Aircraft Performance

EUROCONTROL HindThe ability or opportunity to understand and judge an event or experience after it has occuredight16

Understanding (modern) aircraft performance

The average weight of one elephant and four ants... by Tzvetomir Blajev

Reversion – the other side of automation by Captain Ed Pooley

Blackout: New fi ndings regarding decision-making by Captain Hans-Joachim Ebermann

Winter 2012 Contents Winter 2012

HindThe ability or opportunity to understand and judge an event or experience after it has occuredight16

4 COO’S KEYNOTE Joe Sultana’s 4 European Networks by Joe Sultana keynote 6 EDITORIAL p. 4 6 The average weight of one elephant and four ants… by Tzvetomir Blajev 8 FIGHT OR FLIGHT 8 When the aeroplane is more technically advanced than you by Professor Sidney Dekker 11 TALES OF OPERATIONAL SAFETY 11 Hi Tech! by Alberto Iovino 14 THE VIEW FROM ABOVE 14 Reversion - the other side of automation by Captain Ed Pooley The view from above by Captain 16 FRONT LINE REPORTS 16 Modern Times by Eileen Senger Ed Pooley 18 Modern aircraft performance is not alone by Maciej Szczukowski p. 14 21 121.5 – SAFETY ALERTS by Richard “Sid” Lawrence 21 Safety Reminder Message ‘ICAO Aircraft Operator and Radiotelephony Designators and Abbreviations of ‘Type C’ Call Signs’ 24 Safety Reminder Message ‘Misuse of International Aeronautical Emergency Frequency 121.5 MHz’ 25 Safety Reminder Message Lessons Learned from a Power Outage Incident 26 CASE STUDY 26 The Plane Spotter by Bengt Collin 28 Comment No. 1 by Captain Ed Pooley 29 Comment No. 2 by Dragan Milanovski BLACKOUT 31 Comment No. 3 by Mike Edwards 32 Comment No. 4 by Anita Đuretić Bartolović by Captain Hans-Joachim 34 Comment No. 5 by Dirk De Winter Ebermann 36 FROM THE BRIEFING ROOM 36 Training Aircraft Performance for modern/future ATC systems p.46 by Dragan Milanovski 40 Automation, workload and safety by Captain Harry Nelson

2 FROM THE BRIEFING ROOM (cont'd) TcAS RAs 44 Water/Slush on the runway and what every tower controller should know about it by Gerard van Es p. 54 46 Blackout : New fi ndings regarding decision-making by Captain Hans-Joachim Ebermann 48 Engine failure on take-off – "Hey, What's it doing now?" by First Offi cer Sami Laine 50 Close interactions of the disconnected kind by Loukia Loukopoulos and Immanuel Barshi 54 Unnecessary TCAS RAs caused by high vertical rates before level off by Stanislaw Drozdowski, Captain Wolfgang Starke and First Offi cer Felix Gottwald 60 “So what's it doing now?" Training in new technology environment by Anne Isaac and David Lordl 63 Lost in Modernity by Jean Pariès “So what's it 66 Aircraft automation made simple by Michel Tremaud doing now?" 70 SKYBRARY DOWNLOAD p. 60 70 Thrust Reversers: Flight Crew Guidance

The success of this publication depends very much on you. We need to know what you think of HindSight. Do you fi nd the contents interesting or boring? Are the incident descriptions easy to follow or hard to understand? Did they make you think about something you hadn’t thought of before? Are you looking forward to the next edition? Are there some improvements you would like to see in its content or layout? Please tell us what you think – and even more important, please share your diffi cult experiences with us! We hope that you will join us in making this publication a success. Please send your message – rude or polite – to: [email protected] or to the postal address: Rue de la Fusée, 96 EDITORIAL TEAM B-1130 Brussels editor in Chief: Tzvetomir Blajev Messages will not be published in HindSight or communicated to others editorial Adviser: Captain Ed Pooley without your permission. Graphic designer: Frédérique Fyon

HindSight 16 Winter 2012 3 COO’s KEYNOTE European Networks

Joe Sultana is Chief Operating Offi cer of the Network Management Directorate. He graduated with an Engineering Degree from the University of Malta in 1975 and joined the Air Traffi c Services Unit in Malta in the same year. He obtained ATCO Licences in Aerodrome, Radar and Area Control and was a Watch Supervisor for four years. In 1982, he was appointed Head of Air Traffi c Services in the Maltese Department of Civil Aviation.

Dear Reader,

lease take a moment with me to share a perspective data for European ‘A’ and ‘B’ severity ATM occurrences and Pon what we are doing in the Directorate of Network the vulnerability areas were analysed. On the basis of that Management of EUROCONTROL. Very recently the Euro- analysis we now have some very good ideas where to focus pean Commission has put forward new proposals to make further eff ort and undertake targeted Operational Safety it easier for people to travel and do business within the Studies (OSS). The subject of one of the six potential OSS European Union. This is known as the Single Market Act II we are looking at has a relation with the theme of this Hind- and advocates the opening up of EU airspace so that it op- Sight – it is about the risk of aircraft operations without a erates as a single entity. At NM, the Single Sky concept can transponder or with a malfunctioning one. only become a reality if those of us, involved in ATM network management strive to ensure that the network oper- Operations without a transponder or with a malfunc- ates safely and effi ciently. tioning one constitute a single ‘threat’ with a potential of “passing” through all the existing safety barriers up to “see Indeed, I’m sure that European citizens would like to be re- and avoid”. One of the two incidents from the sample of assured that the ATM network is safe. Achieving and main- LoS-ER that was prevented only by the “Providence” barrier taining a safe network is a complex task that involves many involved a malfunctioning transponder. This single threat players. You will hear the argument that it is the job of the was also found in a sample of ‘A’ and ‘B’ Losses of Separation regulators to keep aviation safe, but I believe it is our com- in TMA operations. This latter sample was constructed main- mon goal as well – it should be the aim of all the actors work- ly to help validate the SAFMAP concept, but it contained an ing together with a distinct purpose. Just as public health is incident that was again prevented only by the “Providence” not only the responsibility of the Health Authorities, safety barrier because of operation without a transponder which is a not a responsibility just of the Safety Authorities. arose after a departure from Controlled Airspace into Un- controlled airspace where there was no transponder equi- To make our contribution to a safe European ATM Network page requirement. more tangible, those of us in the Directorate of Network Management of EUROCONTROL follow a structured col- The Operational Safety Studies are still to be agreed and laborative process with our stakeholders to identify opera- initiated but you, as HindSight readers, will have the oppor- tional safety priorities and thus determine what we can do tunity to read more about them in future editions. to make improvements. Consequently, we have collabora- tively established our Top 5 Operational Safety Risks and Finally, I would remind you as a HindSight reader that have selected two of these for detailed review – Runway HindSight is a communication tool within a portfolio of Incursion (RI) and Loss of Separation En-Route (LoS-ER). The activities we undertake which contribute to a safe Network. detailed review took form of dedicated workshops with six As such I would urge you to make the most of the magazine Air Navigation Service Providers during summer 2012. The by discussing its contents with your colleagues and provid- review was performed with the help of comprehensive op- ing us with your feedback. erational safety barrier models which we have developed which we call ‘Safety Functions Maps’ – SAFMAPS. These SAFMAPS were populated with representative samples of

4 HindSight 15 Spring 2012 5 EDITORIAL The average weight of one elephant and four ants...

How much is the average (or typical) you expect when you are told Mach 0.79 is the average cruis- weight of a group of one elephant and ing speed of an A320? It may help you if the aircraft fl ies just at this ‘normal’ speed without optimising it to take account four ants? of the consequences for the fl ight of maintenance costs, You can choose either one tonne (yes! – 1000kg) or 0.3 mil- passenger delays and fuel. This speed may have meaning ligrams, both are correct! Both those fi gures are valid aver- to an aircraft manufacturer or a certifi cation authority. But, ages or ‘typical’ weights. If the elephant weights 5 unless we are speaking of a great diff erence in the ranges – tonnes and the ants weigh 0.3 milligrams each like in the case of en-route spacing between a wide body jet then the arithmetic average, otherwise and a regional jet, it does not tell you a great deal about the known as mean, can be obtained by add- actual speed of the aircraft you will have on your frequency ing up all the weights and dividing by fi ve today. One can go further with this argument. It is not un- - the number of members of this strange common for procedures and safety assurance calculations group. The result is a mean value of one to assume a single performance value rather than a range tonne for every ant. But there is another within which the value can lie. The result is a rather simplis- measure of average called the “mode”, tic expectation of performance, not only for aircraft, but also which is the most frequently met weight for ATM systems. in our group – 0.3 milligrams – the weight of each of the 4 ants. One can use a Working with ranges has a clear drawback – you do not diff erent kind of average each know precisely where in the possible range the cruising time because the word “aver- speed will be. Or to take another example – in the case of age” has this loose mean- High Intensity Runway Operations at busy airports, the ing. There is also a third time which aircraft spend on the runway needs to be mi- ‘defi nition’ of an average nimised in order to achieve maximum capacity. This run- called “median” but let us way occupancy time for a landing aircraft can vary quite stop here. a lot and depends on a number of factors, including the touchdown speed of the aircraft, its deceleration capabil- All that said about the ity, the availability of Rapid Exit Taxiways, the actual brak- diff erent ways of arriving ing action, etc. This makes the expected occupancy time at an average, what can for a given aircraft only a guess, or expressed more for- Tzvetomir Blajev Editor in Chief of Hindsight Fellow of the Flight Safety Foundation

6 The average weight of one elephant and four ants...

mally, probabilistic. There are well-known human biases Another great example where mea- when our brain “rejects” thinking in strictly probabilistic surement moves the probabilistic terms. In these cases, we have a difference between what guess world of ATC towards a deter- actually happens and what we thought would “probably” ministic one is the Mode S functionality happen. for downlinking such parameters as selected altitude, ground speed, magnetic What can be done? heading, vertical rate, TCAS RA, indicated airspeed and Mach number. Having the ex- One solution is to work towards better measurement, act Mach number, you will not need to try seeking to make the world more deterministic and less and fi gure out how much the cruising speed probabilistic and turning away from the guess work. At will “dance” in the possible range around the an airport which relies on high-intensity runway opera- “average” Mach 0.79 for this A320 that you tions, you can measure the actual duration of runway oc- have now in your sector. cupancy over time and find, for example, an “average” of 50 seconds for this occupancy. This average may not be If you know the actual performance, you can enough to get the capacity you need and, on top of that, run the most optimal plan and deliver the most delivery of the desired occupancy will be very uncertain. effi cient operations. If you do not have the ac- It will vary within quite a wide range and you can get 40 tual performance to hand then you have to work seconds in some cases but also 1-2 minutes with crews with ranges. Working with ranges means we have not familiar with the airport missing the exit by just a lit- to add in extra protections, additional lines of de- tle and rolling slowly to the next available exit. Yet saving fence, one more buff er in case the actual speed just 5 seconds often provides an opportunity to add an- is Mach 0.65 or Mach 0.80. Or, as in Bengt’s case other movement to the hourly total achieved. A common study elsewhere in this issue, the big An124 is slow practice in this case is to reduce the range of possible oc- to vacate the runway and becomes another factor cupancy time by working with the aircraft operators and in the chain of events endangering safety. their crews with a view to raising awareness and encour- aging crews to expect and plan for a given exit. One cannot ask for deterministic safe and effi cient performance while leaving the front line operators working in a probabilistic world.

Or, as the astonished ant would have said looking at its “average” weight – “let’s talk about the elephant in the room...”.

HindSight 16 Winter 2012 7 FIGHT OR FLIGHT When the airplane is more technically advanced than you

by Sidney dekker “University 185, weren’t you going to climb?” I knew it. I knew the question was going to come. Here I was fl ying a small TAA, or Technologically Advanced Airplane, boring along at 2,500 feet and not climbing at all. “Ah, Centre, University 185, I’m still fi guring out the automation,” is my limp reply.

8 And I was. I was on Interestingly, there is an easy way to my own in this new make any airplane climb. I had this airplane, trying to explained to me on one of my fi rst les- fi gure out how to get sons ever. I must have been fourteen or the autopilot into Ver- so. “To climb, you pull the houses lever,” tical Speed mode, and the instructor said. “The houses lever?” to dial in the new altitude “Yeah, the houses lever. You pull, and I had been directed to go the houses get smaller. You push, and to so that it would not over- they get bigger.” “Ah.” I pulled the hous- shoot it, not even worrying es lever. And the houses got smaller. yet about how much engine power I might need to accom- But that was when I was fourteen, plish the climb, all the while and the airplane I was fl ying was any- staying on the GPS track that I thing but technically advanced. In had programmed when still on the fact, it wasn’t much of anything. Now ground. I was thirty-something and half a decade into the twenty-fi rst century and Should I not have studied all this in a I was going to get the automation to bit more detail before getting into the do what I wanted. So I did not pull airplane? I asked myself that question the houses lever. In fact, in this TAA, I too. But let me tell you something. I was afraid of pulling the houses lever. did. And it didn’t help much. When What would happen to all the care- was the last time a guy read the man- fully programmed tracks and restric- uals of a set of interrelated and tech- tions and waypoints and everything nically complicated devices, while not that I had so meticulously put into the having those devices in action or in machine before take-off ? Would I ever interaction, and actually knew what fi nd it again? I was motoring my way to do? Let me know if you are that to a rather big international airport, guy. It would be nice to meet you. granted still at 2,500 feet, and I found it very nice to know that I had all this au- In the meantime, I was getting tomation watching my fl ight for me. I trapped. Trapped into the belief that did not want to risk fl ushing it all away. the best way to solve my automation And of course, I truly thought that problem was to try more automa- there was a way to get this thing to tion. Lots of pilots do that. They think: do what I wanted. I’m that kind of guy, “There is a way to get this thing to do what can I say? Again, tell me if you’re this. I know there is a way.” And then not. It would be nice to meet you. both pilots go heads-down some more and pound away at the keys of Now the controller in this saga was ac- the fl ight management system. And tually very patient. And perhaps that is the airplane either does something the right thing to be – you have that that nobody had expected, or stub- luxury of course. Pilots do not typically bornly keeps doing what it was doing make their automation or their air- without responding to the pilots’ ever planes do funny things because they more insistent pleas to the contrary. are deliberately bloody-minded. They My TAA was doing the latter. It re- themselves get surprised by the auto- fused to climb. mation. 

HindSight 16 Winter 2012 9 FIGHT OR FLIGHT

If pilots are to avoid A decade before my hunt for the verti- automation surprises, manuals that are chock-a-block with cal speed mode in that TAA at 2,500 feet, static, dead details of a machine not in I had been getting my doctorate at The which surprise not only action, and not in interaction. Go and Ohio State University. Researchers there study, see you tomorrow in computer- were working hard on documenting them but controllers too, based training. And then, after that and and trying to understand automation after all those simulator sessions, you’re surprises in the cockpit. Automation then they have to have outside, in the airplane, on the line, surprises, they concluded, happen when an accurate model of how where you will learn the rest. Or learn the automation does something on its how these things really operate. own (or refuses to do something) with- the system works. out immediately preceding pilot input. As I said, we haven’t risen to the chal- It may refuse to comply with a limit on lenge yet. And this is perhaps even a level crossing, for example. Or it may If pilots are to avoid automation sur- more the case for the TAAs that get refuse to climb. Or it may suddenly level prises, which surprise not only them fl own outside of airline supervision and off , with two pilots looking at it, and then but controllers too, then they have to training centres, by pilot-owner-opera- each other, going, “Did you make it do have an accurate model of how the tors who have nobody to tell them this that?” And, of course, neither did make system works. They have to call to mind or that, except perhaps their insurance it do that. It was an automation surprise. the portions of this knowledge, this company. model, that are relevant for the cur- One of the problems of technologically rent situation. They have to recall past If you are the controller, maybe the advanced airplanes (both big and small) instructions to the automation, which best thing to do is try to be patient. If is that indications about the future be- may have occurred when they were you can. Have patience with the guy haviour of the automation are typically still on the ground, or at least some who did not read the manual. Or did weak. There is still no obvious vertical time ago, and which may have been read it and found it to be rather useless. profi le on display in most automat- put in by somebody else. They have to As we have seen in a recent accident, ed cockpits, for example. The be aware of the current and the pro- airplanes can pull off automation sur- vertical intentions of the auto- jected state of those various inputs to prises that aren’t even in the manual. mation need to be read from a the automation, and how they might So a pilot wouldn’t know it – however map display, which shows the all interact. They have to monitor au- diligently he or she studied the books. lateral, not the vertical. So the tomation activities and integrate all of vertical gets conjured into this this into a coherent assessment of the Back in my TAA, I had fi nally been able map with underspecifi ed sym- current and future behaviour of the au- to fi nd the right mode and leave 2,500 bols like moving green bananas tomation. feet. I announced as much on the fre- (I am not making that up) and quency. And I did it without touching dots and lines of various colours. Did you get all that? It is a tall order. A the houses lever! My pride and stub- It is like reading the runes to di- tall order indeed. And the way we train bornness were both confi rmed. “Uni- vine the future. pilots for automated fl ight decks may versity 185, I see you got it fi gured out still have some way to go before it re- now?” The controller sounded as really rises to the challenge. See a new lieved and proud as I did. Or perhaps pilot go into 737 training? He or she’s that is what I wanted him to sound like. Professor Sidney Dekker all smiles walking into the classroom “Affi rmative,” I said. “I have. Thank you is Professor and Director of the Key Centre for for the fi rst time, because eventually, for your patience, Centre. University Ethics, Law, Justice and Governance at Griﬃ th this is the fi rst jet they are going to 185.” “You’re welcome,” he answered. He University, Brisbane, Australia. Author of learn to drive. Four hours later, they told me to contact his colleague on the come out of the classroom, not hav- next frequency and wished me good best-selling books on human factors and safety, ing seen a simulator or airplane yet, luck with the automation on my jour- he has had experience as an airline pilot on the and they are lugging more books than ney northward. I wished it myself too. Boeing 737. they can carry. The smile is long gone. Because at some point, I was going to Their fl ight bag(s) are overfl owing with have to descend.

10 TALES OF OPERATIONAL SAFETY HI, TECH!

by Alberto iovino My six-year-old elder daughter is saving money to buy a PSP. If you experienced a sense of momentary disorientation about what a PSP is, that means there’s probably a good deal of added value for you in continuing to read on...

So, when my daughter fi rst expressed which I already suspected, but can also Following that, through a brief family to me her desire to possess one, I function as a camera, an audio/video council largely dominated by our one- knew what she was talking about, but recorder, a GPS receiver and much and-a-half-year-old younger daughter I immediately had to admit, to myself more, all of this at a quite aff ordable articulating her opinion (I only wish I had if not to her, that my knowledge of the price. been able to better understand what dreamed-of device was actually quite she meant), we as parents came to the superfi cial and defi nitely did not ex- inevitable conclusion. Although we dis- tend to its cost. After some investiga- approved of the idea of our girl becom- tion on the internet (which of course ing a video game addict at such a young provides evidence that I am not age, a denial would have soon bestowed completely out of date), I upon her the status of outcast from the discovered that it not circle of her fully-equipped friends and only allows you to school mates. Perceiving us as unwilling play video games, to say yes, but unable to say no, she bril- liantly resolved the situation by propos- ing to buy it on her own, thus off ering us the psychological alibi for a trade-off between our concerns for her wider mental development and the valuable lesson of achieving a goal through dedication and consistent eff ort. Moreover, we (and no doubt she too) recognised that we would likely be the main fi nancial contributors anyway. 

HindSight 16 Winter 2012 11 TALES OF OPERATIONAL SAFETY

This anecdote offers some analogies with our approach to modern technology. Generally speaking, we each form our own mental picture of a new environment or experience with reference to models we acquired or developed on first exposure to it. Though, by definition, the aim of technology is to change and manipulate the human environment, inevitably the level of technology you were exposed to whilst growing up assumes, to some extent, the character of normality. For my generation, and in my country, cell phones are a novelty that, over a very short time, have led to dramatic changes in many aspects of our ev- Through the subsequent Investigation frequency synthesizers, because these eryday life. For younger people, they by the Swiss BEA/BFU, it was estab- transmit very precisely on the nominal are an essential communication tool lished that the Airbus take-off clear- carrier frequency. However, this does and obsolescence in their present ance had been issued right after an cause a superimposed whistling tone form can already be anticipated from instruction to the ATR to line up; the (but it) is below the audible range of how they are often used as anything ATR flight crew had misheard the sub- human hearing.” but a telephone. Which closes the sequent clearance as being for them. loop on those who, coming from the Both crews read back the clearance How many everyday practices do we age of phone booths, still ask timid- almost at the same time, with the two rely upon, which come from our con- ly, and after a fifteen minute lecture communications overlapping. The rel- solidated background, and which by their customer adviser about the atively stronger signal received from may have become obsolete without wonders of the latest product, where the A340 was all that was heard in the this being appreciated? Clearly, keep- the dial is. tower, but some of the ATR read back ing pace with technological develop- was audible to the pilots of the third ments in one’s own working environ- Professionally, keeping oneself up aircraft, thus allowing them to appreci- ment is about what no longer applies to date is obviously a must. Training ate what was happening. as well as what is new. And this strong- is consistently dedicated to new sys- ly supports a contention that in our tems and functionalities, designed About the fact that the controllers did job, interaction with other domains to address both new ways to do the not get any indications of simultane- is so strong that a good awareness of same thing, and new things to do. ous transmission, something surpris- other stakeholder’s tools is paramount Some more basic subjects, by the ing, to me at least, can be found in the – that’s what this issue of HindSight is way, may simply be neglected. Final Report of the Investigation: about.

On June 18, 2010, an Airbus A340 and “Air traffic controllers questioned were In my days as an airline employee I had an ATR 42 began near simultaneous of the unanimous opinion that they the privilege of working with, for, on take-offs on two intersecting runways would recognise a multiple transmis- and inside the first Boeing 747s that at Zurich, one with but the other with- sion due to a superimposed whistling ever flew. By the way, quite in line with out an ATC clearance. The crew of a tone. This opinion is based on experi- a tradition which has more than once third aircraft alerted the control tower ence with older aircraft-side transmis- applied in time, whenever some sig- to an incorrect read back which had sion equipment, which in the event of nificant technological innovation has been inaudible to the controller, who dual transmission generally caused a been officially launched, the first 747 was then immediately able to call the superimposed whistling tone in the re- to operate on a scheduled flight was ATR to stop just in time for it not to ceiver in the audible frequency range. actually the second. I mean, it was the reach the runway intersection, whilst However, this is no longer the case with one standing from the bench on 22 the A340 performed its departure. modern transmitters equipped with January 1970, as the intended first op-

12 tion defi nitely refused to prove fi t for controllers. Besides any offi cial fl ow traffi c. “Fast”, however, has become a its illustrious role. On that occasion, of information in your organisation, more delicate subject. Pilots aiming to 352 passengers originally expecting to working impressions circulate quickly fl y continuous climbs or descents are depart on the evening of the 21st, ex- in an ops room, and you will soon sometimes reluctant to accept high perienced a night-time delay of more hear, if not personally experience, vertical or horizontal speeds, which than six hours at JFK airport; none of how the newcomer behaves, and will in the latter case may be procedurally them is reported to have given up, not be caught by surprise. Surprise is proscribed anyway. What several pilots thus providing another example of more likely when an aircraft type with will increasingly be looking for is less how humans stand up for their rights, which you are very familiar begins to uncertainty and more predictability especially if they have paid for a ticket be fl own in a way you have not previ- in both space and time (the so called to fl y to Europe and are hoping for a ously experienced. I am not referring 4D trajectories), with “management” story to tell to their friends and family. to isolated aberrations from the nor- progressively replacing any other key mal which may sometimes happen for word, including “control”. These 100 series 747s, which were a wide variety of reasons, but rather state-of-the art technology at that to changes in the way operations are So, understanding technological de- time, still required the presence of a conducted. This takes us a little fur- velopments in a rapidly changing fl ight engineer in the cockpit, like the ther from understanding how modern world requires fl exibility and a willing- 727, and used to carry fl ight attendants aircraft perform, towards obtaining a ness to adapt, which we all possess who looked like Hollywood stars. But clearer picture of how modern aircraft in diff ering degrees. We can gener- by the late eighties/early nineties, the are fl own. alise and say trivially, and with excep- tions, that younger people look better equipped, since they face less novelty How many everyday practices do we rely upon, in the short term and, probably also which come from our consolidated background, because they jumped on board when the average speed was already high. and which may have become obsolete without this Still, one day they will in all likelihood being appreciated? have to face the same feelings of inad- equacy their older colleagues sometimes experience today. As air naviga- same aircraft had aged and you could My ATC generation grew accustomed tion and ATC are gradually becoming see increasing maintenance time being to “the shorter, the faster, the better” something else, curiosity and an open required to keep them on line - main- but it is not like that any more. In times mind are the only way through. tenance involving a good deal of me- of economic crises, strong market chanics and, compared to today, rela- competition and high fuel prices, even In any case, once the re- tively little electronics. According to the the most solid airlines are very careful quired sum has been EUROCONTROL Aircraft Performance with costs, while many other operators raised, I am looking for- Database, they could typically climb at simply struggle not to run out of cash. ward to playing one or 1000 ft/min to 5000 feet whereas for And then there are new environmental two of those video games the most recent (and much heavier) concerns – and the associated poten- myself. version, the B747-8, the correspond- tial cost penalties. “Short” is therefore ing fi gure is 2500 ft/min and even the still largely appreciated, though more Airbus A380 is stated as off ering 1500 and more in a strategically organised ft/min to 5000 feet. So wide-bodied air- context like free-route airspace, rather craft on long haul no longer fl y as “fl at” than by means of frequent clearance after departure as their predecessors amendments that sometimes appear Alberto Iovino used to - although interestingly, they to challenge pilots adapting to the is currently head of ATS Operational Procedures only continue to do better after 5000 increasing capability of aircraft auto- Unit of ENAV Italy. Formerly an airline employee feet at slower airspeeds! mation. Pilots, as Michel Tremaud re- minds us elsewhere in this issue, aim for 8 years, he became an ATCO in 1997, This is the sort of change that is nor- to “stay ahead of the aircraft”, just like working as tower, approach and area controller. mally absorbed in a short time by controllers need to stay ahead of air

HindSight 16 Winter 2012 13 THE VIEW FROM ABOVE REVERSION – the other side of automation

by Captain ed Pooley Most of us recognise that the arrival of high levels of commercial aircraft automation and their major eff ects on the precision with which aircraft performance can be delivered has had a huge impact on the ATM world.

In conjunction with related improve- scope for clearance issue is also great- ments in the same direction in ATM, er. But poor pilot use of automation more aircraft navigational precision in a fast-moving aircraft can quickly has enabled increased effi ciency, fl ex- lead to problems exacerbated by the ibility and capacity all at the existing or expectation of usually more reliable an enhanced level of safety. For auto- outcomes that have allowed more mated aircraft, the likelihood that ac- aeroplanes to use the same skies. And ceptance of an en-route clearance anyway, the skies are full of a complex will be followed by delivery ex- mix of aircraft with a range of perfor- actly as accepted is greater mance capabilities even before you than in former times – add in the pilot factor! and the controller’s Of course, apart from such occasional misuse of automation, the everyday issue if it is functioning properly – and it is very reliable – is twofold. Firstly, how well do pilots understand its capabilities? and secondly, if it or the inputs on which it depends malfunction, how well do pilots cope with reversion to ‘less automation’?

Quite some years ago, but a long time after fl ight with auto pilots and auto throttles became routine even for approaches, almost all the simulator

14 time spent on training and checking pilots on their task competence was conducted without the use of the autopilot. The ‘excuse’ was that to allow Joe... does the emergency it to be used reduced the workload NAV kit work or should I call MAYDAY? which could be imposed upon pilots to see if they could ‘survive’ under high pressure. Such pressure was equated essential process for situations where on acquired and retained knowledge at that time with the pressure that no specifi c procedural response exists, which will only very rarely be needed. might arise if unspecifi ed abnormali- often demands rapid recall of acquired ties arose. Eventually, as this early level and retained technical knowledge, These ‘reversions’ may be internal to of automation moved into the era of both generic to all aircraft fl ight and the ATM system or a consequence of the Flight Management System, di- specifi c to the aircraft type involved. changes to the automation status of rectives changed to a requirement to Such a background goes well beyond an aircraft being handled. Has ATM use the autopilot most of the time. how to get the best out of the SMS and training risen to this challenge? I sus- However, since the required minimum how to optimise aircraft performance pect that, just as in pilot training, in simulator time stayed the same, op- in ‘normal’ operations. But how wide- the areas of background knowledge it erating the aircraft with autopilot out spread is this ‘competency’ nowadays? has not yet caught up with the rapid became something to do in the air- arrival of reliable automation in both craft on a nice day line fl ying. Back in Could there be a parallel in ATM as ATM systems and on the fl ight deck. the simulator, with the exception of a systems are increasingly automated If I am right, it is time to ensure that few key (memorised) emergency task to make sure that ATM performance expensive recurrency simulator time competencies1, the focus in the era of continues to match modern aircraft for controllers is preceded by class- increasingly complex (but also increas- performance? I think so. Performance room preparation for infrequent rever- ingly reliable) automation moved to a of any system which depends on high sions of all sorts which goes beyond combination of the everyday and the levels of automation to deliver effi - ‘learned responses’ for the expected anticipated departures from it. Be- ciency, fl exibility and capacity with and presents ‘unpredictable’ or ‘unex- cause there were now so many SOPs no reduction in safety also demands pected’ scenarios. For such scenarios, for loss of automation scenarios, it was an ability to cope with reversion to there will not be just one particular tacitly assumed that there would be a lower level of system performance. and prescribed correct response but one for most situations provided that Crucially, just as for pilots, this in- several equally acceptable ones. Of (when using a QRH in book form in pre cludes both reversion to expected or course, such background training for ECAM/EICAS days) you could correctly anticipated conditions, which can be the unexpected will undoubtedly also identify it! addressed by prescribed responses provide a deeper understanding of and the infrequent, perhaps very performance issues in the ev- But this understandable focus on infrequent, unexpected and unan- eryday world. mitigating the ‘regular’ causes of ac- ticipated conditions. Here again, the cidents led to far less attention being ability to respond eff ectively is, as paid to the wide range of infrequently for pilots, is likely to be dependent encountered (for any particular pilot) abnormal events, for which a procedural response was (entirely under- Captain Ed Pooley is an experienced airline pilot who for standably) not specifi ed or only par- many years also held the post of Head of Safety for a large short haul airline operation. tially specifi ed. What seems to have He now works as an independent air safety adviser for a range of clients and is currently been overlooked is that what used to acting as Validation Manager for SKYbrary. be called ‘thinking on your feet’, an

1- Such as engine failure on take-oﬀ , emergency descents and responses to activation HindSight 16 Winter 2012 of the Stall Protection System, the TAWS and the TCAS 15 FRONT LINE REPORTS Modern times

by eileen Senger In the last 12 months, two important new aircraft types have begun appearing in Maastricht airspace and elsewhere in the world: the Boeing 747-8 and the Boeing 787. ...

point. The point is that all of us have Now another issue! In these ‘fi fth gen- to gain our own experience with new eration’ aircraft, the pilots’ task has Eileen Senger aircraft types from scratch. The -8? It shifted from aviator to input operator. is an Air Traﬃ c Controller at EUROCONTROL’s is a 747 so we can expect it to climb It is the computer which is really fl y- Upper Area Control Centre in Maastricht. well and fl y fast! And we guessed right. ing the aircraft. When it comes to au- She works in the Hannover Sectors which But the 787? No clue what to expect. tomatic fl ying of a TCAS RA manoeu- cover north-western Germany and is an OJTI. The fi rst time I had it on the frequency vre, I think this is a real improvement. it climbed like a rocket and I was very Pilots sometimes overdid what TCAS pleased, the next time it climbed like it told them to do, not on purpose but came from the aircraft factory from the in the context of the surprise and the I found out by coincidence. Before, other side of the pond… So how does urgency of the situation, occasion- I had read in newspaper and fl ight it perform? I still don’t know. ally even making the potential confl ict magazine articles that both were worse. A computer has no emotions about to be introduced into service. I Is all that guesswork really necessary? so it fl ies the manoeuvre exactly as it is had also read about the usual devel- The data is available and it would be calculated. In our busy airspace, it uses opment problems and setbacks and such a nice service to air traffi c control- minimum airspace for maximum ef- about the expected performance. lers to provide them with a quick data fect, thus not involving other aircraft. And then at once it is there, in my sheet giving them a rough overview On the other hand this will lead to airspace, a square symbol looking of what rates of climb and speeds to fewer and fewer TCAS RA manoeuvres like all other aircraft. And I have no expect. clue how it will behave and perform on its climb out from Frankfurt. I had received no briefi ng, no information sheet with indicative performance data, nothing.

When the Airbus 380 entered into service a few years ago there was more activity beforehand. There were concerns about the wake turbulence. So we received briefi ng sheets about the introduction of this type of aircraft but all they said was: Do not worry about it in upper airspace, it is only an issue for Tower and Approach. Again no performance data, speeds, climb rates etc. So what do you do? You expect the worst. Take an A340-300 on a hot summer day and downgrade from there. I must admit that so far the A380 is doing better than that but that is not the

16 And now all you‛ll have to do is to use the information from the catalogue with the simple formula from the left screen and you‛ll get accurate performance data for that aircraft... being fl own manually – what if the Au- In the section “Climb? YES WE CAN!” last in sequence. The general opinion topilot is not available? And then there he sheds light on exactly those situa- amongst air traffi c controllers is now is a TCAS RA? With the ever increasing tions where “the pilots seem as least as that it is impossible to work properly automation one should not underes- surprised as the controllers to see the with that airline anymore. Why make an timate the impact if that automation aircraft reduce its rate. It seems that eff ort to get them in quicker when they fails. The training challenge increases predicting or knowing what the air- are fl ying so slow anyway! Meanwhile, it and more visits to the simulator for ad- craft (i.e. the computers) will decide is seems other airlines have noticed this ditional and diff erent training become possible and what is not has become fuel-saving strategy and have started more and more important. more diffi cult over the years.” Perhaps crawling as well. I am curious to fi nd out this is not all that surprising given the whether this is merely a short-term phe- When it comes to estimating what per- greater capability of automation. nomenon or whether this is the start of formance an aircraft will be able to de- a change of approach. All we air traffi c liver, it has become a guesswork as well And lastly, we are entering a period controllers can do is watch, learn and – for the pilots! “Are you able to climb not only of change but of uncertainty. adapt our way of working. with 1500 ft/min until passing FL300?” I One of the newest phenomena result- can only recommend a re-read of Phil- ing from the economic crisis in Europe ip Marien´s article “The “OTHER” level which has put airlines under even editoR’S note busts” from Hindsight 101. more pressure than usual to cut costs is aircraft fl ying at very low speeds dur- this problem of the wide range ing cruise. Of course this is a result of of speeds which controllers now company policy rather than the per- see an aircraft type being fl own at formance of modern aircraft types, is even more complicated where However, it is mainly due to modern aircraft operators take a compre- computers and GPS, which moni- hensive view of costs. A signiﬁ cant tor every second of a fl ight in terms infl uence on whether fuel-saving of fuel economy vs. time fl ying, that by slow fl ying is conducive to over- this is now exercised as rigorously as all cost-saving is dependent on the it is by one German airline. Pilots are extent to which scheduled mainte- given a cost envelope in which their nance costs are predicated on fl ight fl ight has to operate. When I was train- hours rather than fl ight cycles and ing I learnt by heart: An A320 cruises on whether aircraft are owned or Mach .78, when it is in a hurry up leased and, if leased, on the applica- to M.80. Today, I see them fl ying at ble payment terms. Since these fac- speeds between M.62 – M.64. And tors can and do vary both between when they are behind schedule M.80. and within the aircraft type fl eets So I never know what to expect. Se- of diﬀ erent operators, there is little quencing has become an adventure. prospect of simple clarity in speeds Some pilots are happy to deviate from anytime soon…unless speed decla- their slow speed “on ATC instruction”, rations on fl ight plans are to be “en- others answer, when asked to speed forced”! up, that they do not mind becoming

HindSight 16 Winter 2012 1- See http://www.skybrary.aero/bookshelf/content/bookDetails.php?bookId=965 17 FRONT LINE REPORTS Modern aircraft performance

✶ ✶ ✶ is not alone! ✶ ✶ ✶ ✶ ✶ ✶ ✶ ✶ ✶ ✶ ✶ ✶ ✶ ✶ ✶ ✶ ✶ ✶ ✶ ✶ ✶ ✶ ✶ ✶ ✶ ✶ ✶ ✶ ✶ ✶ ✶ ✶ ✶ ✶ ✶ ✶ ✶ ✶ ✶ ✶ ✶ ✶ ✶ ✶

✶ ✶ by Maciej Szczukowski I think that we have also reached a time when modern aircraft perfor- Aviation is great – from shiny airplanes in mance and present-day ATC performance are not always in step. When 70-year-old black and white pictures, did they divide? It is a topic for a with smiling pilots and their white scarves and whole book, I guess, but let’s follow up one or two clues. impeccably trimmed moustaches, to today’s complex cockpits and hundreds of fl ashing I admire the variety of airplanes flying on our skies nowadays. I like to lights, with smiling pilots and their white shirts learn about them, see new aircraft types or new airlines landing on the and impeccably tailored uniforms. The huge runway of the airport I work at. I like changes in aviation fashion have been mirrored to know that, although still limited by the laws of nature, engineers are by the evolution of aircraft performance. What able to set new records for maximum altitude, speed, minimum fuel con- once used to be only a trial in Kitty Hawk is now sumption. a huge industry and business with millions of These new high-performance air- people involved. What used to be a simple craft are real pieces of engineering “refuel – clear prop – depart – land” scheme is art, but the airspace is one and they share it with others. These same air- now one of the most complex fi elds not only craft often fly in the same airspace, are served by one and the same air in engineering science but also in IT, manage- traffic management system and use ment, psychology and medicine. the same airports as all the others that are able to “cheat” the law of gravity. Vintage aircraft, business aviation, military, recreational aviation, gliders, paragliders, UAVs, rock- ets, birds, Santa Claus and occasional witches…

18 ✶

✶ ✶ ✶

✶ ✶ ✶ ✶ ✶

tricht UAC. Like- ly level busts can be Do you know the expres- captured by the Air Traffi c sion “the chain is only as strong Controllers as soon as the pilot se- as its weakest link”? This implies that lects the wrong altitude and the same any chain, system, organisation, or ern aircraft, like the Airbus 380 and the is true for identifying mis-set radar ATC, performance is going to be de- Boeing 787, to “war birds” from WWII. headings and mis-set speeds. Cool! termined by the least reliable element And like the proverb above, the reli- in it, i.e. by the least reliable airspace ability of the most sophisticated ele- However, these Mode S equipped user. ments in the chain can be compro- aircraft sometimes share the same mised by the less reliable ones. Soairspace our ATC with network aircraft systemwith no is highlytran- This expression may be an interesting protected,sponder at and all! we Or have an theseinoperative sophis- proverb, but mathematically speak- Let me give you some examples. ticatedone, or modern one jetstransmitting with high incor-perfor- ing it is not true. Systems and organ- mance,rect information. but sometimes Such it takesan aircraftvery lit- isations are designed to have redun- Some advanced Air Navigation Service tlecan for be this invisible airspace or userappear community with the to dancies, back-ups and protections for Providers have recently been intro- lookwrong more position like a chainfor anthan ACC a network. using their less reliable elements. The chain ducing functionalities in their systems Thisonly makessecondary it more surveillance.vulnerable to failure,STCA becomes more like a network and a based on Mode S technology. Mode whichsystems is whycannot I would capture argue the that confl mod- ict failure of one element makes a hole S is a surveillance technology that, ernaff ectedaircraft by performancethese problems and and present- TCAS in the network but does not break it together with the usual surveillance daytoo ATCwill performancebe totally ineff are ective not always in pre- in completely. position, provides much more infor- step.venting collisions. You may say that mation that can be used for all sorts of this will not happen or that it is “ex- The ATC system is a very well pro- not only cool but also useful applica- Isn’ttremely it about improbable”. time they Well, were it reunited? has hap- tected system, layer after layer…but tions. These include aircraft identity, pened more than once, as you can see sometimes, for some situations, it is altitude, speed, heading, vertical rates from the article that has been already more like the chain made by the vary- and downlinked TCAS resolution advi- published in HindSight2.  ing performance of the aircraft this sories. The benefi ts may be enormous 1 system serves. The links in this chain – take a look on SKYbrary at the video 1- http://www.skybrary.aero/index.php/Mode_S range from very sophisticated mod- for Mode S implementation in Maas- 2- http://www.skybrary.aero/bookshelf/books/1418.pdf

HindSight 16 Winter 2012 19 ???? and controllers are sometimes confronted with rapid-onset, dynamically developing situation

In this case My second example highlights the Take another story5 – the incident on the single opportunity for teamwork and cross- 12 January 2011, when the single pilot transponder monitoring on a multi crew fl ight deck. of privately operated Socata TBM850, fi tted was faulty You may have in one and the same air- with some 12 hours of fl ight experience but still operating space aircraft with two or more fl ight in the 28 days before the incident, lost and the altitude in- crew and also small single-pilot air- radio contact on a non-precision ap- formation was incorrect. craft. I know it is simplistic to look only proach, continued the approach with- SSR Mode C indicated FL at the number alone, as there should out landing clearance and landed over 270 but the aircraft was actual- be procedures and other means to en- the top of DHC8-400, which had lined ly fl ying at FL 290. Surveillance was sure an equivalent level of safety no up ready for take off . The PPL-licensed misled, neither STCA nor TCAS was matter how many pilots you have. But Socata pilot had a heavy workload in triggered and the collision was averted do not tell me that the challenges fac- trying to perform an NDB DME ap- only by a ‘last-minute’ see and avoid. ing two aircraft fl ying the same non- proach for the fi rst time in four years You can see more on this in SKYbrary, precision approach are necessarily the and failed to stabilise the approach, which also has a copy of the Investi- same. One aircraft may be operated mistuned the radio, carried on without gation carried out by the French BEA by a highly trained, professional crew landing clearance and fi nally failed to (Bureau d’Enquêtes et d’Analyses pour with regular exposure to non-preci- see an aircraft on the landing runway la sécurité de l’aviation civile).3 sion approaches. The other could also threshold, missing it by pure chance. be legally acceptable in the same air- What we have now is a chain aircraft space, but fl own by a private pilot who And in general how do you think a with sophisticated surveillance per- has not fl own such a non-precision ap- single pilot, fl ying manually copes with formances potentially in the same proach anywhere for several years. copying complex ATC clearances? airspace with an aircraft with critically aff ected surveillance performance or The story of the Qantas A380 uncon- So our ATC network system is highly no surveillance performance at all (at tained engine failure on 4 November protected, and we have these sophis- least secondary surveillance). 2010 is a great example of teamwork ticated modern jets with high perfor- Think about our proverb in which the Captain had the good formance, but sometimes it takes very lit- now! tune to share the fl ight deck not only tle for this airspace user community to with the usual extra co-pilot for the look more like a chain than a network. planned long fl ight but also two more This makes it more vulnerable to failure, pilots – a Check Captain and a Super- which is why I would argue that mod- visor Check Captain. This team of fi ve ern aircraft performance and present- made this story a success with a care- day ATC performance are not always in ful division of tasks4. step.

Maciej Isn’t it about time they werereunited? Szczukowski has been an Air Traﬃ c Controller, for over 10 years, at Warsaw Okecie Airport, Warsaw, 3- http://www.skybrary.aero/index.php/PC12_/A318,_en- route_north_east_of_Toulouse_France,_2010_(LOS_AW_HF) 4- You can see more about this and access the oﬃ cial investigation report at Poland. He also holds a PPL. http://www.skybrary.aero/index.php/A388,_en-route_Batam_Island_Indonesia,_2010_(AW) 5- http://www.skybrary.aero/index.php/TBM8,_Birmingham_UK,_2011_(AGC_LOS_HF)

20 121.5 – SAFETY ALERTS

SAFETY REMINDER MESSAGE IcAO aircraft operator and radiotelephony designators and abbreviation of ‘Type c’ call signs By richard “Sid” Lawrence Synopsis Released on 11 may 2012 eURocontRoL has received a growing number of safety reports from air navigation service providers and aircraft operators related to the use of icAo aircraft operators’ three-letter and radiotelephony (R/t) designators and the abbreviation of ‘type c’ R/t call signs – the latter usually associated with alphanumeric bi-grams (e.g. BA) at the end of the call sign.

Over the past few months the EUROCONTROL Safety Alert service IcAO Provisions “has been approached by a number of n iCAo doc 8585, designators for Aircraft operating Agencies, Aeronautical Author- stakeholders requesting the ities and Services describes the use of, and lists, aircraft operators’ ICAO three-letter publication of a safety alert. In the designators (e.g. HJT) and R/T designators (e.g. ‘HIGHJET’). pages that follow, I will describe three n iCAo Annex 10 Vol li, Chapter 5 of the alerts, covering a variety of top- § 5.2.1.7.2.1.1 states that “An aircraft radiotelephony call sign shall be one of the follow- ics that I hope will spark your interest. ing types… Type c) - the telephony designator of the aircraft operating agency, followed by the fl ight identifi cation”. As previously, my intention is to try and bring new information to the table. § 5.2.1.7.2.2 states that there is no abbreviated form for ‘Type C’ call signs. The aim is to feature more in the way of feedback, responses, comment and analysis to get the most from each Analysis alert. The use of standard R/T phraseology is a cornerstone of safe operations. The correct use of aircraft call signs is an integral part of the pilot/controller communication loop. If you would like to know more about Most commercial operations use ‘Type C’ call signs. Consequently, pilots expect to be ad- the EUROCONTROL safety alert service, dressed by their assigned ICAO R/T designator rather than the company three-letter desig-

register as a subscriber, submit a sug- nator. Routine use of the R/T designator is, therefore, more likely to trigger an accurate and gestion or have a subject that you wish timely response from fl ight crews; the converse may also apply.

to consider then please contact me at Flight crews should be aware of their own R/T designator but controllers are confronted by [email protected]. “ a multitude of designators. It is acknowledged that correlating/memorising the ICAO three- letter designators with their R/T designators is challenging, in particular as there are constant This time, all three featured alerts are changes. In recognition, many ANSPs provide controllers with support information tools/ safety reminder messages. systems to access the designators listed in ICAO Doc 8585. Using a non-standard call sign designator and/or inappropriate abbreviation of the call sign fl ight identifi er can increase the risk of fl ight crews and controllers missing a call or taking/making an incorrect instruction. Moreover, repeat Alternatively, register your interest through SKYbrary: broadcasts by controllers and/or fl ight crews increases http://www.skybrary.aero/index.php/Portal:EUROCONTROL_Safety_Alerts workload on the ground and in the air.  where you can access the Alerts featured here and previous Alerts.

HindSight 16 Winter 2012 21 121.5 – SAFETY ALERTS

SAFETY REMINDER MESSAGE (cont’d)

SoMe eXAMPLeS oF rePorTed iCAo deSiGnATor USAGe

Aircraft iCAo aircraft iCAo r/T reported controller/ Comments operator operator designator designator pilot use

Thomas Cook TCX KESTREL TOMSON Controller uses another R/T designator.

Titan Airways AWC ZAP ALPHA Controllers sometimes do not WHISKEY know the ICAO R/T designator CHARLIE and instead spell out the three-letter ICAO Aircraft Operator designator (AWC).

Regional Europe RAE REGIONAL EUROPE REGIONAL Pilot shortens the R/T designator which is then repeated by controllers. In this case ‘REGIONAL’ belongs to another airline.

Volga Dnepr VDA VOLGA DNEPR VICTOR Pilot spells out the three-letter DELTA ICAO Aircraft Operator ALPHA designator (VDA) instead of using the assigned ICAO R/T designator.

note: (1) Some aircraft operators do not have an assigned ICAO R/T designator but do have a three letter aircraft operator designator. These airlines often, but not always, use their company name as the R/T designator.

ABBreViATion oF ‘TYPe C’ CALL SiGnS – FinAL TWo-LeTTer Bi-GrAMS As per ICAO Annex 10, ‘Type C’ call sign ‘HIGHJET 12BA’ should be read out as ‘HIGHJET, WUN TOO, BRAVO ALFA’ and must not be abbreviated to, for example, ‘HIGHJET BRAVO ALFA’.

However, in France, in accordance with a national decree, controllers routinely abbreviate ‘Type C’ call signs that end with a two-letter bi-gram, but they will not abbreviate other formats such as ‘HIGHJET 3456’ or ‘HIGHJET 345B’.

The introduction of alphanumeric call signs and the use of fi nal two-letter bi-grams is increasing as more aircraft operators use this format of call sign as part of their call sign similarity deconfl iction strategy.

22 Your attention is required Feedback

Air navigation service providers and aircraft operators were n The Titan Airways/ZAP/AWC combination is still causing invited to: problems even after the SRM has been published. ICAO doc 8585 says, “a three-letter designator should reflect n share their experiences, as well as the practices and tech- to the maximum extent practicable, the name of the air- niques they use in their operations to manage the is- craft operating agency or its telephony designator…” In sues described (e.g. sector/frequency management and this case there is no obvious link between the three-letter monitoring processes, FDP/HMI track labelling, availabil- designator, the telephony designator and the airline’s ity of R/T designators to controllers, use of alphanumeric name and this makes it more difficult for controllers to call signs etc); make and remember a mental association between the three elements. The company has considered changing n remind both controllers and flight crews about the ICAO either AWC and/or ZAP to something that is more readily SARPs related to the use of call signs in general, and in recognisable with Titan, e.g. TTA as the 3-letter designa- particular, that: tor. However, it considers that there could be more risk attached to a change than the current position and so - ‘Type C’ call signs consist of the assigned ICAO aircraft wishes to maintain the status quo. EUROCONTROL will operator R/T designator (rather than the three-letter continue to monitor the situation. designator) followed by the flight identification; n To help mitigate the issue described above, Titan Airways - ‘Type C’ call signs should not be abbreviated - whilst includes the call sign “ZAP” in the Field 18 remarks on noting the France exception for call signs ending with the flight plan. However, some ATC systems cannot read two-letter bi-grams; Field 18 data and in those that can it cannot always be displayed to, or retrieved by, controllers easily. Conse- n consider applying, in accordance with ICAO Doc 8585, quently, this tactic of raising ATCO awareness is largely an ICAO R/T designator for airlines that do not already ineffective. have one - aircraft operators to action as applicable.

Further reference

n SKYbrary - Air Ground Communication Briefing Note No2 - http://ww.skybrary.aero/bookshelf/books/110.pdf n SKYbrary - Safety Reminder Message: Abbreviation and Misinterpretation of ‘Type C’ R/T Call Signs - 5 August 2010. http://www.skybrary.aero/index.php/Abbreviation_and_Misinterpretation_of_%27Type_c%27_Call_Signs n IATA Phraseology Survey 2011. http://www.skybrary.aero/index.php/Pilots_and_Air_Traffic_Controllers_Phraseology_Study n ICAO Annex 10, Volume II Chapter 5. n ICAO Doc 8585, Designators for Aircraft Operating Agencies, Aeronautical Authorities and Services. n ICAO Doc 9432, Manual of Radiotelephony.

HindSight 16 Winter 2012 23 121.5 - SAFETY ALERTS

SAFETY REMINDER MESSAGE Misuse of international aeronautical emergency frequency 121.5 MHz

published on 19 June 2012 Synopsis the eURocontRoL Agency has been notiﬁ ed on numerous occasions about the misuse of the international aeronautical emergency frequency, 121.5mHz, most recently involving inappropriate ‘chat’ related to the ongoing eURo 2012 football championship. Your attention is required n Aircraft operators were invited to remind their fl ight crews IcAO Provisions about the correct use of the international aeronautical emergency frequency, 121.5MHz, according to ICAO/na- ICAO Annex 10, Volume V, § 4.1.3.1.1 states that frequency tional requirements and company policy. 121.5 MHz “shall be used only for genuine emergency pur- n Air navigation service providers and state aviation authori- poses” broadly covering the following activities: ties are invited to note the subject and share their experience with similar cases. n the handling of a emergency situations; n air-ground communication with aircraft with airborne equipment failure; EUROcONTROL comment n search and rescue operations and the operation of emergency locator transmitters (ELTs); and The misuse of the 121.5MHz has been raised on a number of n air policing/interception action. previous occasions. Although the message was primarily aimed at the pilot community, since publication EUROCONTROL has note: Some states have fi led diff erences to ICAO SARPs re- also been made aware of inappropriate controller ‘chatter’ on lated to the use of 121.5 MHz – for instance, in the UK it can the emergency frequency. It is clear that, whoever the culprits also be used for practice PAN calls to ensure pilot familiarity may be, there is no place for inappropriate ‘chat’ on 121.5MHz. with the process. Such diff erences are detailed in national The international aeronautical emergency frequency is fi rst and AIPs. foremost a facility to serve fl ight safety needs and its use must be reserved for its intended purposes. Analysis Inappropriate ‘chat’ on 121.5 MHz could interfere with its Further reading legitimate use and should be avoided in order to maintain n ICAO Annex 10, Vol V. the integrity of the frequency for the purposes for which it n SKYbrary Safety Alert: Request for Support Message, is intended. “Guarding 121.5”, 12 March 2007. http://www.skybrary.aero/index.php/Guarding_121.5_MHz

24 121.5 - SAFETY ALERTS

published on 18 July 2012 Synopsis Lessons learned A european Atm service provider reported an incident following a planned intervention on the electrical supply system. during the switch- back to mains power a signifi cant system failure from a power occurred. the eff ects of the event included the unavailability of various operational systems for 01:37 hours. the event was caused by aging outage incident StS (Static transfer Switch) components of the electrical supply system. Analysis n The incident occurred during planned maintenance and reno- n Permanent solution: the degraded STS components will be vation works on the high tension power network. At regular replaced. In addition, the architecture (involving power sup- intervals, interchanges between the public power grid and the ply to the entire IT platform via two redundant STS compo- emergency power generators were performed. During the last nents) and settings of the entire power supply system will be interchange, a power cut lasting some 100 minutes occurred. reviewed and, if found necessary, improved. n The eff ects of this event lasted for 01:37 hours, during which n Decisions in crisis or system degradation events should be time various operational systems were either unavailable or un- practiced to ensure quick reaction in such critical outages. able to distribute messages. The prescribed procedures should be made as simple as pos- n A detailed analysis of power network measurements, equip- sible. ment logs and IT-system logs, followed by factory tests of power n All maintenance interventions on power supply systems system components made it possible to determine the follow- should be preceded by a formal safety analysis of potential ing: signifi cant operational eff ects. - A succession of fl uctuations in electrical frequency synchro- n Crisis management checklists should be developed to pro- nisation between the power system components led to a mote consistent and rapid decision making. slight change from the normal frequency of the electrical power signal. - Factory tests of STS components showed that this fl uctua- Your attention is required tion exceeded the tolerances in these components because n ATM service providers were invited to note the subject and of their variation from design tolerance attributed to ageing share their experience with similar cases. (10 years old). n As a conclusion, the root cause of the unintended power cut was determined to have been STS components erroneously re- cording a degradation in the power quality (a frequency toler- EUROcONTROL comment ance overshoot) and leading to a very short cut in power supply n It is essential that ANSPs can react promptly and effi ciently to operational systems. to emergency situations, degraded modes of operation and longer-term contingencies. Scalable emergency/crisis response regimes should ensure a coordinated approach to Reported actions and lessons learned most eventualities. A key component to successful and safe outcomes is eff ective communication between all aff ected n Short-term solution: the automatic transfer of power source be- parties. tween the STS components was temporarily disabled to avoid n In this context, keeping the Network Manager ‘in the loop’ recurrence of the incident if similar synchronisation fl uctuations is important so that collective decisions can be made to not should appear. only manage the direct local impact but also the side-eff ects that might affl ict other parts of the ATM Network. Further reading n SKYbrary - Safety Reminder Message: ANSP Preparation for Emergency, Degraded modes of Operation and Unusual Situations. http://www.skybrary.aero/index.php/AnSP_Preparation_for_emergency,_degraded_modes_of_operation_and_Unusual_Situations n SKYbrary – Category: ANS Contingency Planning. http://www.skybrary.aero/index.php/Category:AnS_Contingency_Planning

HindSight 16 Winter 2012 25 CASE STUDY case Study - The plane spotter By Bengt collin, eURocontRoL

The Cargo Airport Tower She liked her new job, although the two weeks earlier on the top of a hill working conditions were diff erent. just south of the airport. Sid started It was a grey misty and cold Novem- They could be simply summarised in eating one of his home-made sand- ber day. The wind was from the north, one sentence: “No we don’t have any wiches, herring with orange marma- the clouds almost touched the pine money for improvements”. Everybody lade, topped with vanilla yogurt. Brent trees, in a soft woolly way. Gina was was very pleasantly surprised when checked his camera. Sid switched on sitting by herself in the old, very old, work began to replace the main air- his air band radio; it was their regular tower. It needed updated equipment; side electrical system. Big machines habit to carefully monitor the Tower the whole airport needed an over- were digging up the old system, in- frequency. How could they otherwise haul. stalled by the air force some fi fty years know what aircraft were arriving or ago. The ILS went out of service. departing? “Fascinating, how do these It was the anniversary of her move to magical controllers do it?”, Sid com- the cargo airport, three years to the The Plane Spotters mented as he slowly took a sip from day. Following her graduation from his white tea. “That was a strange call the ATC Academy some eight years As every Saturday, Brent took the com- sign” he suddenly said to Brent. One back, she had worked at one of the muter train northbound. Most week- minute later an enormous aircraft airports in the very north and had ends he dedicated to his hobby. This passed overhead. like an albatross hated it! She hated the cold blistering was not just his hobby, this was his growing and growing in size. It almost weather, hated the snow, almost start- HOBBY, plane spotting. His friend Sid, entirely fi lled the sky above them. “An ed hating herself for the life she lived. An-124 cargo plane, got some nice photos” Brent calmly replied.

The International Tower For the fourth time that day the supervisor changed the runway confi guration. Why should we be off ering departing and landing aircraft a tail wind, one of the controllers asked. It was a fair question, at the Academy one of the basic things he learned was that aircraft should ideally land and take-off into a head wind. However, his time at the Academy had been well before the dressed in a warm local politicians, while enjoying the green anorak matched by an orange benefi ts of the international airport, Being cap, joined him at the main central began arguing that aircraft should not desperate to move, she station. Sid was almost as eager as fl y over their neighbourhood. They informed everybody she knew of her Brent to visit the international air- considered that noise was the main situation. One day Tony called. A pri- port. After thirty fi ve minutes on the local environmental issue, easy to vate company had won the contract train, they changed to a local bus and understand for everybody. So conse- to run a cargo airport about an hour fi nally arrived at their destination fi f- quently, the political compromise was south of the capital, they were looking teen minutes later. They returned to to spread the problem of noise equal- for controllers. the excellent location they had found ly, it was as simple as that.

26 You will get some diversions from the aircraft with exactly the agreed dis- saw the next inbound on the radar south, the supervisor informed the tance in between; Lucia was always screen, it was only two miles out, why runway controller. The ILS has been professional. An An-124 called, the did it not contact him? Unusually high out for almost an hour at the cargo fi rst time I’d talked to the crew of one speed too, he thought as he cleared airport. of these and I’d never seen one before. the departure for take-off . “Tower, be The Macchiato coff ee was excellent. advised that the right wing of the just- The tail wind component on the run- landed aircraft is still out over the runway was almost ten knots. Someone The Approach Control way”. The aircraft remained stationary decided, he did not know when, that at the threshold. The comment made up to ten knots tailwind was perfectly The business jet was still well above him confused, the landed aircraft was alright. the glide slope; Lucia did not like to clearly off the runway according to the break him off though. Something told HMI display. Then seconds later came The Approach Control her that their fuel endurance was not “he has vacated now” and he could that great and to follow a diversion hear the increased sound from the en- Three diverting aircraft joined the with a break from the straight-in ap- gines as the aircraft started rolling. The standard inbound fl ow. It was a mix- proach would extend their airborne business jet was on a half mile fi nal. ture of diff erent types and sizes, an time by another ten minutes, possi- An-124 from an airline Lucia had nev- bly fi fteen. But she did not know for The Plane Spotter er heard of before was followed by a sure of course. Lucia asked the pilots business jet with a foreign registration to reduce speed as much as possible, “It’s started snowing mate, let’s go and fi nally a low-cost carrier plane ten it did not help much. When she told home”, Brent said to Sid as a business miles behind. Plus all the other sched- the business jet to contact the tower, jet passed over their heads. “Emmer- uled traffi c. The wind situation was a the distance to the preceding aircraft dale on television in two hours. It’s bit strange; she knew how to handle was just on the agreed minimum and looking like a good bet if the weather this though, she went on doing her job she knew it would be less moments carries on like this”. as always. Lucia vectored for runway later. The speed was still high too, 36, today being used in mixed mode but the tower controller could handle At the same time, at a cargo airport to for both take-off and landings. The this couldn’t he? “Say again tower fre- the south, the ILS returned to opera- An-124 descended very quickly then quency”, she repeated the frequency; tional. started to slow down earlier than most you could tell by the voice of the pilot aircraft normally did. She understood that their workload was high. “Thanks”, why - the tail wind aloft was stronger the business jet had left the approach so the ground speed was higher than controller. usual. The business jet behind had a much higher altitude and a much The International Tower Bengt Collin higher ground speed. Their fl ight crew works at EUROCONTROL asked for the ILS frequency, fi rst time The An-124 landed and the next de- HQ as an Senior Expert visiting this airport? parture, a Boeing, lined up. “Vacate involved in operational fi rst right”; it was a rapid exit leading ATC safety activities. The International Tower to a parallel taxiway. He could not see the aircraft now, it had started snow- The line of departing aircraft at the ing and the visibility was reduced. He Bengt has a long background as Tower and holding point was growing, not that checked the A-SMGCS display, the Approach controller at Stockholm-Arlanda it bothered him, it never did. The ap- Multilat symbol was off the runway Airport, Sweden proach control delivered the inbound and the SMR reply confi rmed this. He

HindSight 16 Winter 2012 27 CASE STUDY case Study comment 1 by captain Ed Pooley

Not an unfamiliar situation! A potential for confl ict between a late go-around and the departing aircraft or a very late landing clearance, either of which will have the added excitement for all parties of restricted visibility.

approach. Of course they know that performing outside of their comfort if they do so, delay for their landing zone and – I surmise without the ben- Captain Ed Pooley may follow. If that means their fi nal efi t of proper guidance – probably is an experienced airline pilot reserve fuel may be eaten into then additional training. Of course, I know who for many years also held the post of Head increasing numbers of operators now the airport loves to get extra landing require the declaration of a PAN. And fees for zero marginal cost – it prob- of Safety for a large short haul airline operation. if it becomes obvious that some of ably adds to the performance bonus He now works as an independent air safety it will certainly be used, a MAYDAY. of the top team! adviser for a range of clients and is currently Less prescriptive operators leave acting as Validation Manager for SKYbrary. equivalent action to the aircraft com- mander’s discretion. Either way, there A reCoMMendATion is no case for the controller not to adjust their normal way of working to a A full review should be under- We get a tailwind runway with a lack busier situation in order to perform as taken of the robustness of ATM of positive controlling at both po- reliably for Tower as usual. procedures to the range of traffi c sitions. I will concentrate on what loading which may occur, howev- I think of the controlling style and And then the Tower. I fi nd it surprising er infrequently, and of the range leave the question about operating that the controller was permitted to of accepted aircraft types which runways right up to the same tailwind issue low-visibility take-off clearance may go with it, preferably by a component limit as applies to most based on automation that doesn’t suitably qualifi ed and indepen- aircraft types for another day. deliver for the whole range of aircraft dent outsider. types which the airport accepts. And Let’s look at Approach fi rst. Is there once he has been told that the run- any sign of delivering the usual ‘hand way is still not clear and knows the off ’ to Tower? I don’t think so. Radar next landing aircraft is close in and surveillance allows approach con- fast why on earth is the take-off clear- trollers to target closer spacing, but ance not immediately cancelled and if they do that, then positive control- the approaching aircraft told to go ling in required. Speed control for around? sure. Not accepting an aircraft into the landing sequence until it is at an But I don’t see a problem with indi- altitude appropriate to range. Surely vidual controllers here, I see an ANSP this (sometimes) busy international which is being badly managed by airport has at least one holding stack somebody several pay grades re- available? It should be used if neces- moved from the front line. The pro- sary to regulate inbound traffi c. Pilots cedural response to the unexpected always have an option (and these has failed because it off ered the same days often an obligation in their relaxed routine that usually works for SOPs) to decline a clearance which a situation in which it wasn’t going will lead them into an unstabilised to work. Individuals were left plainly

28 case Study comment 2 by Dragan Milanovski

The business jet on fi nal, without a landing clearance, came too close (half a mile) to the active runway which was still occupied by the departing aircraft already cleared for take-off .

It is not quite clear from the story what started reducing at the international low-visibility conditions with regular the exact outcome of this incident was, airport, to which three aircraft (all of runway confi guration changes is “opera- however all the options I could think of them signifi cantly diff erent types) had tions normal” for many controllers, too. were not “pretty”. Except for the plane to divert. At the same time, the runway The approach controller recognised that spotters- it must have been fun to confi guration had to change due to the circumstances were a bit strange but watch the "magical" work of the con- noise spreading to a less preferred op- she never had doubts whether she can trollers, but for them anything would tion wind-wise (10 knots tailwind). handle the traffi c safely. probably be considered fun, given the alternatives. It appears very unfortunate at fi rst It looks like both (Approach and Tower) glance (and probably is), but nothing controllers at the international airport The story describes a set of circum- extraordinary that the “system” could were doing their job as usual and did not stances that signifi cantly contributed not deal with. ILS could go out of make any signifi cant mistakes that con- to the event. The cargo airport op- service at any point in time for many tributed to this incident, but it also looks erator that never had money for im- reasons. Although this is not a com- like they could have done more provements all of a sudden decided mon event, safety should not be com- to prevent it. As soon as the to replace the main airside electrical promised should it happen. Handling fl ight crew of the business jet system. The big machines, involved aircraft diverting to another airport is asked for the ILS frequency, in digging the old system out, were introduced early in controller training the Approach controller probably the reason for the ILS going and, other than a bit more consider- should have realised that  out of service at the cargo airport, just ation and understanding, it does not the situation required spe- as it started snowing and the visibility require any special skills. Working in cial attention and addi-

Dragan Milanovski is an ATC training expert at the EUROCONTROL Institute of Air Navigation Services in Luxembourg. Most of his operational experience comes from Skopje ACC where he worked for a number of years in diff erent operational posts. Now, his day-to-day work involves ATC training design as well as Initial Training delivery for Maastricht UAC.

HindSight 16 Winter 2012 29 CASE STUDY

case study Comment 2 (cont’d)

I am very much against including safety verting to the international airport, it is to be expected that vacating the run- buffers into control actions by default, way might take a bit longer than usual – especially in low visibility conditions but this case was different. (another argument for the Approach controller’s plan B). The Tower controller should also have considered an additional safety buffer in his actions; he should have checked the distance and speed for the next aircraft in the sequence before lining up the departure. After that, he had only one chance for preventing the worse but he also missed it. As he cleared the departure for take-off, he realised there was a fast aircraft on final (two miles) that had not contacted him. By then, he must have been aware that if the departing aircraft did not start rolling immediately the landing aircraft would have to go around. When this did not happen (preceding aircraft slow to vacate) he should have immediately cancelled the take-off clearance for the depar- tional safety buffers. Don’t get me wrong, but we will never know. Her mind was ture, to ensure a safe missed approach I am very much against including safety set on a positive outcome of the situa- path for the landing aircraft. buffers into control actions by default, tion and a hope that the Tower control- but this case was different. For a start, she ler can deal with it. Well, he probably should have considered passing informa- could have dealt with the situation if A recommendation tion about track distance to touchdown (a he had known what to expect earlier. All the actions described above few times during the approach). Since it The Approach controller should have that could have been taken by was a straight-in approach this would not coordinated with the Tower controller the two controllers are very ba- have been a problem for her workload. in time about this fast diverting busi- sic and probably well known to Considering the aircraft’s speed and its ness jet with minimum distance to the both of them. It also appears that failure to reduce it, a “plan B” for increasing preceding aircraft in sequence. The they were professional in their the distance to the preceding aircraft in Tower controller would have probably jobs, with a good understand- sequence should have been employed as kept the departing aircraft on holding ing of the situation and aircraft soon as possible. Even when the sequence point. Finally, transferring the aircraft performance. However, both of was already established and it was evident form Approach to the Tower frequency them failed to realise that it was that the aircraft was above the glide slope also took longer; this was not difficult no longer business as usual, but and faster than expected, it was not too to predict given the situation. The last a situation that required special late to do something about it. She had safety buffer should have been an at- attention and increased safety doubts as to whether the remaining fuel tempt to transfer the aircraft a bit ear- buffers. I would recommend an on board the aircraft was sufficient for tak- lier than usual. additional Human Factors topic ing it out of the sequence (which is rea- in their regular refresher training sonable), but she never asked. There was a Bearing in mind that the preceding dealing with this issue. slight chance that there was enough fuel, landing aircraft (An-124) was also di-

30 Case Study Comment 3 by Mike Edwards

What happened next? ... This story is about a lack of positive The final nail that ensured the incident control both in the air and on the was going to happen – fill the gap – the As the business jet passed the hill upon ground. The Approach and Tower con- day job in mixed mode, one in, one out, which Brett and Sid where sitting, the trollers did what they always did and always fill the gap between arrivals. The FO, whose name was Dick, said to the ignored all of the clues that should gap was going to be tight regardless Captain, whose name was Dom, “Hey have raised the hairs on the back of but in these conditions it was asking for look at those sad anoraks, bet they their necks. There was no defensive trouble. wish they were up here”. Dom did not controlling. Sometimes we need a re- answer as he was busy trying to raise ality check about what we are about The lesson to be learnt – it does not mat- the Tower but was getting no reply. and when necessary add a mile or a ter if you lose the odd gap. You must be The frequency was strangely quiet. minute for the wife and the kids. prepared to vary the plan. This is not about pride. You are there to get every- The International Tower controller, The Approach controller was aware body home safely and never forget it. whose name was Phil le Gap (being that the business jet was fast and of Gallic extraction) phoned the Ap- above the glide path. She was con- As for Dick and Dom, their situation proach controller to ask for the busi- cerned for the fuel state of the aircraft got slowly worse by degrees, the boil- ness jet to be transferred to his fre- and so did nothing. We are there to ing frog, but Dick did not even seem to quency. She was leaving it very late assist the pilot, not second guess. Tell appreciate the rise in temperature. An this time. the pilot what you can see and ask him unstable approach, or what our friends if he is happy or wants to re-position. across the pond call a “slam dunk” and, As the business jet popped out of the The Approach controller was aware when the excrement finally hit the fan, last low cloud about one mile final, that the separation between the a complete breakdown in decision-mak- Dick and Dom were stunned to see a An124 and the business jet was erod- ing. Dick was left with a very strange de- departing aircraft just beginning its ing, but she did nothing, she did not sire for a herring and marmalade sand- roll. The conversation on the business even tell the Tower controller. Team wich. jet was something like this: work guys! Remember that it does not always go okay. Think back to Mexico Dom: Tower Yankee Echo Tango and the business jet that crashed on A Recommendation Dick: Go around approach in similar circumstances. If When you hear the voice inside Dom: Tower Yankee Echo Tango nothing else, think about your own your head telling you to make it Dick: Go around rear end. work, pause, take a deep breath and Dom: Land remember that a few seconds lost Dick: Go around The Tower controller did not change is better than the ultimate cost. his plan at all, despite all of the pieces At which point the aircraft, whose that began to stack up against it. A name was Joey, decided that enough heavy landing aircraft with a pilot that was enough and plonked itself firmly had not planned on being at this air- on the runway. Joey chased the other port, plus some other clues: Mike Edwards aircraft down the runway but gave up is Head of Safety Investigation at NATS (the UK Air and came to a stop. Dick and Dom’s fac- n decreasing weather, visibility and Navigation Service Provider). He has held this role for es were whiter than the snow that was runway state, possible long landing 7 years and prior to that he was Head of Investigation now falling all around them run, slow vacation, unfamiliarity. at London ACC. He had been an ATCO at Edinburgh and n A business jet that was faster than Heathrow before becoming the manager of all student The Approach controller was pleased expected and not on his frequency. controllers and then a Supervisor at London Terminal n with herself as the business jet did not Decreasing ability to see anything Control. He holds a PPL with Group B rating. come out the other side “see, knew it out of the window, your main would work” she said to her assistant. mode of working.

HindSight 16 Winter 2012 31 CASE STUDY case Study comment 4 by Anita Đuretić Bartolović

This story is a great example of how the aviation system is formed and aff ected by various expectations from various domains.

Even though, at fi rst glance, it appears ways a signal, to controllers and pilots workload was high, but she still did that the incident was caused by objec- alike, to increase alertness and atten- not ask any questions nor did she off er tive circumstances - adverse weather, tion and to be more vigilant. any assistance. Maybe they couldn’t inoperative ILS, strong wind, a politi- fi nd the charts they wanted - they cian’s decision on noise abatement Additionally, the ILS at the cargo air- asked for the ILS frequency – they had and runway confi guration – careful port was out of service, so cargo air- never landed there before and they reading reveals circumstances that craft were being re-routed to the in- had tailwind. As the distance between are related to the manner in which ternational airport, thereby increasing the business jet and the preceding air- the controllers performed their tasks. workload in the international tower. craft was decreasing, she assumed that And not in the professional sense of For the An-124 pilots and the business the Tower controller would be able to the word, I would say, rather in the jet pilots alike, it was their very fi rst handle it, but she was not certain. De- human sense. Their behaviour reveals landing at that airport, a fact which spite that, she neither contacted him that they too are human, with their at- should have been considered by the nor warned him of the insuffi cient titudes, thoughts, and feelings. In this controllers, but also by the pilots. distance problem and of the potential case, I would say that the contributing diffi culties that the business jet pilot factors to the incident are largely in The situation where the runway con- was experiencing. This in turn caused the realm of human factors - especial- fi guration had been changed four problems to the Tower controller, who ly the team-related aspects. times and where take off s and land- found himself under heavy load. ings were being performed in a man- Let us start with a few words about ner confl icting with the standards of Finally, stress was also a factor contrib- the objective circumstances the profession due to a political deci- uting to the error – the excessive work- which must not be over- sion, created additional load for the load on the Tower controller as three looked in this type of controllers because it required them things happened simultaneously: the review. The fi rst factor is to operate outside their normal rou- business jet approaching at very high the winter weather with tine. speed, the fact that it was two miles all its characteristics – behind the preceding traffi c and still grey skies, the cold, low The Approach controller concluded had not contacted the Tower, and the clouds, snow, wind. that “the wind situation was a bit information received after a take-off Adverse weather is al- strange”, yet she disregarded it and – clearance had already been issued to at least in her thoughts and actions – an aircraft on the ground that the An- she did not give it much attention but 124’s wing was still over the runway. carried on with her tasks because she All this caused severe stress to the Anita Đuretić Bartolović believed in her ability to handle the controller, and the incident happened. situation. The Approach controller saw Anita is psychologist and human factors the problem of the rapidly decreasing Although a part of the above explana- specialist . Currently, she works in the Human distance between the An-124 and the tion is already related to human fac- Resource Department at Croatia Control Ltd as business jet, but she was preoccupied tors, one of the biggest problems that a psychologist. She is a Lecturer for the subject with her refl ections on why it was hap- occurred is the communication prob- Human factors at the undergraduate study pening, and did not inform the Tower lem, specifi cally the lack of communi- program of Aeronautics, Faculty of transport controller about it. From the conver- cation and poor communication. The and traﬃ c sciences Zagreb. sation with the pilot, she understood law of human communication states that the pilots were busy and their that non-communication is impos-

32 information is limited to voice charac- The law of human communication states that teristics because one cannot see the non-communication is impossible – person and has no other non-verbal signals – such as facial expressions, we even communicate by being absent. gestures etc. – and verbal communication is limited by phraseology. Had the Approach controller known that verbal communication serves to convey sible - we even communicate by believed that they were performing their information, and non-verbal commu- ing absent. When I call a colleague on own tasks in a professional manner nication to convey attitudes and emo- the phone, I verbally transmit a piece and that they could handle complex tional conditions, she would probably of information to him/her, thereby situations. They thought that coordi- have communicated more in verbal helping him/her, but myself as well, nation was not required. Yet this time, mode. because I am sharing important infor- it appears to have been more than mation. If I do not call the colleague on required. Each controller worked inde- the phone, I also transmit information, pendently, forgetting about the team, A Recommendation but since there is no verbal mode of teamwork and coordination. How can such human-factor-re- operation, such information is more lated errors be avoided? The CRM prone to interpretation – it is neither What must not be overlooked in this (Crew Resource Management) reliable nor certain, as the colleague case is the pilot-controller communi- programme which has existed in does not in fact know why I am not cation, which was also deficient. The airlines for years now as training in calling. If I have an important piece of pilots of both the An-124 and the busi- interpersonal skills and TRM (Team information and I do not forward it, ness jet were new to the airport, they Resource Management) is being but I instead believe, maintain or am had never been there before, and they introduced into ATC. The aim of certain that the colleague can man- did not ask much. The controllers did these programmes is to reduce age without it, then it is detrimental not ask questions either, nor did they errors related to poor teamwork, to both me and the colleague. This is initiate communication, especially provide both pilots and control- precisely what happened in the case the Approach controller. Instead of re- lers with behavioural strategies in question. The International Tower questing verbal information on what for improved communication and controller had his hands full with both was going on in the business jet’s more successful teamwork and to arrivals and departures and did not cockpit, the controller made conclu- enhance flight safety.T he focus is communicate with the Approach con- sions based on para-verbal commu- on the skills required for a person troller because he completely trusted nication, which was equal to taking a to function more efficiently as a her to be doing her job professionally guess. From the para-verbal signals in member of a team. It would also as she always did. The Approach con- the pilot’s voice (the tone of his voice, be possible to develop and intro- troller was thinking the same way. pitch, volume) she concluded that duce a training programme in in- both pilots in the aircraft were very terpersonal skills and teamwork I suppose that her not contacting the busy and that the workload was high. enhancement for controllers and colleague was caused by over-reliance When we monitor non-verbal and pa- pilots together. on his capabilities, without verification. ra-verbal signs in communication, we This, of course, also leads to error - just do not aim at reading thoughts but at as overconfidence does. When we are understanding behaviour. Para-verbal not completely certain, we take things and non-verbal communications are for granted, and the assumption leads even more prone to misuse and mis- to error. There was no communication interpretation than verbal communi- between the two colleagues from dif- cation. Since radio communication is ferent positions as both of them be- in question here, the interpretation of

HindSight 16 Winter 2012 33 CASE STUDY case Study comment 5 by Dirk de Winter

Is the event I consider to be a threat for you really aff ecting you? Communicate and you will know. Threat and error management philosophy and techniques are nowadays well established in the worlds of both fl ight crew and controllers. The main idea is that a perfect world does not exist and in real life operations threats and errors are present that have to be managed successfully by all stakeholders to maintain fl ight safety.

advised by NOTAM that the ILS would told the fl ight crew she would give be out of service and in view of the them some extra track miles because Dirk de Winter weather forecast were carrying addi- they were getting too close to the pre- is has over 11,000 hours fl ying time over the last tional fuel. Clearly the fl ight crew was ceding aircraft. The fl ight crew would 22 years. He started as a cadet pilot with SABENA in under pressure: they had to ask for probably have been delighted with 1987 fl ying Boeing and Airbus aircraft. Before starting the ILS frequency, missed the tower the extra time to prepare the approach his fl ying career Dirk obtained an academic Master frequency and were too high and fast and if they really had been in a low-fu- degree in Electronic Engineering at the University of on the descent profi le. These were real el state they could still have declared Brussels. Since January 2009 Dirk has been working threats which mitigated against the an urgency or emergency situation. fl ight crew being able to make a sta- part-time in EUROCONTROL Agency. bilised approach. Knowing the runway Instead she acted in support of the was being utilised in mixed mode with supposed low-fuel state but passed a signifi cant tailwind and that the pre- the increasing real threat (reduced Techniques to identify, anticipate and ceding An-124 was not familiar with separation, rushed approach) on to manage these threats and errors are the airport, the approach controller the next controller and the fl ight crew. part of the Crew Resource Manage- still passed the speeding business jet The tower controller is faced with an ment (CRM) training for fl ight crew or to the tower controller. aircraft rolling late for take-off and the Human Factors training for controllers. business jet appearing fast on short Training focuses mainly on solving The tower controller too was unfa- fi nal. The fl ight crew of the business these issues within their own area of miliar with the size of the An-124 and jet is also under pressure because they competence and there is seldom a fo- expected him to clear the runway im- fi nd themselves on short fi nal with an cus on how threats evolve when they mediately via the rapid exit taxiway. A aircraft rolling for take-off in front of pass from one controller to another or B737 which had probably received a them. from controller to fl ight crew or vice conditional line-up clearance was al- versa. ready lined-up but was unable to take- off immediately as the An-124 could A reCoMMendATion The approach controller was present- be seen still not clear of the runway. Always ask yourself if a threat ed with additional diverting traffi c The tower controller was now faced you’re considering is a real threat from the cargo airport. They were of with an aircraft rolling for take-off and to you or your colleagues. if it’s a unfamiliar aircraft type, and unknown an aircraft in short fi nal being too fast. real threat mitigate it. never pass operators. Instead of mitigating the it on. Small threats will become threat associated with this she fo- Could this situation have been avoid- bigger for your colleague(s) es- cussed on the possibility of low fuel. ed? Was the controller sure the busi- pecially when combined with But was low fuel likely? Maybe the ness jet was low on fuel? Did she re- other unexpected threats. fl ight crew of the business jet were quest his fuel state? She could have

34 EXPECT THE UNEXPECTED!

SAFETY ALERTS Keeping Your Heads Up!

All EUROCONTROL Safety Alerts available on: www.skybrary.aero Subscribe to:

www.eurocontrol.int/safety-alerts EUROCONTROL

AD01.indd 5 03/12/12 16:33 FROM THE BRIEFING ROOM Training aircraft performance for modern/future ATc systems

by dragan Milanovski Knowledge of aircraft performance is essential for the provision of a safe and effi cient air traffi c control service. Every day, every hour, controllers are using this knowledge to provide separation, decide the allocation of cruising levels, create approach or departure sequences, use speed control for sequencing, provide wake turbulence separation and exercise other air traffi c control techniques.

Understandably therefore, aircraft performance has been are becoming technologically more advanced, with a lot of an important part of ATC training since the very begin- tools to support controller decision making. For example, ning. It starts with memorising basic aircraft performance having an arrival or departure manager deciding the se- data for the most common aircraft types and understand- quence is a more and more routine part of the job now. ing how these performance parameters may vary depending on the fl ight conditions. Later on, an element of aircraft performance is included in almost every practical exercise, allowing students to steadily build up the required experience by integrating previously acquired knowledge and understanding into control actions. This learning process is rather long, demanding and it continues during on-the-job training and even during the fi rst few years after validation.

The last decade has seen a general increase in aircraft performance fl exibility on one side and a greater pressure to achieve the most economic fl ight profi le on the other. In general, modern aircraft performance is aimed at keeping the cost of operation as low as possible. In other words, the less controllers interfere with aircraft performance the happier the customers are. When we do interfere reasonably, the probability is rather high that we will get what we ask for… well most of the time... In the same period, ATC systems have become or

36 Training aircraft performance for modern/future ATc systems

The cruising speed in Mach number form as well as rates of I would not exploit the common argument (used whenever climb and descent are available to controllers through en- something new is introduced in ATC) that controllers must hanced Mode S. MTCD is used to project ahead and predict have the required competence available should the automa- potential confl icts and CPDLC has been enhancing control- tion fail. This argument has been abused and over-used in the ler-pilot communication for quite some time now. In reality, past and it has been proven wrong in many instances. Main- not all the ATC systems are equally advanced but this is nev- taining an unused competence over long periods is practical- ertheless the broad direction for expected developments in ly impossible when the automation works as expected and the near future. without failures. Despite this, the subject continues to be one of the most popular in many discussions – but that is another Considering all of the above, inevitable questions arise for story… the two possible directions for aircraft performance learning in future ATCO training. Will controllers still need to do Instead let’s have a look at a few examples in the search for an all that hard work to build up their aircraft performance-re- answer to the questions above. lated competence as we do today when clearly this will be required less and less in the future? Or, will controllers need Use of Rate of Climb (RoC) to ensure a more complex type of competence, considering that they separation on opposite tracks will not need to use these skills very often, but when they do, it will be at a higher level of complexity, which involves This technique is often used to ensure that vertical separa- achieving minimal adverse eff ects on the economics of tion is established when an aircraft is climbing with confl ict- the fl ight profi le? ing traffi c on the opposite track, where the level cross will take place head-on, before the aircraft pass each other. It can also be used as a follow-up, when an aircraft has already been cleared to climb through the level of the opposite direction traffi c at signifi cant distance, but by monitoring the rate it has been determined that unless positive action is taken the aircraft will not reach the expected level at the required distance prior to crossing the opposite direction traffi c.

To achieve the most economic fl ight profi le, modern jet aircraft need to fl y the best climb rate for a given speed. The objective is to reach the cruising level (where jet engines are more effi cient) within the shortest possible time while maintaining optimum forward speed. Requesting an aircraft to increase its rate of climb, even for a small portion of the climb phase, has a negative economic consequence. The rate of climb is increased at the expense of the forward speed, so the aircraft will probably reach its cruising level sooner but will also take longer to get to its destination. It is also worth mentioning that the maximum rate of climb is limited by the minimum climb speed of the aircraft so the rate assigned has to be reasonable and for a short period. Once the restriction is cancelled, most of the excess thrust will initially be used to increase the forward speed at the expense of the rate of climb.

Alternative solutions for these sort of traffi c situations involve either use of an intermediate fl ight level until passing the op- 

HindSight 16 Winter 2012 37 FROM THE BRIEFING ROOM

Training Aircraft Performance for modern/future ATc systems (cont’d)

posite traffi c or radar vectoring. From the economic point of lower speeds than optimal are assigned – the fuel burn view, levelling off at a lower altitude rather than the cruising reduces but the fl ying time to destination increases. Both level has a greater negative impact than assigning a specifi c options have a negative overall eff ect on the economics of rate of climb. Radar vectoring is usually a more effi cient op- the fl ight profi le achieved. Radar vectoring for sequencing tion, but it cannot completely replace the rate of climb solu- generally increases fl ight distance, so it also has an adverse tion (because of airspace restrictions, workload etc). In addi- economic eff ect. tion, it also increases the distance the aircraft will have to fl y. To make things more diffi cult, most of these actions are planned well before the controller can even talk to the aircraft concerned – consequently knowledge, understanding of aircraft performance and experience are crucial for minimising any unavoidable negative cost impact. Another important point is that controllers also have to deal with “unknown factors”, such as routing and requested levels made by an aircraft which are beyond their control area, confl icting traffi c in the downstream sectors, wind, weather, turbulence etc.

Top-of- descent (ToD) and/ Allocation of cruising levels or use of Rate of Descent (RoD) and speed control Finally, let’s have a look into another typical scenario where Another typical situation is when two streams of traffi c are one aircraft needs to descend and pass through the level converging into one, where the controller’s job is to merge of one or more other aircraft which are on a crossing track them by grouping aircraft types with similar performance when the minimum distance between the descending and at same levels. The usual control actions include a combi- the crossing track aircraft would be well below the neces- nation of assigning a lower than requested cruising level, sary minimum radar separation. The two most common speed control to ensure longitudinal separation and/or ra- solutions involve either delaying the aircraft top-of-descent dar vectoring for sequencing (mainly to delay an aircraft). point so that it passes above the crossing traffi c or initiating its descent early and setting a minimum rate of descent Allocating a lower level for the cruise has a negative impact to ensure that vertical separation is re-established at a safe on aircraft fuel burn. Usually, assigning one or two levels be- distance prior the crossing point. low for a part of the cruise phase of fl ight is manageable, but more than that might have signifi cant consequences later For jet aircraft, the most economic fl ight profi le is with the on in the fl ight. Requesting aircraft to fl y at higher speeds engines at idle from the ToD until the aircraft arrives at the than the optimal also increases the fuel burn but reduces appropriate altitude and speed at the point where the ap- the fl ying time to destination. The opposite happens when proach for landing begins. If the ToD is delayed, the aircraft will have to descend with a higher rate after passing the crossing traffi c. Increasing the rate with engines at idle is generally not a problem because it can be resolved by adjusting aircraft pitch attitude. The secondary eff ect, which is also a limiting factor for the maximum rate of descent, is the concomitant increase in speed. If ToD is delayed within reasonable limits, then the excess speed can usually be dissipated by extending the speed brakes. In extreme cases, the aircraft may then have to fl y additional track miles at lower levels to reduce the speed. On the other hand, if initiating descent early and assigning a minimum RoD until passing the level(s) of crossing traffi c is deemed more appropriate, then the descending aircraft will reach the approach altitude earlier than planned and will probably have to maintain level fl ight at these lower altitudes.

38 Radar vectoring to ensure radar separation whilst vertical separation does not exist – in this case a very short period of time, as one of the aircraft will be descending – is usually considered a less appropriate solution as it increases the distance that the aircraft will have to fl y. But in exceptional cases, where the vectoring also provides a shortcut or when the minimum distance is close to the minimum radar separation and vectoring does not involve signifi cant turns (>5 degrees), it might be the best option.

The “unknown factors” mentioned in the previous example have an even more signifi cant role when the downstream sectors are lower and often busier airspace.

tive control actions which will usually have a negative cost impact. At the same time, future ATC tools will probably be available to support the controller decision-making so that we can provide the best possible service whilst minimising the negative eff ects of both , the “unknown factors” and cumulative eff ects. The required ATCO competence regarding aircraft performance will probably change, but this change is likely to be gradual and evolving to meet the future demands. Until then, here is a general guidance for assigning aircraft performance restrictions.

Whenever possible, aircraft should be allowed to fl y the shortest possible routes at their requested cruising levels, own speeds and rates. When positive control action must be taken to ensure separation, the economic factors should also be considered with high priority before selecting the Now let’s have a look at all three examples together, let’s most appropriate solution to a given traffi c situation. Per- say it is the same aircraft that is aff ected during diff erent formance restrictions should be kept to the minimum nec- phases of fl ight. The negative cost impact can accumulate essary to ensure separation and as much as possible should as the aircraft passes through diff erent sectors. Luckily, in be distributed among the aircraft concerned in a balanced reality the overall eff ect can be that successive impacts can- way. Adding extra buff ers by default (“just in case”) is not cel each other out or, more likely, their cumulative eff ect will a good practice and neither is applying double restrictions be lower than just a pure sum of the individual restrictions. (“just to be sure”). It is always a good idea to assign a limit for For the time being we have neither the overview nor any any restriction too, since this will allow pilots to plan ahead. control on how the various performance restrictions are When this is not possible, do remember to immediately can- accumulating during a fl ight, but at least there are devel- cel the restriction once it is no longer required. opments for the future that look promising (4D trajectory management).

Instead of an answer Dragan Milanovski is an ATC training expert at the EUROCONTROL Institute of Air Navigation Services in Luxembourg. It is challenging to predict the future, but looking ahead is Most of his operational experience comes from Skopje ACC where he part of the job and I will give it a try. The traffi c situations worked for a number of years in diff erent operational posts. Now, his described above are likely to continue to exist in the future. day-to-day work involves ATC training design as well as Initial Training I am sure you can think of many others too. Furthermore, delivery for Maastricht UAC. fl ying free routes will probably add to the need for posi-

HindSight 16 Winter 2012 39 FROM THE BRIEFING ROOM Automation, workload and safety

by Captain Harry Nelson During my first period as a test pilot I worked at the Blind Landing Experimental Unit (BLEU) at the Royal Aircraft Establishment, Bedford in England. One of the research topics that was very high on our work agenda was workload.

The research was led by Dr Allan Ros- ment and I thank Hindsight for giving control manually, sort out whatever coe, who was well ahead of the times me an opportunity to share some of has happened and then, when happy in his thinking and work. What it meant what I have learnt. with the flight path, energy situation for us pilots was that we were heart and technical configuration, select the rated on every flight and every task, If we look at the aviation definition of autopilot back on again. However, an some of which were very demanding automation, it is “something that is de- increasing trend is for pilots to use up indeed. We also used eye marker tools signed to decrease workload”. Oh, that significant workload capacity in get- to “see” what we were looking at. We it were so simple! If we take the autopi- ting the automatics to do what they coupled this to skin acidity measure- lot, we have seen the progression from want, sometimes at the expense of ac- ments and we also rated each task in basic attitude hold systems through to curate flight path control. .Exception- terms of workload. In fact we used a modern day, very sophisticated auto ally, this can lead to situations where workload rating scale to assess our flight path control systems which are the safety of the aircraft may be put in level of workload. It was developed perfectly capable of controlling an air- question. from the famous Cooper-Harper rating craft from just after take-off through scales for aircraft handling qualities. to touch-down. The day of the auto Ok, so I have introduced a new term, For those interested, a quick look on take-off is, I suspect not too far away as workload capacity. We all know what the internet will provide many details more and more experience is gained that is or we all have an idea of what it on these interesting areas of research. through RPV types of aircraft and then is about. Let us look at it in a bit more who knows what may come next? detail. In brief, I learned a lot about workload, which was to stand me in good stead over the rest of my test flying career. I ...an increasing trend is for pilots to use up also believe that there are real parallels significant workload capacity in getting the with workload in the air traffic environ- automatics to do what they want, sometimes at the expense of accurate flight path control. Harry Nelson has had a flying career spanning some 46 years which has focussed on Certainly the autopilots of today do re- flying training and test flying as the two main duce workload and they are extremely It may be helpful for all controllers (and activities. A graduate of the Central Flying successful at doing it. This, coupled to pilots) to consider that at any moment School and the Empire Test Pilots School he has the greater reliability of modern jet you have a given workload capacity and operated in all parts of the world and worked transport aircraft can lead to pretty low it changes with many variables. Your general arousal states for the crew. Most health, your fitness level, your degree at 5 flight test centres throughout Europe end- flights are completed in a very easy and of stress, your training level, your expe- ing up in Airbus where he now holds the post low workload state. But what happens rience, your fatigue level, your age, and of Executive Operational Advisor to Product when things do not go as expected? your circadian rhythm are just some of Safety. He has over 10000 flight hours on over the many factors affecting your capac- 76 types of aircraft. For most pilots of my era, the natu- ity for work when examined on an in- ral tendency is to immediately take stantaneous basis. You will notice from

40 Automation, workload and safety

even this list that some of these are under your control and others are not. capacity limit, several unfortunate ef- hearing may well be aff ected. He hears For ease of understanding I will group fects start to become evident. Firstly sound but maybe does not register mental workload together with what he sheds tasks by priority. It may be the content of the communication in I call motor function workload or that one of the fi rst to go is the lon- a normal way. If he hears it then he physical workload. When a pilot takes ger-view monitoring of his situation. may not be able to resolve what he is control from the autopilot, his motor He stops “tracking mentally ahead of being told and messages may have to function workload takes a step change the aircraft or at least he may not look be repeated several times before they upwards and immediately demands so far ahead in terms of threats and “get through”. Of course this tendency the use of more of his spare workload things to avoid. Later, with a further in- will also cut across the potentially capacity. If he was using this capacity crease, he focuses only on what is hap- helpful Crew Resource Management for mental tasks like trying to resolve pening inside the fl ight deck and fi nal- (CRM) behaviours that all airline pilots an issue with the aircraft or communi- ly he may tend to “fi xate” on an issue are aware of today and eff ectively iso- cate on a detailed route change, then or a parameter or a course of action at lates that crew member in his “close to he potentially may have a problem. As the expense of others which may be becoming overloaded” state. As the workload gets closer and closer to his more important or more helpful. His overload condition takes its full grip, 

HindSight 16 Winter 2012 41 FROM THE BRIEFING ROOM

Automation, workload and safety (cont'd)

he may well be focusing on only one instrument or even one parameter.

A simple analogy may be useful here. Add a crying child in the back of the their individual team members so that Imagine you are driving an older gen- car and things can go critical as many they can keep the whole operations eration car and you want to change husbands and wives will testify – and safe but we also know from the real the radio channel. On an autoroute, hopefully laugh about it later. world that occasionally expediency motorway or autobahn it is easy. On a rules and “there is simply no one else”. two lane road with traffi c coming to- Learning to recognize your personal The same applies to pilots. Once again, wards you, albeit separated by a white symptoms as you reach your work- I must put some of the responsibility line, you need to take some care be- load limit is something I would com- onto the shoulders of each pilot and cause more of your capacity is being mend to everyone but you must go ATC controller. Only you know how used in ensuring the trajectory of your further and also decide before you get you feel right at this moment. Only you car remains on your side of the road. into such a situation, what you will do know whether the new baby kept you And fi nally, if it is night and you are about it as you see those symptoms awake all last night and you are feeling driving along narrow country lanes it starting to impact your performance. really tired. Instead of being ”macho” becomes a task not without risk and You need to formulate an action plan. about it, recognize your potentially you may either switch the radio off One of the best action plans before degraded workload capacity state and or develop a new technique to do it. you hit the “black hole” as I call it, is to inform the other pilot, the supervisor, You reach for the knob without look- call for help. Inform someone immedi- or the controllers operating the adja- ing and then verify with a quick glance ately as soon as you feel that capacity cent sectors. They can help and all will that you have the right one. Then you is becoming limited. Ask the other pi- have experienced similar situations. tune by ear using minimum eye move- lot to take over control. Leave it and it ments as you carefully steer the car may become too late. We know from our Human factors along the diffi cult bendy road. During studies that there is an optimum this action it is quite possible that you Looking from afar I know that much arousal and activity state in terms of would not hear a passenger in the car work has been done on this subject in workload. Too little and our battle is talking to you or letting you know that the world of ATC regarding the num- with boredom and inattention and all you had just passed the intended turn ber of aircraft a given controller can that can follow from that. Too much off point. You can build up this work- handle during “normal fl ow” and also and we can hit the black hole. We work load scenario by imagining that you how many if the situation changes, best when working within our capac- are driving on a route unknown to you let us say by one aircraft declaring an ity in an alert and active manner. That so as well as the motor function eff orts emergency. I am also sure that ATC must be the target of each one of us needed to steer, accelerate etc you are supervisors try to be aware of the ca- as we go about our business in this also thinking hard about the route. pabilities and workload capacities of safety-driven industry.

42 The mechanisms i have learnt to help me through those potentially very high workload periods include the following:

In general I try to shed unnecessary workload so as to maintain a (1) greater level of spare capacity.

If it is quiet and low arousal that is the threat, imagine an emergency (2) and run through it in your mind what you would do in detail. If you have forgotten something go and check the books.

If it is busy, I try to be a bit schizophrenic by fulfi lling my primary task (3) for sure but also trying to stand outside myself and “observe” my own behaviour as if I am in one of those video car racing games where the car is ahead of you. This way it becomes possible to see some potentially risky behaviours before they happen.

As soon as I sense the early warning signs of a signifi cantly build- (4) ing workload situation, I ensure that the other pilot is aware and get In conclusion him to take more of the non fl ying tasks, leaving me free to concentrate on fl ight path and energy control. If he gets overloaded it may Pilots and controllers can help each oth- slow down communication or delay a procedure but if I, the fl ying er in this workload issue. It is reasonably pilot, get overloaded the situation would be much more serious. It easy for an experienced pilot to judge was interesting to note that Captain Sullenberger, in the Hudson how hard a controller is working and I accident, left his co-pilot to deal almost completely with the drills and am sure that the reverse is true. Why is it attempts to relight while he focused on the water landing. then that I hear pilots putting even more pressure on controllers who are dealing Of course the doctors are also right. It is important to stay fi t and to with, for example, a low visibility opera- (5) ensure the right amount of rest and food. In the RAF years ago it was tions situation. Complaints about hold- a punishable off ence not to take breakfast. So it became normal to do ing times, expected approach times so, a routine that still works and frankly for me, it remains the most and the rest do not help anyone. They important meal of the day. add to the overall “noise” and cause ir- ritation to all real professionals. What Finally, a word about the “black hole”. The hole is like a whirlpool in both the pilots and the controllers need (6) that you tend to get drawn progressively into it without the apparent is clear minimized information. From strength to get out again. If one fi nds oneself at the very edge of the the pilots, the controllers need to know black hole, the only mechanisms I know for a recovery are to take a if there is a real fuel shortage or any mental seat in the video game viewing place and take a fresh look other operational constraint so that the at what is going on and to force oneself to examine all the instru- right prioritization can be made. From ments, starting with the attitude indicator, to seek out those instru- the controllers, the pilots need to know ments that are giving good information. I would guess that the cor the changing weather situation, when responding situation for controllers is fi xation on one “tricky” aircraft they can expect to start the approach at the expense of others that may become threats to the overall safe so that their passengers and company situation. Releasing the thing or parameter that you have fi xated on can be informed, fuel can be managed in favour of good parameters is not easy. We all have the desire to and maybe in exceptional circumstanc- make the facts fi t the belief or decision that has already been made. es, the aircraft can be diverted or an Even when faced with overriding evidence that the initial assumption emergency declared. The rule has to be was wrong, we still cling to it and try to get a “fi t” from the other pa- the greater the workload – the greater rameters. It is vital to re-examine the data in front of you in a fresh the assistance we need to be giving way. each other.

HindSight 16 Winter 2012 43 FROM THE BRIEFING ROOM Water/slush on the runway and What every tower controller should know about it

by Gerard van es, Senior Consultant, nLr-ATSi, The netherlands. Analysis of accidents that occurred in the last 20 years has shown that the risk of overrunning the end of the runway on runways covered by liquid contaminants such as water or slush is about 10 times higher than on a dry runway.

The hazardous eff ect of water/slush the tyres displacing the water or slush problem of water/slush on the run- on aircraft fi eld performance was fi rst and drag due to impingement of the way is more acute for turbine engine brought into prominence after the ac- spray on the aircraft thrown up by aircraft than for piston engine aircraft cident to the BEA Airspeed Ambassa- the tyres. It was shown that this drag because of their higher ground speeds dor aircraft at Munich in 1958 in which increased with increasing water/slush and their increased susceptibility to in- 23 people were killed. The increasing depth. It was also discovered that gestion and impingement due to their prevalence of tricycle undercarriages there was a possibility of loss of en- design. and higher aircraft operating speeds gine power, system malfunctions and in the late 1950s were associated with structural damage due to spray inges- Let us have a look at some typical this new hazard to aircraft operations. tion or impingement. Also directional numbers of the eff ect of water/slush In the early 1960s investigations on control problems were found when on take-off performance. Just 13 mm the eff ects of water/slush covered crosswind conditions existed. Further- (0.5 in.) of slush can subject a large runways were carried out in the Unit- more the problem of very low braking jumbo jet to a drag that is equal to ap- ed States, the United Kingdom and friction between the tyres and surface proximately 35% of the thrust of all its France. Tests were conducted using was identifi ed in which aquaplaning of four engines. This number increases to catapult-driven test carriages as well the tyres played an important role. The 65% for 25 mm (1 in.) of slush making as actual aircraft. These early tests gave a clear picture of what water and slush on the runway do to an aircraft that takes off or lands. It was found that the acceleration during take-off was reduced due to the drag eff ects on

Gérard van Es works as a Senior Advisor fl ight safety and operations for the NLR-Air Transport Safety Institute - Amsterdam, the Netherlands. He is currently involved in the European working group for the prevention of runway excursions.

44 Some examples

(1) On 14 August 2005, a British Airways Regional Embraer 145 overran Runway 27L at Hannover by 160 metes after fl ying a stable approach in daylight but then making a soft and late touchdown on a water-covered runway. Dynamic aquaplaning began and this was followed by reverted rubber aquaplaning towards the end of the paved surface when the emergency brake was applied. The aircraft suff ered only minor damage and only one of the 49 occupants was slightly injured. http://www.skybrary.aero/index.php/E145,_Hanover_Germany,_2005_(RE_HF_WX)

(2) On 10 November 2010, a Kingfi sher Airlines ATR 72 made an excessively steep and unstabilised tailwind approach in light rain to runway 27 at Mumbai in visual daylight conditions. After touching down late, the aircraft was steered off the side of the runway when it became obvious that an overrun would otherwise occur. The Investigation found that ATC had failed to advise of water patches on the runway and aquaplaning had occurred. It also found that without aquaplaning, the available distance from the actual touchdown point would have been suﬃ cient to stop the aircraft in. http://www.skybrary.aero/index.php/AT72,_Mumbai_India,_2009_(RE_HF)

(3) On 24 November 1998, a KLM uk Fokker 100 overran runway 20 at Southampton after a late and fast daylight touchdown in rain was followed by poor braking. The Investigation found that the assessment of the runway as ‘wet’ passed by ATC prior to the incident was correct but that sudden heavy rain shortly before the aircraft landed had caused a rapid deterioration to somewhere between ‘wet’ and ‘fl ooded’. Slow drainage of water from the runway was subsequently identifi ed and the runway was grooved. http://www.skybrary.aero/index.php/F100,_Southampton UK,_1998_(RE_HF_WX)

a take-off impossible. In general for a can be difficult because aquaplan- instance more water/slush can give multi-engine transport aircraft, just ing is likely to occur on water/slush lower take-off weight penalties. Not 13 mm (0.5 in.) of water/slush can in- covered runways. This will increase all aircraft manufacturers account for crease the take-off distance by some the landing distance compared to these affects during the landing. 30-70%. a dry runway. However, although it sounds strange a thicker layer of wa- There is another important diff erence Additionally, there is another potential ter/slush can be better for landing between an aircraft taking off and one hazard associated with taking off and performance than a thin layer. The landing on a runway contaminated the presence of slush. There is possibil- drag generated by the water/slush with water/slush. The former can as- ity that the slush will be taken into the helps to stop the aircraft. The more sess the situation before and during air on probes and in wheel wells and water/slush you have on the runway the early stages of the take-off roll then freeze quickly as air temperature the higher drag on the aircraft. This whereas the latter has just a few sec- drops in the climb. also applies to rejected take-offs and onds to complete a much more sub- can lead to strange performance jective assessment. Night operations Slush can have an adverse eff ect on restrictions when taking off from can make both judgements much the landing performance too. Braking water/slush covered runways. For more diffi cult.

conSideRAtionS FoR contRoLLeRS

n For pilots it is extremely important to have the most ac- n Controllers almost always rely on the aerodrome op- curate, complete and up-to-date information regard- erator to provide information on the runway surface ing the runway condition and weather conditions that condition. Inaccuracies in these reports are always could infl uence this (e.g. heavy rain showers). possible and diffi cult to identify by the controller. During daytime the controller might observe areas n Controllers should realise the potential impact of a with water puddles or slush on the runway and in- water/slush covered runway has compared to a wet form the crews about this. runway. There is a big diff erence in infl uence on operational safety between a wet runway and a water/slush n Any ‘pilot reports’ passed to subsequent aircraft by covered runway. ATC in respect of water or slush should be accompanied not only by how old they are but by an ‘un- n Air traffi c control plays an important part in this infor- offi cial’ comment as to whether it appears from the mation provision. There have been cases in the past in Tower as though the situation has materially which incorrect or outdated information regarding the changed. runway condition was provided by the controller to the pilots, leading serious incidents.

HindSight 16 Winter 2012 45 FROM THE BRIEFING ROOM

by captain Hans-Joachim ebermann How could this almost fatal blackout have occurred? Was the crew totally incompetent?

While they may have been incompe- BLACKOUT tent, research reported by Etienne Koechlin, Head of the Cognitive Neu- roscience Laboratory at the ENS (Ecole new fi ndings regarding Normale Supérieure) in Paris to a recent Conference suggests a more like- decision-making ly scenario. One third of the entire brain looks after decision-making in the prefrontal cor- tex. Three areas of the prefrontal cor- A 747 en route from Germany to Los An- Las Vegas is extremely busy and is tex can be distinguished: the middle geles. The Captain takes the fi nal crew situated in a valley that does not al- sector controls motivation, the lateral rest break. The plane had been a little low long radar vectors. The approach area controls the selection of action heavier than usual on departure. En is steep and, as the aircraft joins the options and the lower area processes route, the winds are somewhat stronger fi nal approach track with a very high emotions, personal preferences, etc. than expected and the planned even- crew workload, it is travelling much All three areas work independently, tual fl ight level is not achieved. Nearing too fast. So fast that later inSERIALITY the sub- OFbut EXECUTIVEcommunicate CONTROL with each other Las Vegas, the two First Offi cers on the sequent interview with his fl ight safe- constantly. But the brain can only fl ight deck decide to call the Captain ty manager, it becomes clear that the ever make one decision at a time even ■ Frontal lobes can make only one decision at one time back from his break, because they be- aeroplane may well have failed to stop though two or three situations need- lieve that on arrival in LA they will no on the runway had a landing■ They cannbeenot co ntroling the a concurrentdecision executioncan be monitored si- longer have minimum diversion fuel on attempted. Although this must of multiple have tasks multaneously(routines/procedures) and the actions initiated board and want to recommend a fuel been more or less clear to each of the after a decision can be monitored to stop in Vegas. The Captain initially wants three pilots on board, nobody at fi rst see whether the desired outcome is to fl y on but the First Offi cers’ manage said "go around" even afterStimulus the 1 1000ftPERCEPTION DECISIONachieved.ACTION to convince him that they should land gate was passed. It was not until very Stimulus 2 PERCEPTION DECISION ACTION and refuel. Reluctantly and still tired, late, close to touchdown, that one of Translated into the FORDEC decision- the Captain initiates an en-route diver- the two First Offi cers fi nally said "go making model, this means that steps sion to an airport which is unknown to around" and thus prevented the cer- F, E and C (Facts, Execution and Check) all three of them. tain crash. will be processed in parallel, while

SERIALITY OF EXECUTIVE CONTROL GAIN CONTROL ON SELECTION PROCESSES

Increased outcome values at stake ■ Frontal lobes can make only one decision at one time Too low arousal Lower Higher Saturation ■ They cannot control the concurrent execution Stress of multiple tasks (routines/procedures) Max

Stimulus 1 PERCEPTION DECISION ACTION Neuronal activityNeuronal

Stimulus 2 PERCEPTION DECISION ACTION Min No control Selected routine Selected routine Selected routine Action routines

Figure 1: The brain can only make one decision at a time. Figure 2: An overload may temporarily prevent the brain from taking any decisions whatsoever.

46 GAIN CONTROL ON SELECTION PROCESSES

Increased outcome values at stake

Too low arousal Lower Higher Saturation Stress Max Neuronal activityNeuronal

Min No control Selected routine Selected routine Selected routine Action routines steps O, R and D (Options, Risks and De- There are several ways of preventing cision) will only be processed in succes- overload situations arising, or mitigat- sion. These are physiological laws which ing the consequences if they do: cannot be infl uenced by training (see fi gure 1). n Staff selection: the individual resilience and ability to cope with To speed up and simplify decisions, af- stress of those applying to be pi- ter each decision with a satisfactory lots should be as great as possible outcome the brain stores a routine or at the time of their selection. strategy to which it will refer in a similar future decision-making situation. n Training: four recurrent simulator events are the industry standard. Therefore experienced pilots are able Initial training should have an to take decisions more quickly and with empirical basis in terms of scope greater certainty. On the other hand, if and quality. Making cuts with no objectively dangerous situations such thought for the long-term eff ects as unstabilised approaches have been spells suicide for airlines. individually found on several previous occasions to be manageable, this can be n Better individual stress and fatigue critical. The brain then stores the "con- management tinue instead of go-around" routine and modifi es this routine only after a failure. n Active intervention in overload In this case, after a landing overrun. situations to rectify matters: This appears is a somewhat unsuitable learning process. Which is why our SOPs In the latter respect, accident re- and limits are so important, because view shows that whilst interven- only they defi ne the boundary between This brings us back to the example tion is usually very good from the safety which is objectively necessary at the beginning: there was no moti- Captain to the First Offi cer, but it and safety which is individually (and vation on the part of the Captain to is often poor where the Captain wrongly) perceived as manageable. make an en-route diversion; the crew is the person who is overloaded. did not know Las Vegas; the aggres- Particularly where he is the also Equally important here is our training, sive radar vectoring was a surprise; the PF and the First Offi cer is the Pilot for example in the simulator, where by plane was allowed to fl y too fast; the Monitoring (PM, a more recent al- handling as many diff erent problem sit- go-around was diffi cult; ATC and traf- ternative designation for the PNF). uations as possible, routines and strate- fi c monitoring also played a part, etc. After a signifi cant deviation from gies are stored in the brain, to be relied It is therefore highly likely that while an SOP in particular, the First Of- on in an emergency. Savings in training, the fl ight crew involved were "know- fi cer may no longer be sure when such as shortened transitions and only ingly" clear about the consequences he should intervene. Such signifi - three instead of four recurrent simulator to be expected, they were for a short cant deviation from an SOP can be events per year, are therefore potential- time unable to decide to abort the ap- a consequence of an overload but ly unsafe. proach. it can be diffi cult to judge if it is. Captains who routinely fail to fol- It is obvious from what has been said so This is a possible explanation of the low SOPs (keywords: private pro- far that the brain cannot deal adequate- concept of target fi xation: temporary cedures, operational pressure) are ly with situations where it is overloaded. overload leads to this state of "inability liable to discourage intervention Too many stimuli and/or too many tasks to take a decision". People knowingly from their First Offi cer at precisely to be handled in parallel place us under rush headlong towards an accident the time when they might really excessive strain. Enormous stress is gen- and if they survive are subsequently need it as they fi nd themselves in erated particularly where serious conse- unable to explain their behaviour. It is overload. quences are likely - an accident such as therefore obvious that overload situa- an overrun, but also "just" a failed com- tions should wherever possible not be This alone is a pretty convincing petency check. Such overload can lead allowed to arise in the cockpit because argument for adhering to SOPs. to a situation where the brain is tempo- they cannot be managed with a suffi - rarily no longer capable of taking any ciently high probability that a safe out- decisions (see fi gure 2). come will result.

HindSight 16 Winter 2012 47 FROM THE BRIEFING ROOM Engine failure on take-off "hey, what’s it doing now?"

"BestAir 11, runway 18, cleared for take-off , left turn..." The old Airbus 340 of Best Airways starts its take-off run and is gaining speed majestically. Just when it is passing the halfway point of the runway, a fl ock of birds ahead thinks they had the clearance fi rst. As the inevitable collision occurs, the number 1 engine takes a direct hit and emits some fl ames to show its displea- sure. Soon the aircraft's nose starts to rise…

In normal operations an aircraft should fl y according to the given departure clearance, which may be using a published SID or using some other ad hoc clearance.

Acceptance of such clearances assumes that all engines are functioning normally. For engine-out situations the aircraft operator to operator. The operators when the aircraft passed 130 knots, operator is responsible for checking may also have outsourced obstacle "Oh my… birds! many birds!" As several obstacle clearance and per- clearance evaluation and the plan- large birds hit the aircraft, a number of formance data and detail- ning of engine-out procedures to an thumping sounds were audible in the ing for each departure any external service provider - although fl ight deck along with a momentary vi- variations which might be of course that does not aff ect their re- bration of the airframe. The cabin crew needed to retain suffi cient sponsibility in this matter. It is normal heard this too. The engine failure drill obstacle clearance. to keep these procedures as simple as appeared on the ECAM. The Captain possible. The procedures are planned maintained directional control and af- The engine failure re- for each runway separately. Typically, ter the "Rotate" call from the First Of- sponse procedures ad- an engine-out initial fl ight path will fi cer, he began a rotation towards the opted may vary from continue on the extended centre line engine-out target pitch attitude… of the runway to a pre-determined distance and/or height, after which a In modern air transport aircraft, the turn towards an engine-out holding take-off performance is established Sami Laine position may be specifi ed. These pro- before each fl ight. This can be done completed an MSc in accident and safety in- cedures are planned so that an aircraft using an EFB or more traditionally by vestigation at Cranfi eld University. He is a First can continue climbing after level ac- reference to the performance manual celeration and clean up of the aircraft or by use of pre-calculated take-off Oﬃ cer on the Finnair A340/320 fl eet, fl ight confi guration. weight tables and a speed booklet. safety analyst and Editor of the Finnair safety Performance calculations take into magazine. His earlier working experience was …In the fi ctitious example we began account a number of factors such as with the Maritime Rescue Coordination Centre, with, the Captain was the pilot fl ying runway characteristics and conditions Helsinki. and the First Offi cer was the monitor- (e.g. runway slope, length, possible ing pilot. "V1", the First Offi cer called contamination etc), weather factors

48 (e.g. temperature, wind velocity), air- navigate, communicate”. However, them to follow any the instructions craft weight and intended engine even with this priority, many aircraft given by the cabin crew… thrust settings (full or reduced to ex- operators expect their pilots to at least tend engine life). inform ATC of a problem that will af- Landing performance is calculated in fect compliance with a clearance early a similar way to take-off performance, Applicable regulations define margins on without getting involved in what with corresponding safety margins. for each part of the take-off and initial the alternative will look like by follow- Landing performance calculations climb, which must be met for every ing with a "stand-by". This allows flight must also take account of go-around take-off. The main principle is that the crew resources to be focussed on the performance. Again the operator is most demanding requirement defines initial piloting and engine failure tasks. responsible for ensuring that the en- the maximum weight of the aircraft gine-out climb gradient would meet for take-off. The required take-off dis- Typically, when abnormal check-lists obstacle clearance requirements. If the tance is defined as the distance from a are being performed, the pilot flying normal published go arounds can't be standing start to a height of 35 feet and temporarily takes over communica- used in the case of an engine failure, both all engines and the engine-out tions with ATC to allow the monitoring the operator can use higher decision scenarios are considered. The obstacle pilot to focus fully on the prescribed altitudes or publish a special proce- clearance of 35 ft may be reduced to engine failure tasks. Whether this hap- dure, which may differ from the ATC 15 feet if the runway is wet or contami- pens or not, the priority for the pilots is expectations - at least if the flight crew nated. The obstacle clearance require- to establish and maintain overall situ- don't remember to tell ATC about it. ment is based on calculated net flight ational awareness. path, which is the gross, or theoretically … Finally the aircraft joined the ILS achievable, flight path reduced to ac- … The flight joined the holding pat- for runway 18 and the Captain made count for an aircraft flown in a typical tern over the VOR about 20 miles a smooth touchdown, selected full re- way rather than with perfection. from the airport. Once the aero- verse on the three still-functioning en- plane was in the hold, the First Of- gines and the auto brake took effect. ... many aircraft opera- ficer finished the engine failure Almost the whole length of the run- checks according to the ECAM and way was used and as the brake tem- tors expect their pilots started to prepare landing distance peratures rose, the tyre safety plugs calculations for an overweight land- began to melt and some of the main to at least inform ATC ing. The First Officer took over as gear tyres deflated. As the aircraft of a problem that will pilot flying whilst the Captain called could not now taxi, it blocked the only the cabin crew chief and explained runway at the airport. The passengers affect compliance with the situation and requested ap- had to disembark to buses down ex- propriate preparations for land- ternal stairs. Although the flight ended a clearance early on... ing. He also made a public address without any drama, the working day announcement to the passengers of the crew and the ATC still had some … Meanwhile the aircraft started to to keep them informed and asked way to go. lift off from the runway surface. After checking, the First Officer called "positive rate" and the Captain responded Sidestory / fact-box – the pre-departure briefing for engine failure by calling “gear up”. From the Tower it looked like the aircraft barely avoided during take-off given by the pilot designated as the pilot flying. hitting the runway lighting installa- These briefings will vary from operator to operator and between aircraft types. They are often only given tions. There was no transmission from in full on the first flight if the crew will be operating more than one flight. But the following example of the pilots and it quickly became ob- one shows the increased workload for pilots during the initial stages of an engine-out situation. Remem- vious that the aircraft was not going to follow its departure clearance. The ber, the priorities of cockpit crew are aviate, navigate, communicate! aircraft had passed the SID initial turn "After V1 continue take-off, when you call positive rate, I call for gear up. Initial target 12.5 nose up, when and appeared to be flying straight indicating - follow SRS. Fly and trim, after trim and at least 100 feet radio altitude – autopilot on. After ahead. At last there was a PAN call ad- 400 feet I call for the ECAM actions. In case of a flame out, continue to master switch, if fire or severe vising of an engine failure and the pi- damage, continue to fire bottles according to ECAM. At acceleration altitude, push to level off, accelerate, lots’ intention to turn towards the VOR clean configuration. At green dot speed (optimum engine-out climb speed) set safe altitude and pull the after they had passed 1700 feet… altitude knob. The local procedure here is to continue ahead to 8 miles, and then turn left to the beacon. Safe altitude here is 2300 feet. The SID assumes an early turn, so inform ATC with mayday and tell them In the flight deck, the priority is to fly to standby..." the aeroplane. The priority is “aviate,

HindSight 16 Winter 2012 49 FROM THE BRIEFING ROOM close interactions of the disconnected kind

The relationship between a pilot and a controller is a complicated one. It is critically intimate, yet pragmatically distant. It is built on mutual trust, yet cannot aff ord blind reliance. Safely seated at his or her station inside a building, the air traffi c controller issues instructions and clearances to the pilot of an aircraft way up in the sky, often many miles away. The interaction, which often lasts no more than just a couple of minutes, is highly pro- ceduralised. And yet, despite its highly critical function of getting an aircraft safely to or away from the ground, and the fact that it is carried out between experts, conscientious professionals, there are occasions when this interaction goes wrong. When it does, it compromises the integrity of a fl ight and potentially puts an aircraft (and its crew and passengers) dangerously close to an accident. Let’s examine three such cases:

Case 1. captain (pilot fl ying) reporting: “On descent into ATL on the CANUK Seven… we were asked to keep our speed up … [and] given the clearance to cross CANUK at 12,000 ft at 250 kts. We began by Loukia Loukopoulos and immanuel Barshi our descent to comply with the restriction. Approximately 30 miles Note: this article is based on voluntary reports by from CANUK, …our clearance [was revised] to level at FL230. Our de- pilots and air traffi c controllers to NASA’s Aviation scent rate was close to 4000 ft/min. Safety Reporting System System (ASRS), which give [There was] no way we could level at the reporters’ perspective on events that they believe FL230… [the PM] transmitted "Un- able to comply"… as we descended compromised safety. As such, it refers to fl ight through FL210… operations that take place in the United States [ASRS 878704, march 2010, B757, national airspace system. iFR on descent]

50 close interactions of the disconnected kind

While fl ying into a busy airport, the can continue leading the aircraft to a safe crew of this aircraft, is expecting (per Case 2. captain (pilot monitor- landing. Without the quick awareness the published procedure) to cross CA- ing) reporting: “We were descend- aff orded by a moving map, such as the NUK “at 14,000 or as assigned by ATC.” ing…with clearance to descend one available on later generation aircraft Having let the aircraft automation cal- via. We did not get a runway (24R) of the same type and model, the crew culate the optimal path to the 14,000 until almost the end of the STAR wastes valuable time (and undoubtedly ft restriction (in line with company pol- … I selected the runway and the experiences frustration) by attempting icy), the crew is letting the autopilot transition, but could not close the to enter “invalid” information into the determine the appropriate angle and discontinuity that showed up on computer. In fact, the aircraft autopilot speed that will bring the aircraft to CA- the ight management computer. has already (in accordance with its de- NUK, at the right altitude and the right The pilot ying reselected the same sign) “dropped” the commanded auto- speed. The controller, concerned with and executed. We both then real- mation mode in response to the invalid managing the fl ow of traffi c, perhaps ized ENGLI was behind us and LNAV information, and the aircraft automation also in an eff ort to help the aircraft ar- had disengaged… Since this was a is not in the mode the crew expects. rive sooner at its destination, issues a [Boeing 737]-500 without a moving new instruction. This instruction re- map, I relied on the FMC to know The crew relies on the approach control- quires the aircraft to reach a lower al- which xes the aircraft had already ler to provide timely information, and titude sooner, something that requires passed. .. About the same time, ATC also relies on the automation to the point a steeper descent. When the controller gave us a vector and cleared us di- of not always paying careful attention to reacts to changes in the traffi c fl ow, rect to KONZL. We then ew the ILS their location and path. Both ATC and however, and issues a diff erent level- 24R uneventfully to a landing.” the automation are so reliable so much off altitude, the aircraft is in a high rate of the time that pilots are sometimes of descent – so much so that the crew [ASRS 929900, January 2011, lulled into over-reliance. An early issue hardly has time to respond to the con- B737-500, iFR on descent] of the approach and runway clearance troller that it will not be able to comply would have saved the crew from fi ghting with the new instruction before the the automation. But so aircraft has already passed the desired Long before reaching the top of de- would have more care- level-off altitude by 2,000 ft. scent point, the crew has entered the ful attention to their ac- designated arrival route (STAR) in the tual path and location.  An aircraft in motion has a lot of mo- fl ight computer and has (in line with mentum and high inertia. The faster company policy) selected the appro- it moves, the harder it is to change its priate autopilot mode (LNAV), which direction and the more time, space, is now taking the aircraft through the Loukia and distance it requires to change or designated waypoints in compliance arrest its movement. It is also possible with the necessary path and speed Loukopoulos that during steep climbs and descents, restrictions. The only piece of informa- has a PhD in Cognitive Psychology and an the controller’s display of the aircraft tion that is still missing is the tail end of Aerospace Experimental Psychology altitude could be misleading because the arrival, the approach and landing designation from the United States Navy the altitude encoder on the aircraft’s information. As soon as the control- where she served before joining NASA Ames’ transponder lags behind the actual ler provides that, the pilot monitoring Human Systems Integration Division. altitude. attempts (in line with company procedure) to enter that information in She is currently also a human factors consul- the aircraft’s fl ight management computer, so that the aircraft automation tant to the Hellenic Air Accident Investigation and Aviation Safety Board and involved in a number of aviation human factors research and teaching activities.

HindSight 16 Winter 2012 51 FROM THE BRIEFING ROOM

close interactions of the disconnected kind 'cont'd)

Many airlines now require their pilots craft require the initial selection of a whether the crew is using the aircraft to engage the automation to the full great deal more thrust than the crew automation to guide the aircraft, and and to make all adjustments to fl ight is expecting (or used to). At busy air- if so, at what level. Diff erent air carriers path via the fl ight management com- ports, where the controllers work hard have diff erent policies regarding use puter. Yet, although its proper use can to sequence arrivals and departures of automation, and diff erent pilots, to lead to an accurate and effi cient fl ight, carefully, a single go-around can mess the extent that they can exercise dis- its programming can present the crew up many good plans. This may be an cretion, have diff erent preferences for with substantial workload. unanticipated consequence of the ef- when and how to use the automation. fects on the pilots of late changes or it Diff erent aircraft, even of the same might be a direct and almost inevita- model, “wear” diff erent technologies Case 3. captain (pilot fl ying) re- ble consequence of an ill-judged back so controllers cannot know whether porting: “We were cleared for [the] taxi clearance in the face of traffi c on every aircraft of the same type neces- approach and to land on runway short fi nal. sarily has, say, a moving map display 21... Citation tra c ahead touched on board. Controllers have a gener- down and was asked to hold short It is not the point of these selected ally good understanding (mostly built of [the] south runway for departing cases to say that controllers or pilots through experience) about diff erences ight... Tower then … instructed the make mistakes (which they, like all hu- in descent capabilities of, say a Boeing Citation to back taxi on runway 21 mans, undoubtedly make). The point 737 versus an Airbus A320 - as a func- [so as] to turn o at one of the taxi- is to illustrate that, to a certain de- tion of aircraft design - but lack more ways that they had already passed. gree, the intimacy of the relationship in-depth knowledge about speed, al- At this point we were 300 feet above between controllers and pilots can titude, or other criteria dictated by air the ground within a mile to touch- also be accompanied by a paradoxi- carrier policies that would aff ect the down… We were in the process of cal disconnect. This disconnect stems details of how the particular aircraft is beginning a go-around when Tower from the fact that controllers don’t al- fl own and what instructions the crew instructed us to cancel landing and ways know enough about aircraft and can comply with. climb to 4,000 feet.” pilots’ capabilities and limitations, or about the demands and constraints of As a result, whether it is the outcome [ASRS 885498, April 2010, iFR on the cockpit as an operational environ- of a sincere intention to help (expe- initial approach] ment. Other than its make and model, dite traffi c, assign a requested runway, what else does the controller really etc.), or of an intense focus on the ulti- know about the fl ight capabilities of mate goal of managing complex traffi c The fi nal approach to land phase is the aircraft s/he is controlling? Other fl ows from their radar scope, control- a busy time for pilots even if all goes than the airline’s name, what does the lers sometimes make judgments and according to plan. For many pilots, a controller know about the policies and calls that inadvertently introduce risk go-around, especially one that has procedures the pilots must comply to a fl ight. Drastic, unexpected chang- not been anticipated, is a potentially with? es in altitude level-off s, landing runway stressful time despite their simulator changes with little notice, late runway training. As a consequence, it is not The actions of the controllers (and assignments, and other such instruc- unknown for pilots to exceed speed their repercussions) in the cases cited tions on approach may introduce and altitude restrictions on a go- above have a direct relevance to these around, since many modern jet air- questions. These controllers cannot tell

52 considerable extra workload in highly The issue is not new. Recommenda- automated aircraft (reprogramming tions have been produced1 to help of the computer, reviewing charts, address certain aspects of this discon- Immanuel Barshi re-briefi ng, assuring compliance with nect, and the subject of stabilised air- Immanuel Barshi is a Senior stabilised approach criteria, conduct- craft approaches has been presented Principal Investigator in the ing checklists, etc.). Sometimes, an specifi cally for controllers2. There have Human Systems Integration Division early decision to perform a go-around also been eff orts to alleviate such dis- at NASA Ames Research Center in may be the best option. Leaving a go- connects through familiarisation pro- California, USA. His current research around until nearer the ground in the grammes that encourage controllers addresses cognitive issues involved hope that the necessary pilot response to ride the “jump seat” and gain a view in the skilled performance of astro- to late changes can be completed can of operations from “the other side.” nauts and pilots, as well as mission end up making the task more diffi cult, Such programmes were suspended in controllers and air traﬃ c controllers, especially if the fl ight crew rarely expe- the USA after the 2001 terrorist attacks their ability to manage challenging riences go-arounds. There is also the but they were due to be reinstated concern felt by some passengers as this year. Based on our own research, situations, and their vulnerability to the expected imminent touch-down we are passionate believers in focus- error. Dr. Barshi holds PhDs in Lin- suddenly changes to a steep climb. ing any type of training on real-world guistics and in Cognitive Psychology. operations. It would therefore be in- He holds an Airline Transport Pilot Pilots are frustrated when such inter- teresting to explore the actual level of certifi cate with A320, A330, B737, action disconnects occur. In their own participation in such familiarisation/ and CE500 Type Ratings; words: training programs, as well as to exam- he is also a certifi ed ine just how they are structured and fl ight instructor for Case 1: “My only thoughts as to what elements and means would be airplanes and helicop- how this happened … [was the] required to really acquaint controllers ters, with over 35 years controller’s failure to understand with aspects of operations that are of fl ight experience. [the] aircraft level-o capabilities.” critical to their jobs and that could alleviate the occasional disconnects.

Case 2: “ATC should give us more 1- Eﬀ ective Pilot / Controller Communications. Airbus Flight Operations Brieﬁ ng Notes. Available at time to program the correct runway http://www.skybrary.aero/bookshelf/books/172.pdf arrival.” 2- Stabilized Stabilised Approach Awareness Toolkit for ATC. Developed jointly by the Civil Air Navigation Services Organisation (CANSO), the Flight Safety Foundation, EUROCONTROL and Cotswold Airport. Available at http://www.skybrary.aero/index.php/Solutions:Stabilised_Approach_Awareness_Toolkit_for_ATC Case 3: “From our perspective, ATC failed to have adequate awareness of the tra c they were controlling. In the future, I think back taxis should only be allowed if there is no con icting landing tra c within 10 miles.”

HindSight 16 Winter 2012 53 FROM THE BRIEFING ROOM Unnecessary TcAS RAS caused by high vertical rates before level off

by Stanislaw drozdowski, Captain Wolfgang Starke and First offi cer Felix Gottwald

TCAS warns fl ight crews of an imminent risk of collision by generating resolution Advisories (rAs) to the fl ight crew. However, monitoring conducted in core european airspace in 2011 and 2012 shows that roughly three out of four TCAS rAs are in level off geometries1. The causal factor in most of these rAs is a high vertical rate of climb or descent by one of the aircraft involved during the last 1000 feet prior to level-off when the adjacent level is occupied. This is despite iCAo publishing a recommendation in november 2008 to reduce the vertical rate to 1500 ft/min in the above situations.

in the fi rst part of this article, TCAS expert and editor of eU- roConTroL’s ACAS Bulletins Stanislaw drozdowski explains why such rAs are generated. in the second part of the article, Wolfgang Starke and Felix Gottwald, both current commercial pilots and members of the German Air Line Pilots' Association Air Traffi c Services Committee give the pilots perspective.

54 1- Source: Analysis of downlinked RA messages for SESAR project 4.8.3 The TCAS experts point of view Unnecessary by Stanislaw drozdowski Stan is a Senior ATM Expert at EUROCONTROL HQ in Brussels, working in the area of ground and airborne safety nets. He The performance of modern aircraft allows pilots to climb focuses on current operational issues as well on enhance- and descend with high vertical rates. While this can provide ments to existing systems. Before joining EUROCONTROL, operational benefi ts (i.e. fuel or time savings), it can become Stan worked as an ATM system engineer with Northrop TcAS RAS problematic when aircraft continue to climb/descend with Grumman in Baltimore (USA) and as an en-route air traffi c a high vertical rate close to their cleared level when the ad- controller in Poland and New Zealand. jacent level is occupied or another aircraft is descending/ caused by high climbing towards the adjacent level. TCAS RAs due to high vertical rates Always follow the RA before level-off vertical rates Pilots and controllers sometimes judge these RAs as opera- TCAS will issue an RA when it calculates a risk of collision tionally not required and refer to them as “nuisance” RAs. based on the closing speed and vertical rates. A high verti- However, in real time the pilot cannot (and should not) as- cal rate before level-off may cause the TCAS logic to predict sess whether the RA is in fact operationally required. Once before level off a confl ict with another aircraft even when appropriate ATC an RA has been issued it must be followed without delay instructions are being correctly followed by each crew. This and it takes precedence over any ATC instructions. is because TCAS does not know aircraft intentions – autopilot or fl ight management system inputs are not taken into Therefore, it is best to avoid approaching the cleared level account because TCAS must remain an independent safety with a high vertical rate when the pilot is aware of another net. aircraft at the adjacent level – based on ATC traffi c information, observation on the TCAS traffi c display or as a result If, simultaneously, another aircraft is approaching an adja- of a Traffi c Advisory (TA). In this way the occurrence of nui- cent level, the combined vertical rates make RAs even more sance RAs can be minimised. likely. The majority of all RAs occur within 2000 feet before level-off at the cleared level. TCAS will typically generate: icAo recommendation an “Adjust vertical speed, adjust” RA (in version 7.0) which requires a reduction of the vertical rate as indicated on the in order to reduce the number of RAs caused by high fl ight instruments; or a “Level off , level off ” RA (in version 7.1) vertical rates before level-off , icAo in november 2008 which requires a reduction of the vertical rate to 0 ft/min published a provision recommending the reduction of (i.e. a level-off ). In extreme cases, involving very high verti- vertical rates to 1500 ft/min or less in the last 1000 feet cal rates TCAS may diagnose that insuffi cient time remains before level-off , when the pilot is made aware of an- to assure safe separation by a reduction in vertical rate and other aircraft at or approaching an adjacent fl ight lev- instruct a crossing RA, announced in the cockpit as “Main- el, unless instructed by Atc to maintain a certain verti- tain vertical speed, crossing maintain”. cal rate. Some States have published or are considering publishing similar or even more restrictive measures to be applicable in their airspace.

The above is based on the text that fi rst appeared in Issue 15 of EUROCONTROL’s ACAS Bulletin. These bulletins discuss real-life TCAS events in order to spread the lessons learned and encourage best practice by pilots and controllers. Issue 15 is dedicated to Unnecessary RAs due to high vertical rates before level-oﬀ and can be accessed on the EUROCONTROL website and on SKYbrary.

http://www.skybrary.aero/bookshelf/books/1804.pdf 

HindSight 16 Winter 2012 55 FROM THE BRIEFING ROOM

Unnecessary TcAS RAS caused by high vertical rates before level off (cont'd)

56 Unnecessary TcAS RAS caused by high vertical rates before level off (cont'd)

The pilot point of view

by Wolfgang Starke and Wolfgang is a Bombardier Dash 8-Q400 line training Felix Gottwald captain with Air Berlin. He has previously fl own Boeing 737-NG and classic aircraft. Wolfgang is a member of the Air Traffi c Services and Airport and Ground Environment Following on from the fi rst part of the article, why do pilots committees of the Vereinigung Cockpit (German Air Line sometimes only fl y shallow descents and sometimes ap- Pilots’ Association) and IFALPA (International Federation proach their cleared fl ight level with a high and sometimes of Air Line Pilots’ Associations). inappropriate vertical rate, knowing that there is traffi c separated by just 1000 feet? To answer these questions we need to understand how the climb and descent phases are Felix is an MD-11 fi rst offi cer with Lufthansa Cargo. fl own by pilots, and how the autopilot/fl ight director (AP/ He has previously fl own the A320 series aircraft. FD) systems on board the aircraft operate. Felix is a member of the Air Traffi c Services committee of the Vereinigung Cockpit (German Air Line Pilots’ Associa- Climb profi le tion) and IFALPA (International Federation of Air Line Pilots’ Associations). He is also an IFALPA representative to In modern aircraft fuel burn reduces signifi cantly with in- the IFATCA Technical Operations Committee. creasing altitude. At the same time, true airspeed increases and the speed of the aircraft over the ground is higher. In order to achieve the most economic fl ight profi le, the Standard Operating Procedures (SOPs) of most airlines re- Descent profi le quire pilots to apply full climb thrust while adjusting the indicated airspeed (IAS), or at higher altitudes the Mach Practically all aircraft now operating within European air- number, by increasing the pitch of the aircraft. This results space feature vertical navigation (VNAV) functions in their in aircraft always fl ying their maximum available climb rate fl ight management systems (FMS). These VNAV paths are at the optimum speed. calculated at certain airspeeds with the engine thrust at fl ight idle. This is a simplifi ed version of the climb case, n Asking pilots to maintain a low rate or reduce their engine thrust is kept constant while indicated airspeed is vertical speed will either lead to a reduction in engine adjusted by changes in the aircraft's pitch. thrust or higher airspeeds. Both situations result in a non-optimal fl ight profi le. Increasing the vertical speed of an aircraft during the descent when the engines are at idle leads to an increase in n Asking pilots to increase their vertical speed or main- airspeed. This is only possible up to the maximum allowed tain up to a higher fl ight level will require them to trade airspeed, although increasing the descent rate further can off an aircraft's indicated airspeed against vertical be achieved to a certain extent by using speed brakes - speed. Such a trade-off is possible for short-term ma- usually in the form of lift spoilers on the upper surface of noeuvring like following a TCAS RA, but the available the wings. But speed brakes also disturb the very sensitive climb rate after the manoeuvre will be signifi cantly re- aerodynamic properties of the wing, wasting a lot of the duced for a period longer than the manoeuvre itself. aircraft energy, energy that had previously been generat- Additionally, ATCOs are unaware of the speed of the ed by burning fuel. And of course, passenger comfort may aircraft relative to its minimum climb speed. be aff ected by the noise and airframe vibration which can occur with speed brake deployment. All of the above has one common consequence: higher and more economic cruise altitudes will be reached later In most instances, increasing the rate of descent causes into the fl ight, which increases the amount of fuel used an aircraft to be at a much lower altitude than planned at and the fl ight time. Nowadays pilots rarely carry signifi - a given distance from the destination airport. Similar to cant additional fuel for economic reasons so the increased the climb, the consequence will again be prolonged fl ight fuel burn at lower altitudes could limit their options later times, increased fuel burn and higher operating costs. Ad- in the fl ight. Longer fl ights also mean higher cost for main- ditionally, both the risk of having a bird strike and noise tenance, air crew salaries etc. abatement issues increase at lower altitudes. 

HindSight 16 Winter 2012 57 FROM THE BRIEFING ROOM

The pilot point of view (cont'd)

Nevertheless, increasing descent rate is possible most of tioned to level flight too quickly, autopilots have an altitude the time, but pilots may have difficulty making speed re- capture or altitude acquire mode. This mode is based on a ductions if required to maintain high rates of descent whilst predetermined g-load, typically in the range of 0.25 to 0.3g, doing so. during level off. The point relative to the level-off at which the altitude acquire mode starts to operate is dependent on Timing the start of the descent the vertical rate of the aircraft. At low vertical rates, such as 1000 ft/min, the altitude acquire mode will typically start to As already mentioned, the aircraft's descent profile is reduce the vertical rate between 300 and 100 feet before planned using idle thrust. Therefore it is not always pos- level-off. At very high vertical rates, such as 6000 ft/min, the sible to increase the rate of descent to the rate requested by reduction in the vertical rate should be expected to start ATC. If ATC keeps an aircraft at high altitudes and the pilots more than 1000 feet before level off. are not allowed to initiate the descent in time, or are only instructed to do so at a low vertical rate, the descent rate Generally modes such as the altitude acquire work well; required for the remaining track miles to the destination will nevertheless they do have one significant safety issue in become more and more problematic. This can result in one common. Whenever pilots try to modify their aircraft flight of two scenarios; the pilots are required to descend at the path once altitude acquire mode is active, it is likely to drop maximum rate, or they are unable to descend without flying out if already engaged, which in turn may well lead to a lev- additional track miles. In the later scenario, on top of fuel el bust. To prevent aircraft from overshooting their cleared burn and flight time penalties, the additional track miles will flight level, many airline SOPs restrict the use of VS mode be flown closer to the ground and near an airport, so the near to level-off. These SOPs render pilots unable to reduce margin for error is reduced and the chance of receiving a a high vertical rate whilst keeping the automatics engaged. TCAS RA is increased! Although IFALPA (International Federation of Airline Pilots' Associations) is calling for a consequent improvement in A good question is why do pilots not programme a little the design of modern autopilots, this change has not been “reserve” into their FMS to cater for unexpected ATC con- made yet. straints? To be honest, operationally this would be a good idea. However, as descents work well most of the time, plan- A version of this specific to Airbus aircraft is the so-called ning to be at lower altitudes earlier than necessary would “Alt. Star trap” (a star symbolises the altitude capture mode be less efficient than descending at high rates or even ad- in Airbus-speak). Once the autoflight system has captured ditional track miles from time to time. the level-off altitude, it maintains the climb rate and there is no way for the pilot to change the aircraft's mode to re- Level-off procedure – influence duce the vertical rate – the crew are “trapped” in the altitude of the autopilot/flight director capture mode. One possibility is flying the aircraft manually, which is hardly ever practiced during the en-route stages of ICAO recommends that the vertical rate be reduced to not a flight. The other is to select a new altitude, change to verti- more than 1500 feet per minute during the last 1000 feet cal speed mode, reselecting the original altitude and hope altitude prior level-off. If followed, this would reduce the all goes well within the few seconds before the Autopilot number of unnecessary RAs during level-off quite signifi- overshoots the cleared level. But such non-standard actions cantly. But not all operators have yet incorporated this guid- increase the likelihood of errors and reduce safety margins. ance into their SOPs. So why do pilots not always follow this recommendation? Pilot and controller errors Part of the answer lies with today's autopilot/flight direc- Of course, it is not always the system design which makes tor (APFD) design. These systems fly the aircraft in the most things worse; sometimes pilots just forget to reduce the ver- economical manner and, by doing so, will use the maximum tical rate. But imagine the workload when approaching busy available climb rate for a given speed. However, as a flight terminal areas like those around London, Paris or Frankfurt. would be highly uncomfortable if a high rate of climb transi- Often pilots have to focus on other things, knowing that the

58 TCAP – A solution to reduce nuisance rAs?

In an eff ort to reduce the number of unnecessary preven- tive TCAS RAs to ‘Adjust Vertical Speed’ during the ap- autopilot capture mode is not perfect, but still will capture proach to level off , Airbus have developed the TCAS Alert the cleared level. And there are also occasions where the Prevention (TCAP) system for Airbus A380 and A350 air- controller instructs a vertical rate RA. A year ago, one of us craft. travelled on the fl ight deck jump seat after a duty. When climbing through approximately FL210, ATC informed us about a company Boeing 737 levelling off 1000 feet above Recent data indicates that between 50 and 75 per cent of our cleared level of FL220. At the same time TCAS fi rst gave all such ‘nuisance’ RAs are caused by high vertical rates in a traffi c advisory (TA), quickly followed by an RA instruct- geometries when one or sometimes two proximate air- ing us to descend. At the time climb rate was about 1000 ft/ craft are about to level off . To reduce these RAs, ICAO rec- min. We heard later that the company 737 received an RA ommends reducing the vertical speed to less than 1500 to reduce vertical speed (an “Adjust vertical speed, adjust” fpm during the last 1000 feet prior to level off . However, RA) – a high rate being fl own at the specifi c request of ATC!. even up to date autofl ight systems fail to achieve this recommendation. So prior to level off , pilots may (if permit- Communicating each other's intentions ted by their SOPs) decide to change their fl ight guidance mode to manually reduce vertical speed. From a safety Of course, mistakes do happen again and again and we will point of view this is potentially hazardous, as it can in- never able to change that. Yet, in aviation we have to fi nd crease the chances of overshooting the cleared level. ways to mitigate the risks! Controllers and pilots are both experts in our very special jobs - pilots should not try to TCAP is an enhancement of the autofl ight system which make adjustments to traffi c separation, nor should ATCOs ensures a reduction of vertical speed prior to level off to try to “fl y airplanes”. In our view, if a modern aircraft needs prevent these TCAS from nuisance RAs. TCAS itself is not to be at a certain position at a certain airspeed or altitude, then it is better for a clearance to state exactly this require- changed by TCAP because it is merely an enhancement to ment rather than make the sort of aircraft control request - the autofl ight system. TCAP prerequisites are necessary to such as “descend xxx with a rate of yyyy ft/min” which might allow TCAP to automatically reduce the vertical speed af- suit older aircraft better. A clearance pilots love to hear is ter receiving a TCAS TA which has resulted from a proximi- “be there at this altitude with that speed” because then we ty which may subsequently lead to an unwanted TCAS RA. can do our job by manoeuvring our aeroplane effi ciently to achieve that objective. Besides, this is current ICAO provi- Pilots who have used TCAP certainly appreciate it, but sion of PANS-ATM. TCAP can only be seen as a fi rst step to improve modern autofl ight systems. In fact, IFALPA have had a policy since If as a controller you still need to instruct a defi ned verti- 2010 that pilots should not be required to interfere with cal rate for separation, please always stick to ICAO and tell the normal autofl ight system process for achieving level us when the specifi ed vertical speed is no longer required. capture to prevent unwanted TCAS RAs as a result of an Otherwise we are not able to plan ahead, which is a prereq- excessive vertical speed. uisite for safe fl ying. Whilst the possible disruption to air traffi c control will be Teamwork for ATC and pilots is best accomplished when reduced as the numbers of nuisance RAs decrease, the both parties know each other's plan. For this we have to real solution is for all autofl ight systems to reduce the keep up proper communications and make sure our respec- vertical speed as level off is approached to a value which tive intentions are known. This allows everyone to do their job well. does not even produce a TCAS TA. In other words, the system should be compatible with the ICAO recommendation and automatically reduce vertical speed to 1500 ft/ ediToriAL noTe min or less during the last 1000 feet prior to level off . See HS12 (Winter 2011) for more detail on this: http://www.skybrary.aero/bookshelf/books/1417.pdf

HindSight 16 Winter 2012 59 FROM THE BRIEFING ROOM “So what’s it doing now?” Training in new technology environments

by Anne isaac and david Lord “Fifteen foot banks of identical switches with small code numbers displayed in a nuclear power plant, sophisticated military aircraft that are so expensive to operate the pilots rarely fl y them and ships that collide while the offi cers are observing each other on anti-collision radar all suggest human-machine problems in high technology systems”

Defi ning the problem another design would have resulted in better, more economical and safer performance. Engineers typically design machines according to engineering principles, rather The assumption underly- than behavioural principles. That is, when ing the implementation of a machine has been designed, relatively new technologies, such little consideration may have been given as electronic fl ight data to how easy it is for a person to use or op- systems, is that with the erate. Life is fi lled with such examples. In automation of functions some cases, the poor design is a nuisance which were once allocated and not particularly dangerous, such as to human control, the pro- the size of door handles or the place- cessing resources of the opment of spare tyres in cars that require erator, their ‘spare capacity’, you to unpack the entire boot to access. will be freed to deal more ef- In others, the design is positively hazard- fectively with other required ous, such as the placement in some new tasks. However, while the use control consoles of the switch to ‘amend of new technologies may be fl ight information’ next to the ‘screen essential in order to deal with shutdown’ button. the ever increasing information processing demands of the avia- There are several reasons why engineer- tion system, the long-term per- ing principles dominate the design fi eld. formance implications of extended Many engineers who have not had the use of the new technologies on hu- benefi t of human factors training believe man information processing and performance are that humans are able to adapt readily to largely unknown. There is a possibility that new almost any environment in which they are technologies, intended to reduce workload and placed; whether a work place is too hot or consequently enhance memory, will undermine cold, too quiet or noisy, humans manage situation awareness and safe aviation practices. to perform their work. The real question, There is a need for the task demands of the aviation system however, is whether they perform these to keep the operators alert and actively involved in mean- jobs adequately and safely, and whether ingful ways.

60 Operational complexity versus functional capability Anne Isaac leads the Human Performance development work in the issues for pilots: the further diffi culties of co-ordinating the new technolo- pilot/ controller interface in NATS, UK. gies, and human capabilities, between the fl ight-deck and She gained her PhD in Cognitive Neuropsychology at air traffi c control environments? Otago University in New Zealand. Her previous work has been in the development of incident investigation There are some signifi cant diff erences between the abili- tools and techniques in European ATM, ties of machines and the skills, abilities and traits of hu- the introduction of TRM into the ATC mans, and it is crucial in all high-risk environments in environment and the introduction of which new technologies are introduced, to research carefully how these two very capable ‘systems’ work together Day to Day Safety Surveys techniques optimally. These principles of human-centered automa- into NATS. She has written several tion (Billings, 1991) advocate the design of automated book chapters, academic papers and systems with the human operator at the centre, rather the book Air Traﬃ c Control: than trying to exclude them. However, it can be seen that the human performance factors. many new technological systems are often designed and installed only with regard for the operator’s ergonomic requirements with no consideration from a systemic or operational stand point.

To consider human factors properly at the design stage is costly, but the cost is paid only once. If the operator must compensate for incorrect design in his training programme, the price must be paid every day. And what is worse, we can never be sure that when the chips are down the correct response will be made. Or, as Rudyard Kipling put it: But remember please, the law by which we live, David Lord fl ew a variety We are not built to comprehend a lie, of aircraft in the Royal Navy as a We can neither love, nor pity nor forgive – pilot, fl ying instructor and If you make a slip in handling us, examiner. His operational You die. experience included Northern Ireland, the Falklands The Secrets of Machines War, Bosnia and Haiti, the latter whilst on an exchange tour with the US Coast Guard. He commanded the Commando Helicopter Force at RNAS Yeovilton and issues for both teams/crews: Until pilots and controllers are provided with eff ective completed 3 seasons as Swordfi sh display pilot. After means of mastering the automation and technology in leaving the Fleet Air Arm he fl ew B737s with British their workplace, with training processes and operating Midland (bmi) for 4 years and then joined Aviation Train- protocols that ensure survival, the successful reversion ing International as the Apache program Chief Aircrew to manual core skills and standardised protocols for the Instructor at Middle Wallop. David joined Flight Safety avoidance or mitigation of technology-related errors and International in 2007 and was appointed as the Deputy hazards, we are likely to go on hearing the phrase… Head of Training in 2011. “So what’s it doing now?” 

HindSight 16 Winter 2012 61 FROM THE BRIEFING ROOM

It is therefore essential for both ANSPs and airlines which introduce new technology, to follow Billings’ principles of automation. The three areas which need constant and comprehensive consideration are the selection of the right technology, the development of appropriate procedures and the selection of the most appropriate training.

n Selection of appropriate human/machine technology – it is essential that an automation philosophy, policy and guiding principles be developed to enhance the choices made when new technology is introduced and multi-disciplinary teams work together to detail the interface and operability of these advanced systems. It should also be realised that the more complex systems become, the less the operators will understand the linkages between the diff erent teams and their specialist functions.

n Development of appropriate and robust procedures – leg- acy procedures are often adopted when new technology is introduced, in the belief that the operational staff will behave in the same way. Technology usually brings at least two behavioural changes; fi rst, the operators will quickly adapt to the support which the technology brings, becoming less involved in thinking and intervention. Secondly the operators assume that the technology is always correct and become less ‘afraid’ of dealing with failures since they are rarely seen. Both these behaviours can lead to slow adaption and sometimes misuse of procedures which are often so subtle that the system itself does not recognise the degradation.

n The development of new and appropriate training materials and methods – automation requires behavioural modifi cation in the operating environment. These diff erent behaviours and problem-solving techniques must be identifi ed, possibly from the automation principles, and highlighted in the training at ab-initio level. There is also a need for both controllers and pilots (and in many cases engineers) to col- lectively share their learning in a collaborative approach to cross-disciplinary aviation training.

Dave: Open the pod bay doors, HAL. HAL: I'm sorry, Dave. I'm afraid I can't do that... Dave: HAL, I won't argue with you any more! Open the doors! HAL: I know you and Frank were planning to disconnect me. And that's something I cannot allow to happen...

Stanley Kubrick’s 2001: A Space Odyssey, 1968

62 Lost in by Jean Pariès Having reached a veteran’s age, may I indulge myself modernity with a personal memory?

In 1992, I was invited by Dan Maurino, then the head of ICAO's programme on Human Factors and Flight Safety, to give the closing address to the second World Symposium he was organizing in Washington DC. In these cases, one generally accepts with a fl ush of plea- sure, then bitterly regrets. So, I was suff ering the agony of the white page, when I remembered that the fi rst ICAO Human Factors symposium had been held two years earlier in a country that no longer existed (USSR), and in a city which had changed its name (Lenin- grad). I decided to talk about the challenges of change for safety. The world is changing at an impressive rate, I said in essence, so what will aviation look like in 10 or 20 years from now? What are the safety challenges we will have to meet? Is there a plane today that foreshadows this future? I was then immersed in the investigation into the crash of an Airbus A320 at Mont Saint Odile near Strasbourg, France. I bravely answered: I think this aircraft exists, it is the Airbus A320. I heard something like an off ended whisper run across the meeting room. Designating an aircraft which had raised so much con- troversy and had suff ered so many accidents in its introductory years as an archetype of the future was probably a bit provocative. Twenty years later, I believe it was a rather good guess. But anyway, the point I want to make here is that, whatever the answer, I un- knowingly asked the question which is underlying this issue of Hindsight: what is a "modern" aircraft? Is it possible to speak of "modern aircraft" as there are “modern times” or “modern art”? Doesn’t it simply mean the air- 

HindSight 16 Winter 2012 63 FROM THE BRIEFING ROOM

Lost in modernity (cont'd)

craft which have recently left the as- sembly line? Well, obviously, some have an entirely new design, some are evolutions from older models, while they may benefit a "new" cockpit or newer engines. Which are "modern"?

According to the Merriam-Webster dictionary, “modern” means either “contemporary” or “up-to-date”, i.e. involving recent techniques, methods or ideas. I think modernity also often includes the dimension of a disputed change from what existed before (a quarrel between the ancients and the moderns..). But what is new is not necessarily modern. Modernity further implies a "sustainable change", that is to say, a change that sets the path of future changes, and defines the general trend for a new way of doing things. So we move from one modern the entire knowledge of the world is likely evolution, and match it to their time to the next one, from one “age” to theoretically accessible to virtually ev- model of the environment. They will the next one, in different time scales. eryone. And Michel Serres goes on: it fly, navigate, and communicate inten- Where will our current modernity take would be a waste of time and energy tions and trajectory forecasts with an us? Let's take a step back. A big step: to try and keep that knowledge inside accuracy of just a few seconds. They let's look at things at the scale of the our students’ brains. Sooner or later, it will define and negotiate with their history of mankind. A whole series of will be lost, as were lost, all along man- "colleague" aircraft, and with what will revolutions triggered transitions from kind’s history, all those skills suddenly stay as the ground-based component one age to the next, and changed rendered useless by the correspond- of the traffic management system, the our relationship to the world: carved ing socio-technical revolution. stone, fire, agriculture, bronze, iron ... According to the philosopher of sci- What if we apply this kind of vision to ences Michel Serres, three of them aviation? I guess a first outcome is that were even more important – writing, what defines an aircraft’s modernity printing, and computers – because is its cockpit, because it is where the they have changed our relationship to handling of knowledge (cognition) knowledge. Writing made it possible lies. A second outcome is that mod- to archive knowledge outside human ern aircraft are potentially connected memories (to outsource long-term to all the knowledge in the world. memory) and to access it without the And they will use it to calculate pres- constraints and fragility of oral trans- ent and future actions, and execute mission. With printing, it became pos- them. They know, or will soon know, sible to provide a mass access to that the weather, the traffic, airport acces- external memory, without needing to sibility, the price of fuel. They will in- cross the closed door of a few privi- corporate a complete digital simulator leged libraries. With computers, both of themselves, and know their internal long-term memory and the central status through omnipresent detectors computation unit were outsourced. and monitors. They will know their And with Internet and other networks, performance limitations, and their

64 Jean Pariès graduated from the French National School of Civil Aviation as engineer, best navigation trade-off between generation of aircraft fl ies, as long as and then joined the DGAC safety, fuel effi ciency, weather, envi- autopilots fail and disconnect, or do for several positions deal- ronment, and passenger comfort. So surprising things, there will be a need ing with air safety regulations. what will be left to human operators: for pilots with manual skills enabling pilots, controllers? them to back-up. And since Lisanne He was a member of the ICAO Human Bainbridge’s “ironies of automation” Factors & Flight Safety Study Group from “What is left to the human brain?” asks in the early 80’s, we have known that its creation in 1988. In 1990, he joined the Michel Serres. He answers: creativity, maintaining those skills is both a need Bureau Enquêtes Accident as Deputy Head, imagination, serendipity, ethics. Let’s and a real challenge, as they are atro- and Head of Investigations, where he led the translate this into operator language: phied day after day like unused mus- technical investigation into the Mont Saint- sense-making, adaptability, judgment, cles, with pilots watching when things Odile Accident, 1992. Currently Jean is CEO - common sense, airmanship, survival go well and suddenly required to fl y of Dédale SA. instinct. Is it enough to save human when things go wrong. jobs in cockpits or control rooms? It's He holds a Commercial Pilot Licence with more than enough: it is essential! Be- But in the longer term, most “manual Instrument, Multi-engines, Turboprop, and cause there is something the analyti- fl ight” skills will inexorably be lost. Instructor ratings and a Helicopter Private cal computation of “intelligent” com- Lost in modernity. The next genera- Pilot Licence. puters will not, for still a long while, tion of “modern” aircraft will probably be able to cope with: the unexpected, be “fl y-by-autopilot” only. The issue the irreducible uncertainty and unpre- will not be manual skills, but automa- sponses (automation is only the ulti- dictability of a fl ight, of thousands of tion reliability: a failure of the “perma- mate form of predetermination). This fl ights interacting within a worldwide nent autopilot” will not be an option strategy makes the system more and network. Airport delays, blocked run- anymore. Nevertheless, the next gen- more reliable within its envelope of ways, fl ocks of birds and other kinds eration will share with the current one designed-for uncertainties, and more of fl ying objects, unprecedented an extended version of the “ironies and more brittle outside it. The com- combinations of failures, passenger of automation”. I call it the “ironies petencies needed to cope with the emergencies, volcanoes, wars, ter- of predetermination”. The “modern” unexpected at the front line are lost rorist attacks, and so on. But it means safety strategy seeks the anticipation in this continuous eff ort to eradicate one should not fi ght the wrong battle. of all potential threats, and the pre- surprises. There is no “fundamental Needless to say, as long as the current determination of all the needed re- surprise” in the simulator, only listed emergencies. But the real world is ir- reducibly unpredictable, and safety strategies should rather get people both prepared… and prepared to be unprepared! Front-line operators should be trained to cope with the unexpected. Human-machine coop- eration should be revisited in the next (cockpit, control room) generation to better support human operators in their fundamental role: managing the unexpected, managing uncertainty, making judgments and decisions. They should be provided with a clear display of their current position within the operational envelope, as well as of their margins for maneuver. A paradigm shift is needed. There may even be a word for it: resilience engineering.

HindSight 16 Winter 2012 65 FROM THE BRIEFING ROOM Aircraft automation made simple by Michel Tremaud

it is very useful for air traffi c controllers to have some Understanding Automation understanding of the pilot’s working environment. This includes the fundamentals of aircraft automa- The design objective of an automatic fl ight system (AFS) is to tion (understood in this article as automatic fl ight provide assistance to the crew throughout the fl ight (within guidance), how pilots interface with automated sys- the normal fl ight envelope and with normal operation of all tems and how optimal use of automation can con- the systems it depends on), by: tribute to the overall management of the aircraft fl ight path. n relieving the pilot-fl ying (PF) from routine handling tasks and thus allowing time and resources to enhance his/her Although the extent to which automation is used has situational awareness or for problem-solving tasks; or, grown a great deal over the past 50 years, and many n providing the PF with attitude and fl ight path guidance diff erent levels of systems integration and automa- through the fl ight director (FD), for hand fl ying. tion remain in the skies today, the guiding principles which underlie automation have remained essential- Basically, the AFS provides guidance to capture and mainly the same. tain the selected targets and the defi ned fl ight path, in accordance with the modes engaged and the targets set by Providing an aircraft is functioning normally, the the fl ight crew on the fl ight guidance control panel (usually high levels of automation which may be available referred to as the fl ight control unit – FCU – or mode control are able to provide pilots with an increasing number panel – MCP) or on the fl ight management system control of solutions to the task which they must accomplish, and display unit (FMS CDU). such as complying with ATC requirements. When seeking an understanding of any automated system, i will try and describe some of the fundamental as- but particularly the AFS and FMS, it helps if the following pects of the use and monitoring of automation. de- questions are considered: signers of automated aircraft systems envisage that strict adherence to the following guiding principles n How is the system designed? and golden rules of operation will enhance pilot n How does the system interface and communicate with situational awareness and prevent so called "auto- the pilot? mation surprises". of course this model has humans n How can the system be operated in normal and abnormal in charge and so controllers should recognise situations? that pilot mismanagement of automation at times is unfortunately a fact – and the evi- The following are both important for optimal use of automatic dence suggests that the risk of this is direct- fl ight guidance: ly proportional to the complexity of the automation or procedure involved! n The integration of autopilot/fl ight director (AP/FD) and autothrottle/autothrust (A/THR) modes (i.e. the pairing of modes); n Mode transition sequences; and, n Pilot-system interfaces for both:

− Pilot-to-system communication (i.e. for selecting guid- Michel Trémaud retired from Airbus as senior director and ance targets and arming/engaging AP/FD - A/THR head of safety programs / initiatives. His career also included positions modes); and, with the French Bureau Veritas, Air Martinique and Aérotour. Beyond − System-to-pilot feedback (i.e. for checking the status retirement, Michel has continued to support safety initiatives led by of modes armed or engaged and the correctness of EUROCONTROL and by the Flight Safety Foundation. active guidance targets).

66 Aircraft automation made simple

AP - A/tHR integration Flight crew/system interface

Integrated AP - A/THR systems feature an association (pair- The FCU constitutes the main interface between the pilot ing) of AP pitch modes (elevator or stabiliser control) and and the autofl ight system for short term guidance (i.e. for A/THR modes (throttle or thrust levers). immediate guidance).

An integrated AP - A/THR operates in the exact same way as a human pilot:

n The elevator/stabiliser is used to control pitch attitude, airspeed, vertical speed, altitude, fl ight- path-angle, vertical navigation profi le or to capture and track a glide slope beam. Figure 2 - A340 FCU (typical) n The throttle/thrust levers are used to maintain a given thrust or a given airspeed. The FMS CDU constitutes the main interface between the pilot and the autofl ight system for long-term guidance (i.e. Indeed, throughout the fl ight, the pilot’s objective can be for the current and subsequent fl ight phases). seen as to fl y either:

n performance segments at constant thrust/power (e.g., take-off , climb or descent); or n trajectory segments at constant speed (e.g., cruise or approach).

Depending on the task to be accomplished, maintaining the airspeed is assigned – automatically – either to the AP (elevators) or to the A/THR (throttles levers/thrust control), as shown in Figure 1.

A/tHR AP / Fd

throttles/ elevators thrust levers

Performance Given thrust Speed Segment or idle Figure 3 - A340 FMS CDU (typical) trajectory Vertical Segment speed Speed Altitude When performing an action on the FCU or FMS CDU to give Vertical a command to the AFS, the pilot has an expectation of the proﬁ le aircraft reaction and, therefore, must bear in mind the fol- Glide slope lowing questions:

n What do I want the aircraft to fl y now? Figure 1 - The AP / FD – A/THR Modes Pairing n What do I want the aircraft to fl y next? 

HindSight 16 Winter 2012 67 FROM THE BRIEFING ROOM

Aircraft automation made simple (cont'd)

User Strategy This implies an awareness and understanding of the following aspects: The following principles should guide the operation and su- n Which mode did I engage and which target did I set for the aircraft pervision of automation so that pilots can stay ahead of the to fl y now? aircraft and be prepared for possible contingencies. n Is the aircraft following the intended vertical and lateral fl ight path and targets? Taking advantage of automation to reduce workload n Which mode did I arm and which target did I preset for the aircraft The use of automated systems is intended to reduce work- to fl y next? load and signifi cantly improve the time pilots need to respond to unanticipated changes such as ATC instructions or To answer these questions, the roles of the following controls and adverse weather conditions. Some unplanned, abnormal or displays need to be understood: emergency conditions can also be best dealt with by the use of automation – but controllers should be aware that signifi - n FCU (mode selection-keys, target-setting knobs and display win- cant failures often lead to a reduction in the extent to which dows); automation can assist. n FMS CDU (keyboard, line-select keys, display pages and messages); n FMA (Flight Mode Annunciator) on PFD; and, Most aircraft operators expect both AP and A/THR to be rou- n PFD and ND (Navigation Display) displays and scales (i.e., for cross- tinely engaged, especially in marginal weather conditions or checking active guidance targets). when operating into an unfamiliar airport.

Using the AP and the A/THR enables pilots to pay more attention to ATC communications and enhances their overall situational awareness, particularly in congested terminal areas and busy airport environments.

Of course, pilots need to maintain their manual fl ying skills too because they never know when they might be required. Training sessions in the full fl ight simulator are important for this but, in appropriate circumstances, pilots can be expected to periodically elect to control the aircraft manually or per- Figure 4 - A340 PFd and nd (typical) haps to select a lower level of automation than they usually use, to maintain all round profi ciency. Eff ective monitoring of these controls and displays promotes and increases fl ight crew awareness of the available/active guidance for Using the correct level of automation for the task fl ight path and speed control. This includes: On the latest highly automated and integrated aircraft, several levels of automation are available to perform a given task. n modes (i.e. AP/FD modes being engaged or armed); and, n targets (i.e. altitude, speed or vertical speed or vertical navigation, The optimum level of automation depends on: heading or lateral navigation). n The task to be performed: − short-term task (i.e. tactical choice, short and head-up The safe use of the AP, A/THR and FMS needs a three-step approach: action(s) on FCU, immediate aircraft response); or, n Anticipate: − long-term task (i.e. strategic choice, longer and head- − Understand system operation and the results of any action. down action(s) on FMS CDU, longer term aircraft re- − Be aware of the modes being engaged or armed. sponse); − Understand mode transitions or reversions. n The fl ight phase: n execute: − departure; − Perform action(s) on FCU or on FMS CDU. − en-route climb/cruise/descent; n Confi rm: − terminal area; or, − Crosscheck and announce the eff ective arming or engagement − approach; and, of modes and the correctness of active guidance targets (on n The time available: FMA, PFD and/or ND scales and/or FMS CDU); − pre-planned selection or entry or ‘last-minute − Observe the aircraft response and resulting change’; trajectory. − normal selection or entry; or,

68 Once automated systems have been programmed, the pilot al- n supervision of the resulting AP/FD guidance and A/THR ways retains the ability to change the level of automation and operation on the PFD and ND (i.e. pitch attitude and bank guidance for the task. This might include: angle, speed and speed trend, altitude, vertical speed, heading or track …). n adopting a more direct level of automation by reverting from FMS-managed guidance to non-FMS guidance (i.e., us- Both pilots must always be aware of the status of the modes ing the FCU for modes selections and targets entries); armed or engaged and of the selected and active guidance n selecting a more appropriate lateral or vertical mode; or, targets. n reverting to hand flying (with or without FD guidance, with or without A/THR), for direct control of aircraft vertical trajec- Being ready and alert to take over, if required tory, lateral trajectory and thrust. Supervising automation can be summed up as simply “Flying with your eyes" - observing cockpit displays and indications to Ultimately, the optimal level of automation likely to be the one ensure that the aircraft response matches your mode selec- the pilots feel comfortable with for a task in the prevailing con- tions and guidance target entries, and that the aircraft attitude, ditions, which will be dependent on their knowledge and expe- speed and trajectory match your expectations. rience of the aircraft and its systems. If any doubt exists regarding the aircraft flight path or speed It has been noted that pilots with significant levels of experience control, pilots are encouraged to revert to a lower level or au- on an aircraft type tend to use automation in a simpler way than tomation rather than attempt to re-programme automated pilots who are recently qualified on a type, who tend to explore systems unless an obvious entry error is detected. higher levels of automation ... with the resulting risk of error or loss of mode awareness. Whilst there are routine ways to disconnect automation, if an AP or A/THR operation needs to be overridden following a mal- Being aware of available guidance at all times function such as a flight control runaway, pilots are expected The FCU and the FMS CDU are the prime interfaces for the flight to use the immediate disconnection methods provided. Only crew to communicate with the aircraft systems (i.e. to set targets in emergency situations is it expected that pilots will manually and arm or engage modes). override an engaged AP or A/THR.

The PFD and ND are the prime interfaces for the aircraft to com- Conclusions municate with the flight crew, to confirm that the aircraft systems have correctly accepted the mode selections and the target en- Automation should match the pilot's mental model for fly- tries: ing the aircraft from gate to gate and should therefore be intuitive. However, effective use of automation requires that n PFD (FMA, speed scale and altitude scale): proper pilot understanding of it is achieved through ground − guidance modes, speed and altitude targets; training followed by supervised use during line training and n ND ( heading / track scale or FMS flight plan): finally consolidation in normal operations. − lateral guidance. It should be acknowledged that automation may malfunc- Any action on the FCU or on the FMS CDU (keyboard and line- tion and that pilots may mismanage automation but it select keys) should be confirmed by cross-checking the corre- should equally be recognised that complex procedures (ap- sponding annunciation or target on the PFD and/or ND, and on proach or go-around/missed-approach) or challenging ATC the FMS CDU display. instructions may complicate the pilot task and his/her use of automation. The use and operation of the AFS must be evident to both pilots at all times by: As an air traffic controller, you may have an idea of the level of automation which a particular aircraft type provides but you n announcement of changes made or observed to the status of will not usually know the extent to which it is being used and AP/FD modes and A/THR mode on the FMA (i.e. mode arming how your instruction may affect this use. However, perhaps or engagement, mode changeovers); this article has opened your eyes to the scope and pilot use n announcement of the result of any change of guidance target of automation in the flight deck and prepared you a little to on the related PFD and/or ND scales; and, appreciate both its potential and its pitfalls for pilots.

HindSight 16 Winter 2012 69 SKYBRARY Skybrary download

If you need to fi nd out something about aviation safety, we suggest you go fi rst to www.skybrary.aero. It doesn’t matter whether you are a controller, a pilot or a maintenance engineer, SKYbrary aims to have either the answer you are looking for or a direct route to it.

Description

If by any chance you can’t fi nd what you want, please Thrust Reversers on jet aircraft provide a signifi cant way of increasing the remember that SKYbrary is a dynamic work-in- rate of deceleration during the initial stages of both a landing roll or a re- progress which needs continuous user feedback and jected a take off from high speed. benefi ts from user support. Be sure to tell the SKYbrary editor about any diffi culty you may have The following remarks are generic in nature and must, therefore, be had making it work for you. If you can directly help us considered in the context of instructions and guidance provided for by identifying material we could use or even fi ll a gap specifi c aircraft types by aircraft manufacturers and aircraft operators. by writing some content yourself then please tell us too! System Dependency on air/ground status The option of thrust reverser deployment on an airworthy aircraft depends We aim to provide wide coverage through both on whether the system has been signalled with ‘air’ status or ‘ground’ sta- original articles and, especially, by hosting the best of tus, the latter being a pre-requisite. Aircraft certifi cation requires multiple what’s already been written so that a wider audience defences against reverser deployment when ‘in fl ight’ but during the short can access it more easily in one place. period of transition between ‘air’ status and ‘ground’ status and between SKYbrary is also the place where you can access: ‘ground’ status and ‘air’ status, there can be both system use and system malfunction issues which are directly attributable to the status being signalled. n all the documents of the Flight Safety Foundation operator’s Guide to Human One example of the latter is a link by design between slat retraction and Factors in Aviation reverser unlocking. n the largest collection of selected offi cial Aircraft Runway Performance accident & serious incident reports from around 1. In both the landing roll and after a rejected takeoff decision, thrust revers- the world anywhere in one place online ers have the greatest eff ect when deployed whilst the aircraft is at high n an expanding facility to search iCAo document speed. This will correspond to the period when directional control is reli- text. ant on rudder inputs rather than nose wheel steering systems. - don’t delay deployment without a good reason (such as correct- An article taken from SKYbrary is reprinted in each ing runway alignment) HindSiGHT. For this issue, we have chosen “Thrust reversers : Flight Crew Guidance”. At fi rst sight of that 2. Depending on the regulatory system under which an aircraft is operated, title, and given that the majority of our readers are in the broadly speaking whether it is European or North American, an allow- ATM community, you might wonder why! Actually, as ance for the eff ect of thrust reverser deployment is likely to be respec- controllers in the VCR see rather a lot of aircraft making tively either included in or excluded from the runway performance data good use of their thrust reversers, our guess is that it will which fl ight crew are instructed to use. be interesting to learn a little about how they function in relation to other deceleration methods – and be able see - Be sure you are aware which assumption is made in the aircraft the sort of accidents and incidents in which their use, performance data you are required to use. abuse or un-commanded deployment has fi gured.

70 SKY

Thrust Reversers brary : Flight Crew Guidance

3. The relative benefi t of timely thrust reverser deployment is near- Rejected Landings ly always considerably less than the timely deployment of lift spoilers / ground spoilers / speed brakes. This is because of the In almost all cases, the activation of thrust reversers after touchdown way in which the increased pressure on the main landing gear will remove the option to reject the landing because the time neces- which they create enhances the eff ectiveness of braking. Note sary to regain eff ective thrust will use considerable runway distance. that contrary to the situation with thrust reversers, the eff ect of If such runway distance is available, it will almost always be more lift spoilers / ground spoilers / speed brakes is always included in eff ectively utilised in continuing with the attempt to stop. If it is not aircraft landing performance data. available, then other options to avoid a hazardous runway excursion - Whilst it is important to deploy thrust reversers promptly are likely to be preferable to an attempt to get airborne again. In any and check their correct activation, it is even more impor- case, many aircraft types are operated under a blanket prohibition tant to fi rst ensure that the lift spoilers / ground spoilers / on a go around once thrust reversers have been deployed. speed brakes have deployed correctly. Rejected Take Off s 4. When a landing is being carried out on a wet/slippery and/or Whilst the selection of reverse thrust will normally be part of the re- otherwise potentially limiting runway, the desire to achieve sponse to a decision to reject a take off at high speed, it should not a touchdown in the touchdown zone can sometimes result in be assumed that reverser deployment will necessarily occur if the initial runway contact which is even fi rmer than may have been reason for the reject decision is related to a possible or actual loss of intended. Whilst it is unlikely that a positive bounce will not re- airworthiness. In some situations, thrust reverser deployment might sult in the instinctive delay of reverser deployment, it is less well not be advisable anyway. known that the selection of reversers during a transitory unload- ing of landing gear after a fi rm touchdown can result in a cycling Thrust Reverser Unservicability between aircraft air and ground status before consistent ground When one or both thrust reversers have been identifi ed by mainte- status is achieved. Many reversers will lock out in transit if this nance as unserviceable, it is usually permissible to despatch under MEL happens and normal deployment with the aircraft in defi nitive relief. In the absence of MEL conditions which entirely prohibit the use ground status will not be possible until the selector has fi rst of any remaining serviceable thrust reverser(s), any such use should been returned to the reverser stowed position. be predicated on the existence of fl ight crew guidance provided or - Crew briefi ng for potentially challenging landings could endorsed by the aircraft operator. This guidance should be expressly usefully include reference to the need for reverser de- briefed prior to every landing with such status and then followed. ployment to occur without delay but only when lift spoiler / ground spoiler / speed brake deployment has been Use of Reverse Thrust confi rmed – after their manual deployment if necessary, Use of Reverse Thrust on low wing aircraft with mounted engines since this will act as confi rmation that sustained ‘ground’ should be limited to the time when the aircraft is on an active run- status has been achieved. way. Use of even idle reverse during runway exit and initial taxi in can result in engine damage due to ingestion of FOD or contamination of the air conditioning system with excess surface de-icing chemicals sometimes found on taxiways.

For Accidents and Incidents click here: http://www.skybrary.aero/index.php/Thrust_Reversers:_Flight_Crew_Guidance#Accidents_and_Incidents

HindSight 16 Winter 2012 71 If you are interested in downloading the entire HindSight collection: www.skybrary.aero

N°1 N°2 N°3 N°4 N°5 N°6 January 2005 January 2006 June 2006 January 2007 June 2007 January 2008

Putting Safety First in Air Traffic Management COMMUNICATION COLLISION HINDSIGHT IS A “Hindsight” The ability or opportunity to understand and judge AIRSPACE an event or experience after it has occured. AIRSPACE WONDERFUL THING CONFLICT AVOIDANCE “Hindsight” The ability or opportunity to understand and judge INFRINGEMENT an event or experience after it has occured. By Tzvetomir Blajev “Hindsight” DETECTION “Hindsight” DETECTION THE RUNWAY AND YOU Coordinator - Safety Improvement Initiatives, The ability or oportunity to understand and judge The ability or opportunity to understand and judge “Hindsight” and Editor in Chief of HindSight. an event or experience after it has occured. an event or experience after it has occured. The ability or opportunity to understand and judge WORKLOAD By Ian Wigmore an event or experience after it has occured. WORKLOAD “Hindsight” See page 22 “With the benefit of hindsight I would Win a free trip for two to Paris: See page 22 The ability or opportunity to understand and judge THE APPLICATION OF have done it differently”. See page 26 IS AIRSPACE PENETRATION an event or experience after it has occured. OFFSET TRACKS How often do we hear responsible people If we learn the right lessons we will stand with your colleagues - think what you AN ATC PROBLEM OR NOT? ANOTHER SUNNY DAY IN SWEDEN saying these words? Often,it is an attempt a much better chance of reacting correct- would do if you had a similar experience. IN EUROPEAN AIRSPACE to disguise the fact that they had not ly when we are faced with new situations We hope that you too will join in this See page 3 BY BENGT COLLIN prepared themselves for some unusual where a quick, correct decision is essen- information sharing experience. Let us WORKLOAD A STRANGE CONCEPT BY ROLAND RAWLINGS situation. Yet hindsight is a wonderful tial.This magazine is intended for you, the know about any unusual experiences See page 5 thing and can be of great benefit if used controller on the front line, to make you you have had – we promise to preserve BY PROFESSOR SIDNEY DEKKER See page 8 intelligently to prepare ourselves for the know of these lessons. It contains many your confidentiality if that is what you SeeSee pagepage 66 unexpected. There is much to be learnt examples of actual incidents which raise wish. Working together with the benefit from a study of other peoples’ actions - some interesting questions for discussion. of HindSight we can make a real contribu- good and bad. Read them carefully - talk about them tion to improved aviation safety. Front Line Report by Bert Ruitenberg The Briefing Room Editorial 121.5 - Safety Alerts Learning from Experience

Hindsight is a Wonderful Thing 1 Avoiding Action Phraseology 4 Runway Incursion 8 EUROCONTROL Safety New Clearance Related Loss of Separation 11 WORKLOAD VERSUS BOREDOM Enhancement Business Division 2 to Levels 4 THE HUMAN FACTOR COLUMN Controlled Flight Into Terrain 13 BY BERT RUITENBERG About HindSight 3 Aircraft Turn Performance LOSS OF SEPARATION BY PROFESSOR SIDNEY DEKKER BY BERT RUITENBERG Early Turns 5 Unauthorised Penetration See page 7 Acronyms used in This Issue 24 of Airspace 15 See page 7 See page 7 Undetected Simultaneous THE BLIND SPOT The Editorial Team 26 Transmission 5 Wake Vortex Turbulence 17 See page 15 AIRBUS AP/FD TCAS MODE: Contact Us 27 Hand-over/Take-over of Runway Excursion 19 Operational Position 6 INVESTIGATINGINVESTIGATING CONTROLLERCONTROLLER Disclaimer 28 Level Bust 21 NEAR COLLISION A NEW STEP TOWARDS Feedback 22 The Phonological WHAT? RUNWAY INCURSIONS - AT LOS ANGELES BLIND SPOTS SAFETY IMPROVEMENT SeeSee pagepage 2020 See page 25 IT WILL NEVER HAPPEN TO ME ... BY DR. BARRY KIRWAN By Paule Botargues SeeSee pagepage 1212 See page 18 See page 25

European Air Traffic Management - EATM

EUROCONTROL DAP/SAF January 2005 EUROCONTROL DAP/SAF January 2006 EUROCONTROL DAP/SAF June 2006 EUROCONTROL DAP/SSH January 2007 EUROCONTROL DAP/SSH June 2007 EUROCONTROL DAP/SSH January 2008

8: winteredition’09 “Hindsight” the ability or opportunity to understand and judge an event or experience after it has occurred.

EUROCONTROL HindThe ability or opportunity to understand and judge an event or experience after it has occuredight09 EUROCONTROL EUROCONTROL N°7 HindThe ability or opportunity to understand and judge an event or experience after it has occuredight10 HindThe ability or opportunity to understand and judge an event or experience after it has occuredight11 HindThe ability or opportunity to understand and judge an event or experience after it has occuredight12 July 2008 Hind ight WEATHER Are you responsible Level Bust... Airspace Infringement - Runway excursion “Hindsight” The ability or opportunity to understand and judge an event or experience after it has occured. Production for safety? or Altitude Deviation? again?! Putting Safety First A WEATHER Pressure Controllers and pilots teaming up in Air Trafﬁ c Management GHOST STORY to prevent runway excursions By Prof. Sidney Dekker Clashing moral values p08 by Captain Bill de Groh, IFALPA © European Organisation for Safety of Air Navigation See page 8 (EUROCONTROL) January 2009. By Professor Sidney Dekker

This publication has been prepared by the Safety Improve- NEW! CASEment Sub-GroupSTUDY: (SISG) of EUROCONTROL. The authors acknowledge the assistance given by many sources in the Next please p15 Some hidden dangers “THE FIRSTpreparation OFFICER of this publication. IS MY By Anthony F. Seychell of tailwind MOTHER-IN-LAW”The information contained herein may be copied in whole or in part, providing that the copyright notice and disclaimer by Gerard van Es are included. By Bengt Collin Safety & the cost killers p16 SeeSee pagepage 1616 The information contained in this document may not be To see or not to see modiﬁ ed without prior permission from EUROCONTROL. By Jean Paries 50 YEARS AFTER MUNICH A letter to aviation prosecutors Level Busts: cause or consequence? by Bert Ruitenberg The role of ATM in reducing The views expressed in this document are not50 necessarily YEARS AFTER MUNICH those of EUROCONTROL. See page 39 by Tzvetomir Blajev by Professor Sidney Decker the risk of runway excursion The consequences of p23 EUROCONTROL makes no warranty, either implied or ex- Let’s get rid of the bad pilots by Jim Burin pressed, for the information contained in this document; nei- commercial pressure can be fatal ther does it assume any legal liability or responsibility for the I separate therefore I am safe The ‘Other’ Level Busts by Professor Sidney Dekker accuracy completeness and usefulness of this information. By John Barrass by Bert Ruitenberg by Philip Marien GET YOUR R Airbus altitude capture enhancement SKYbrary POSTEPOSTERge See centre papage Air Traffic Controllers do it too! to prevent TCAS RAs Lesson from (the) Hudson by Loukia Loukopoulos by Paule Botargues by Jean Paries CND January ’09 Winter 2010 Summer 2010 * Piste - French, 1. (ski) track, 2. runway Winter 2011 Summer 2009 EUROCONTROL DAP/SSH July 2008

In the next issue of HindSight: Safety versus Cost

EUROCONTROL

Putting Safety First in Air Traffi c Management

This publication has been prepared by the Safety Improvement Sub-Group (SISG) of EUROCONTROL. The Editor in Chief acknowledges the assistance given by many sources in its preparation.

The information contained herein may be copied in whole or in part, providing that the Copyright is acknowledged and the disclaimer below is included. It may not be modifi ed without prior permission from EUROCONTROL.

diScLAimeR The views expressed in this document are not necessarily those of EUROCONTROL which makes no warranty, either implied or expressed, for the information contained in it and neither does it assume any legal liability or responsibility for its accuracy, completeness or usefulness.