, he performed several reporting requires aircraft to be on a common fre- aerobatic maneuvers. Then, while beginning a quency. … Independent fl ight operations in the “moderately steep” climb from an altitude of less same area but operating on different frequencies than 1,000 feet, the engine failed. increase the risks associated with visual fl ight.”

FLIGHT SAFETY FOUNDATION • FLIGHT SAFETY DIGEST • MAY 2005 43 A CCIDENTS/INCIDENTS

of a red “GOV” light and changed the operations Faulty Training Cited in manual to show that helicopter pilots were re- Mishandling of Emergency quired to have a profi ciency check ride every two years and a route check in alternate years. Eurocopter AS 350B3. Minor damage. No injuries. ROTORCRAFT Helicopter Strikes Mountain on he helicopter was being fl own at 700 feet off Flight in Clouds, Fog Tthe eastern coast of Canada for surveillance of a lobster-fi shing dispute. As the pilot conducted Bell 206B JetRanger. Destroyed. a right turn, the red “GOV” (governor) warning Four fatalities. light illuminated and the cockpit alarm sounded. The pilot continued the turn and headed toward he helicopter was being fl own on a daytime the shore, where he planned a precautionary land- Theli-skiing fl ight in Switzerland and was trans- ing. He observed an increase in rotor revolutions porting three passengers to a mountain glacier. The per minute (rpm) above the limit and felt a severe helicopter struck a mountain at 10,400 feet. rotor vibration. At the time of the accident, fog and low clouds “The pilot lowered the collective and reduced prevailed in the area, with mountain peaks twist-grip throttle, but there was no apparent obscured. reduction in rotor rpm,” the incident report said. “Believing that manual control of the throttle was lost, the pilot reopened the throttle to the Stuck-pedal Demonstration ‘FLIGHT’ detent and tried to reach the overhead Ends in Rollover Accident fuel-control mode-selector switch to move it to the Robinson R22 Beta II. Minor damage. manual position; however, the severe vibrations No injuries. made it diffi cult to activate the caged switch.” isual meteorological conditions and calm The pilot then attempted to decrease main-rotor winds prevailed for the instructional fl ight rpm by raising the collective, but there was no V in South Africa. The instructor intended to dem- apparent change, and the helicopter continued its onstrate to the student what might happen if the rapid descent. After landing, a severe ground reso- right pedal were to become stuck while applying nance (destructive vibration induced by rotor-blade power during hover fl ight. oscillation) developed; the pilot hovered the helicopter, but the vibrations continued, and he again “The recovery technique involved raising the collec- landed the helicopter and activated the fuel shutoff tive while simultaneously rolling off the throttle and lever on the ceiling to shut down the engine. reducing main[-rotor revolutions per minute (rpm)] The report said that the pilot “had not received and tail-rotor rpm to keep the helicopter aligned with adequate fl ight training for the red ‘GOV’ light the fl ight path,” the accident report said. “The inten- emergency and did not realize that the twist-grip tion was to land the aircraft following the exercise; throttle still controlled fuel fl ow to the engine. however, during the attempted landing, the left skid Consequently, the emergency was mishandled, gear touched down fi rst, positively, with slight lateral resulting in a severe overspeed of the aircraft’s drift, which resulted in a rolling motion from which dynamic components.” the instructor was unable to recover.”

At the time of the incident, the operator required The main-rotor blades struck the ground, and the neither a pilot-proficiency check nor a pilot- helicopter fell onto its left side. competency check, and the operator was unaware that the pilot had received “less-than-adequate The report said that the probable cause of the ac- training” on the helicopter type, the report said. cident was that the instructor allowed the left-skid- ﬁ rst touchdown on an uneven surface and that After the incident, the operator issued a memoran- he was unable to overcome the student’s “strong dum to its AS 350B3 pilots to discuss illumination reaction on the controls.” ■

44 FLIGHT SAFETY FOUNDATION • FLIGHT SAFETY DIGEST • MAY 2005 Now you have Flight Safety Foundation the safety tools

Approach-and-landing Accident Reduction to make a difference. Tool Kit

The Flight Safety Foundation is a comprehensive and practical resource on compact disc to help you prevent the leading causes of fatalities in com mer cial aviation: approach-and-landing ac ci dents (ALAs), including those involving controlled ﬂ ight into ter rain (CFIT).

Put the FSF to work for you TODAY! • Separate lifesaving facts from fi ction among the data that confi rm ALAs and CFIT are the leading killers in avi a tion. Use FSF data-driven studies to reveal eye-opening facts that are the nuts and bolts of the FSF ALAR Tool Kit. • Volunteer specialists on FSF task forces from the international aviation industry studied the facts and de veloped data-based con clu sions and recommendations to help pilots, air traffi c controllers and others prevent ALAs and CFIT. You can apply the results of this work — NOW! • Review an industrywide consensus of best practices included in 34 FSF ALAR Briefi ng Notes. They provide practical in for ma tion that every pilot should know … but the FSF data confi rm that many pilots didn’t know — or ignored — this information. Use these benchmarks to build new standard operating pro ce dures and to im prove current ones. • Related reading provides a library of more than 2,600 pages of factual information: sometimes chilling, but always useful. A versatile search engine will help you explore these pages and the other components of the FSF ALAR Tool Kit. (This collection of FSF publications would cost more than US$3,300 if purchased individually!) • Print in six different languages the widely acclaimed FSF CFIT Checklist, which has been adapted by users for ev ery thing from checking routes to evaluating airports. This proven tool will enhance CFIT awareness in any fl ight department. • Five ready-to-use slide presentations — with speakers’ notes — can help spread the safety message to a group, and enhance self-development. They cover ATC communication, fl ight op er a tions, CFIT prevention, ALA data and ATC/aircraft equipment. Customize them with your own notes. • An approach and landing accident: It could happen to you! This 19-minute video can help enhance safety for every pilot — from student to professional — in the approach-and-landing environment. • CFIT Awareness and Prevention: This 33-minute video includes a sobering description of ALAs/CFIT. And listening to the crews’ words and watching the accidents unfold with graphic depictions will imprint an un for getta ble lesson for every pilot and every air traffi c controller who sees this video. • Many more tools — including posters, the FSF Approach-and-landing Risk Awareness Tool and the FSF Approach-and-landing Risk Reduction Guide — are among the more than 590 mega bytes of in for ma tion in the FSF ALAR Tool Kit. An easy-to-navigate menu and book marks make the FSF ALAR Tool Kit user- friendly. Applications to view the slide pre sen ta tions, videos and pub li ca tions are included on the CD, which is designed to operate with Microsoft Windows or Apple Macintosh operating systems. Order the FSF : Member price: US$40 Recommended System Requirements: Nonmember price: $160 Windows® Mac® OS Quantity discounts available! • A Pentium®-based PC or compatible computer • A 400 MHz PowerPC G3 or faster Macintosh computer • At least 128MB of RAM • At least 128MB of RAM • Windows 98/ME/2000/XP system software Contact: Ahlam Wahdan, • Mac OS 8.6/9, Mac OS X v10.2.6–v10.3x Mac OS and Macintosh are trademarks of Apple Computer Inc. registered in the United States and other countries. Microsoft and Windows are either registered trademarks or trademarks membership services coordinator, of Microsoft Corp. in the United States and/or other countries. +1 (703) 739-6700, ext. 102. The FSF ALAR Tool Kit is not endorsed or sponsored by Apple Computer Inc. or Microsoft Corp. What can you do to improve aviation safety ? Join Flight Safety Foundation.

Your organization on the FSF membership list and Internet site presents your commitment to safety to the world.

Flight Safety Foundation • Receive 54 FSF periodicals including Accident Preven tion, Cabin Crew Safety and Flight Safety An independent, industry-support ed, nonproﬁ t orga ni za tion for the Digest that members may reproduce and use in exchange of safety information their own publications. for more than 50 years • Receive discounts to attend well-estab lished safety seminars for airline and corporate aviation managers.

• Receive member-only mailings of special reports on important safety issues such as controlled ﬂ ight into terrain (CFIT), approach-and-landing accidents, human factors, and fatigue coun ter measures.

• Receive discounts on Safety Services including operational safety audits.

Want more information about Flight Safety Foundation? Contact Ann Hill, director, membership and development by e-mail: or by telephone: +1 (703) 739-6700, ext. 105.

Visit our Internet site at .

We Encourage Reprints Articles in this publication, in the interest of aviation safety, may be reprinted in whole or in part, but may not be offered for sale directly or indirectly, used commercially or distributed electronically on the Internet or on any other electronic media without the express written permission of Flight Safety Foundation’s director of publications. All uses must credit Flight Safety Foun da tion, Flight Safety Digest, the speciﬁ c article(s) and the author(s). Please send two copies of the reprinted material to the director of pub li ca tions. These restrictions apply to all Flight Safety Foundation publications. Reprints must be ordered from the Foundation. For more information, contact the director of publications by telephone: +1 (703) 739-6700, ext. 116; or by e-mail: . What’s Your Input? In keeping with the Foundation’s independent and nonpar ti san mission to disseminate objective safety infor ma tion, FSF publications solicit credible contri bu tions that foster thought-provoking dis cus sion of aviation safety issues. If you have an article proposal, a completed manuscript or a technical paper that may be appropriate for Flight Safety Digest, please contact the director of publications. Rea son able care will be taken in handling a manuscript, but Flight Safety Foundation assumes no responsibility for material submitted. The publications staff reserves the right to edit all pub lished submis sions. The Foundation buys all rights to manuscripts and payment is made to authors upon publication. Contact the Publications Depart ment for more in for ma tion. Flight Safety Digest Copyright © 2005 by Flight Safety Foundation Inc. All rights reserved. ISSN 1057-5588 Suggestions and opinions expressed in FSF pub li ca tions belong to the author(s) and are not nec es sar i ly endorsed by Flight Safety Foundation. This information is not intended to supersede operators’/manufacturers’ policies, practices or requirements, or to supersede govern ment regulations. Staff: Roger Rozelle, director of publications; Mark Lacagnina, senior editor; Wayne Rosenkrans, senior editor; Linda Werfelman, senior editor; Rick Darby, associate editor; Karen K. Ehrlich, web and print production coordinator; Ann L. Mullikin, pro duc tion designer; Susan D. Reed, production specialist; and Patricia Setze, librarian, Jerry Lederer Aviation Safety Library Subscriptions: One year subscription for 12 issues includes postage and handling: US$280 for members/US$520 for nonmembers. Include old and new addresses when requesting address change. • Attention: Ahlam Wahdan, membership services coordinator, Flight Safety Founda tio n, Suite 300, 601 Madison Street, Al ex an dria, VA 22314 U.S. • Tele phone: +1 (703) 739-6700 • Fax: +1 (703) 739-6708 • E-mail: