By Dr. Sunjoo K. Advani
Dr. Sunjoo Advani is chairman of the International Committee for Aviation Training in Extended Evelopes (ICATEE) and member of the Flight Simulation Group of the Royal Aeronautical Society. His company, International Development of Technology, based in Breda, The Netherlands, is specialized in designing and integrating engineering and research simulators for flight, medical rehabilitation, and driving applications. His areas of work include training system specification, simulator motion and visual cueing, and human factors. Tel: +31 655 737 345; Email: s.advani@idt-engineering.com
[Editor’s Introduction: Dr. Advani was a speaker at the RAeS Annual International Flight Crew Training Conference in London during September 2011. Following his presentation, he and I discussed how becoming comfortable and competent controlling an aircraft really begins with the first flight. I mentioned that I vividly remember how uncomfortable I was as a very young student pilot when my instructor first said "we're going to practice stalls." At that point in my career, anything over 30 degrees of bank and 5 degrees nose up or down was something I considered an unusual attitude.
A few hundred hours later, the idea of max performing my aircraft by flying to the edge of a stall and even doing inverted vertical rejoins had become second nature. I flew a supersonic aircraft that had one very clear directive -- "if you enter a spin, eject" yet we routinely operated on the very edge of the envelope with our students.
How did my skills and confidence go from a basic fear of losing control to having the ability to operate comfortably in all attitudes of flight and a full range of speeds? In my case, it was accomplished through a series of well designed training exercises, each of which built on the previous lessons learned.
It has often been said that UPRT requires competent and experienced instructors but how do we get to that point? How do we break the current instructional bias towards teaching only what is required for a flight check? How do we get our young instructors to develop enough skill and confidence to actually teach their students the basics of flight control through their aircraft's entire envelope so that when the student moves on a foundation has been established?
Dr. Advani generously agreed to share with IAFTP members what the ICATEE is doing in this regard.]
Loss-of-Control In Flight (LOC-I) remains today the number-one cause of fatal commercial aviation accidents. Loss-of-Control is often the result of airplane upsets, such as stalls or over banking. These are areas of the flight regime that, if entered with today’s training, are likely to lead to unpredictable and fatal outcomes, even though the likelihood of getting into these flight conditions may very low. With current licensing training, pilots have limited competencies to recover from significant upsets.
As an industry concentrating on improving training, we ask ourselves, what could possibly be done to prevent the recurrence of such jet upset incidents? How can we teach proper avoidance strategies, and if necessary, recovery techniques? What constitutes effective upset prevention and recovery training?
The challenge is to find a common causal thread, and define an effective solution.
The Flight Simulation Group of the Royal Aeronautical Society agreed to look forward by creating the International Committee for Aviation Training in Extended Envelopes (ICATEE). It is now over seventy members strong, with airframe manufacturers, airlines, aviation authorities and safety boards, simulator manufacturers, training providers (including upset recovery specialists), research institutions, and pilot representatives.
An integrated approach to effective UPRT
ICATEE has concluded that the only way to defining training solutions is to first clearly delineate the training needs. The inflow of pilots with exposure to an all-attitude, all envelope flight environment is rapidly declining as the availability of airmen with prior military exposure or similar civilian experience is decreasing. Experience with high angles-of-attack, increased g-loading, rapid maneuvering and situations that could induce spatial disorientation can be of benefit to recovering from potential upsets in transport-category aircraft.
No “Silver Bullet” Solution
Today, we simply do not adequately train pilots consistently to recognize, avoid, prevent and recover from upsets . Practical test standards for stall training that have emphasized “minimum loss of altitude” (even when altitude loss is not a factor) may be leading to negative training transfer when recovering from these threats.
Upset prevention and recovery involves three distinct and critical levels of mitigation. First, there is awareness, which includes a thorough understanding of aerodynamics, and knowledge of airplane upset causes. Secondly, in the event a threat begins to emerge, the pilot must use recognition and avoidance techniques in order to stay clear of the threat, without further compounding the situation. However, distractions, failures, or other factors can still lead to an actual upset; hence, the final hope is to accomplish an effective recovery to bring the airplane back to a controlled state. In some situations with large transport aircraft, such as the incipient spin, or worse – a fully-developed spin, recovery may not even be possible, making recognition, avoidance and prevention the most critical training elements. To train all three, a combination of academics and practical skills development is essential.
Academics
For academics training, there exists today an industry-developed guide book called the Airplane Upset Recovery Training Aid. Unfortunately, it is not commonly integrated into training, and its use is not mandated. Furthermore, this aid is limited to large (100-plus passenger) swept-wing jets, whereas these problems also occur in regional aircraft, including turboprops.
Practical Training in Airplanes and Simulators
Aerobatic-capable aircraft, or specially-configured aircraft, can teach maneuvering skills and exposure to an all-attitude all-envelope environment. Familiarization with the limit loads and general flight dynamics can be very effective. The transfer of these skill sets to the multi-crew, glass cockpit, automation-enhanced environment takes place outside the actual in-flight lesson through a process of “differences training”. Flight simulators and academics can act as the bridge between basic skills, and application in the type-specific environment.
Clearly, it is not practical to teach UPRT in actual transport aircraft. Simulators (the logical replacement) are limited to the data acquired from the actual test aircraft. They cannot replicate the forces (g-effects) encountered in upset conditions, nor can they easily generate the sudden startle effect that occurs when a pilot discovers there is little time to apply life-saving control actions. The NTSB has identified several aircraft upset accidents in which inappropriate use of the simulator, and lack of instructor knowledge could have contributed to resulting accidents.
Simulator Enhancements
What can we do in today’s simulators and training curricula to better manage UPRT issues? Many factors have to be considered . Current flight simulator data can lead the pilot into believing that the aircraft remains docile and controllable at all times, even when it may not be. Enhanced data (derived from actual aircraft, or simply “representative”) can help pilots in becoming aware that, while initiating recovery of a stall, a roll control reversal, for example, or other effects could be encountered. Experts in ICATEE also agree that enhancements, albeit not even type-specific, may be a major step forward. Fortunately, with the involvement of research organizations like NASA, UTIAS and several academic experts, we will have the advantage of being able to extend existing models, and carry out more detailed analyses before drawing final conclusions.
Why Train in the Stall Zone?
Proposed FAA rule changes will require pilots to train avoidance and recovery from stalls, and are intended to contribute significantly to reducing LOC-I accidents. As stated in USA Today (11 May 2011), when referring to the Colgan Air 3407 crash, pilots are currently trained to avoid entering a stall at all costs but are never shown how to recover once they have entered a stall. “One key to preventing such accidents in the future will be more realistic training in simulators,” according to FAA Administrator Randy Babbitt.
Still, why would a pilot need training beyond the stall warning?. Several recent airplane accidents all occurred with the stick shaker activated, with plenty of opportunity to recover, and where timely stall intervention did not occur. Inattention, inadequate or improper use of automation were also cited as causal factors. Therefore, exposing pilots to the general nature of these conditions is absolutely essential.
The Psychology of Startle
What psychologists call “startle” can affect the pilot’s decision-making capabilities and narrow his/her reactions down to basic primal instincts; it is in these conditions that the rote learned or first-reflexes are applied. In some cases, the initial reaction has proven wrong, and the time available for corrective action is limited.
The Instructor’s Crucial Role
Currently, simulators provide limited feedback regarding the condition of the airplane close to or inside an upset condition. In today’s practice, trainees regularly exceed the aerodynamic, structural, or aeroelastic envelope, as this is not displayed in the simulator. Providing this important feedback to the pilot and instructor could be a straightforward enhancement.
ICATEE Industry Implementation
Through a comprehensive training needs analysis, ICATEE has been able to define the training objectives, needs and means. There is no one, single place or environment where complete upset prevention and recovery training can be conducted. Integration and standardization of multiple resources in a properly structured manner is the key. By properly integrating academics, aircraft training, and simulator training, commercial aviation will achieve the maximum reduction in LOC-I related accidents.
Integrated UPRT is being embraced by major airlines. Recently, KLM Flight Academy announced it has mandated a module of UPRT to participants in its airline transport pilot program, using academics and aerobatic-capable aircraft provided through Phoenix-based APS Emergency Maneuver Training. Both organizations are ICATEE members.
ICATEE plans to recommend enhancements to the simulator requirements in ICAO document 9625 when revised in 2012. ICATEE will also issue a Manual of Upset Prevention and Recovery Training to ICAO in the same year. Furthermore, ICATEE will release through the Royal Aeronautical Society a UPRT Research and Technical Report, describing the processes utilized in ICATEE to find suitable solutions, and how future research will be used to fine-tune such solutions.
What do you think?
1 Brooks, R.L., “Aerobatics Versus Upset Prevention and Recovery Training”. CAT Magazine, Halldale Media Group, Issue 2, 2011.
2 Advani, S.K., Schroeder, J.A., and Burks, B., “What Really Can Be Done in Simulation to Improve Upset Training?”. In Proc. of AIAA Modeling & Simulation Technologies Conference, Toronto, Aug. 2010, AIAA 2008-CP
October 31st, 2011 | 
Category:
Good article. States the major limitations with upset training:
1. Upset training in corporate or airline type equipment is too risky compared with the potential for loss.
2. Present training simulators do not have the true “sense” of stalls or uncontrolled flight.
3. Training in current simulators is limited recognize a potential event that could lead to uncontrolled flight like, windshear, GPWS escape, turbulence etc. These events are usually given with notice to the pilots, thereby, removing the startle-effect.
4. Pilots are trained to recover from a stick-shaker without confirming the aircraft is really stalled. I call it monkey see stick-shaker, monkey recover. Pilot could be putting their aircraft into a uncontrolled situation or recovering at an unsafe altitude in a knee jerk reaction.
5. Many stall situations are at cruise altitude due to attempting to climb beyond the capability of the aircraft. Many pilots can not differentiate between low-speed and high-speed stalls due to poor training and a lack of confirmation of symptoms prior to taking action, ie. Air France
6. An over-dependency on automation strips pilots of the basic airmanship and basic aerodynamics of their aircraft.
The recent list of LOC accidents in transport-category aircraft leaves no doubt this problem is real and needs to be addressed. I am familiar with the APS program and I believe their model (intensive academic training combined with flight instruction in aerobatic-capable general aviation aircraft) is the best solution to this problem.
Another poster mentioned an “over-dependency on automation strips pilots of the basic airmanship and basic aerodynamics of their aircraft.” I specialize in teaching pilots to fly glass cockpit airplanes (I am the author of the book “Glass Cockpit Flying”) and I disagree with the idea that pilots of glass cockpit airplanes are less skilled at airmanship. My experience has convinced me that the problem in glass cockpit airplanes is that glass cockpit instrumentation can sometimes confuse a pilot who is not totally proficient with these devices as to the actual nature of their situation. In other words, it’s not that they don’t know what to do in a particular situation but rather they are confused about what that situation actually is.
Aerobatic-capable general aviation aircraft do not fly like transport category airline and corporate level aircraft. Aerobatic-capable general aviation aircraft are stressed to fly aerodynamics not transport category. In an Airbus 300-600 at 250 KIAS, 2 1/2 inches of rudder travel or 33 foot pounds of pressure is full-scale deflection of the rudder. Apply that twice and your rudder is no longer attached to your aircraft. The Airbus simulator does not mimic this, it takes full rudder displacement to get full scale rudder deployment. In an aerobatic-capable general aviation it takes full rudder travel to get full deflection of the rudder. Fly-by-wire aircraft are computer controlled and even less correlation of feel to rudder pedal to rudder position.
The use of aerobatic aircraft is an excellent tool to teach basic aerodynamic but there is NO correlation between the feel, performance and limitations between transport category aircraft and general aviation aircraft. If you want to teach the basics of spins, stalls, rudder rolls etc. great but any attempt to make a correlation is negative training to a transport category pilot. They just don’t respond in the same manner. Boeing and Airbus have both put out limitations on flying transport category aircraft in these upset training. Use of rudder only procedures are very dangerous. Rudder assisted recoveries can be dangerous because only engineering simulators have the true feel of the stall, roll, pitch etc.
Having worked at a major airline with 15,000 pilots and testing a cross section of them, basic flying skills can not be maintained unless pilots practice them. Basic stick and rudder skills decrease quickly if not practiced. Many pilots of automated aircraft do not maintain this standard. In my work with the Air Transport Association’s HF, CRM and Automation Sub-committee show this is problem throughout the airline and corporate world. Younger pilots seem even more prone to use automation to a level of almost total dependency.
Simulator tests, during a recurrent, consisting of having pilots takeoff an automated aircraft, in VFR conditions with no autopilot and no autothrottles, no flight director and fly a VFR pattern and return for a landing. Nearly 70% could not perform this task to ATP standards. Why? Because they turn the autopilot on at 400ft AGL, start using the Flight Guidance Panel and then LNAV/VNAV. Another test is to turn of the FD off on an ILS with a crosswind and watch the S-turns on final.
Why did Colgan and Air France occur. Among other reasons no basic airmanship and aerodynamics training and no practice flying the aircraft without automation. Just like basic instruments, the aircraft can be instrument equipped, the pilot can be qualified but if you have not flown in the weather in six months your in trouble.
I think we all can agree that this is a serious safety issue and that the training of transport category pilots needs to address this issue. We also all can agree that all of the proposed ways of accomplishing this goal (simulators, aerobatic-capable general aviation aircraft) are imperfect. However, we should not let the unattainable perfect be the enemy of the good. We are not going to have pilots practicing unusual attitudes in Airbus or 787 airplanes so we are never going to have perfectly realistic training in this area. That does not mean that we cannot use these other resources to improve the ability of transport category pilots to deal with LOC events.
colt 27 is slightly off the mark when he states that there is “NO correlation” when using aerobatic aircraft to teach Upset Recovery to Transport Category pilots. Several studies, and my personal observation, shows quite the contrary. Just as the skill sets all pilots learn in small GA aircraft (take-off/landing, instrument procedures, general aircraft handling, etc.) transfer to bigger aircraft — so do Upset Recovery skill sets learned in an aerobatic aircraft.
What colt27 may not know, is that when pilots are trained in Upset Recovery techniques in aerobatic aircraft — good courses train only to the limits of the clients aircraft. The aerobatic aircrafts excess load limits are just used as a safety buffer. The feel (proprioception or g-force feel) does fully transfer and mimic the feeling the pilot will experience in a transport aircraft. 2.5 Gs in an EA300 will “feel” the same as 2.5 Gs in a B737. We go out of our way to de-emphasize stick force (and have a couple of tricks to enforce that to the client).
Good knowledge/skill-set transfer is further validated when clients follow the on-aircraft phase with a transport category level-D simulator phase (the training sticks) and by “real-world” experience from our alumni.
I do though, fully agree with colt27′s comments on automation. For a lot of reasons, as professional pilots many of us have forgotten how to “fly the aircraft.”
Overall interesting article. Loss of control is certainly an issue that needs our attention, but I believe that we are missing a point here.
Before the pilots lost control of the airplane, they had lost something else: situational awareness!
It is therefore essential that pilots are (re)trained on understanding and recognizing the pre-stall and stall regimes.
Unless you understand and recognize the situation you are in, there is no hope that you will take the correct actions.
Academic training and training in the present simulators will already bring us a long way towards reducing this risk.
Going into post stall regime in simulators using calculated or wind-tunnel data is a waste of money and can create a false sense of competency of being able to get out of these regimes. Nobody knows if you can get an Airbus or a Boeing out of a spin in one piece since nobody has done it.
In the initial, ab-initio, training the intelligent use of an aerobatic trainer as described by bwillett is essential.
Prevention is the key to success to reduce these accidents.
Two current articles caught my attention about the basics of this subject. The first was “Learning to Love Stalls” by Joseph E. (Jeb) Burnside in the November issue of Aviation Safety Magazine (access by subscription only) in which he states:
“Stalls should be respected, not feared. Understanding how they develop and progress can go a long way toward eliminating any unreasonable fears. Among the concerns expressed by a brand-new student pilot I was talking with recently was what I took to be a strong fear of stalls. I didn’t have the opportunity to ask him where he learned stalls should be feared. I did, however, relate they were important but—at least when understood—weren’t anything to fear. Which is not to say they shouldn’t be respected. Key to understanding stalls, of course, is knowing why and how they occur, why we practice them and how we can use the knowledge and experience gained during that practice to prevent more dramatic behavior, like deep stalls or spins, especially when close to the ground.”
The second was “The no-spin zone” by Barry Schiff in the December issue of AOPA Pilot (access by subscription only) in which he describes the time he failed a flight test for his flight instructor certificate in 1956:
“It was an era when instructor applicants had to demonstrate spins during the flight test … the government inspector said ‘that was good. Now let’s see an over-the-top spin to the right … an under-the-bottom spin to the left …’”
Barry confessed he didn’t know what either of these spins were but quickly learned. In this one page article he describes both his learning experience and the importance of the knowledge he gained.
Unfortunately, neither of these articles is available for free public access. Do you have similar experiences that you are willing to share as an IAFTP training practice?
“New guidelines coming for stall training” is a related article that appeared in Aviation Week on 25 Nov 11:
http://www.aviationweek.com/aw/generic/story_channel.jsp?channel=comm&id=news/awst/2011/11/28/AW_11_28_2011_p44-397089.xml&headline=New%20Guidelines%20Coming%20For%20Stall%20Training
“We moved away from aerobatic training in the 1980s, says Avani. And the training pilots do receive can teach the wrong response …”
I feel that automation has led the pilots to believe too much on technology and lot less on the fundamentals or theory of flight.
Why do we keep facing these situations? A pilot learns the basics during the initial stages, as they progress in their careers, firstly there is no refresher, secondly they are told to rely more on automation.
I have observed that the new breed of pilots take automation for granted. If automation fails or gives an incorrect command they are hesitant to get back to raw data. This is because they haven’t flown an airline transport aircraft without automation . Todays training syllabi is so compressed that there is hardly any time for the pilot to think. The need of the hour is to discuss about rationalization or commercialization of training programs.
If theory of flight is discussed more often issues like handling the aircraft or giving incorrect inputs will not arise.
The move towards increasingly automated flight decks is considered by many to be a casual effect in the decline in pilot handling skills and may be a contributory factor in the increase in the number of LOC-I and flight upset incidents in the industry. The use of full flight simulators to train newly graduated pilots and current line qualified pilots to recognise and recover from LOC-I and flight upset scenarios, as mentioned, is now under investigation through the RAeS and the ICATEE. While modern Level D Full Flight Simulators offer many training benefits, current conventional motion platforms are unable to present pilots with the effects of sustained ‘g’ acceleration cues. The latter is deemed a necessary element in recognising and recovering from flight beyond an aircraft’s normal operating envelope.
For more than 30 years Cranfield Aerospace Limited (www.cranfieldaerospace.com), the commercial arm of Cranfield University (UK), has built up a considerable international reputation in the design, development and delivery of ‘G’ cueing seats for military applications in both centre thrust fixed wing fighter/training aircraft and helicopters. In the majority of training requirements the combined use of ‘G’ seat cueing and ‘G’ suits has obviated the need for conventional motion platforms.
Through a co sponsored agreement with Cabair College of Air Training, we at Cranfield Aerospace are now ready to offer up to the commercial airline industry a Flight Simulation Training Device (FSTD) fitted with a motion cueing seat for evaluation in a range of potential applications in commercial pilot training, the primary being the use of the seat as part of teaching, recognising and recovering from loss of control and flight upset situations.
In that respect we remain open and ready to receive constructive feedback from which we would plan to launch our case to those aviation organizations and committees engaged in studying the ways and means to optimize LOC-I training. I believe a successful outcome to our evaluation of seat motion cueing offers up a potentially new dimension in the use of flight simulators for LOC-I training.. In my view the use of simulation to provide high fidelity LOC-I training far outweighs the possible recourse to airborne training – a costly and potentially hazardous alternative.
SKYbrary provides the following reference in regard to the basics of recovery from unusual aircraft attitudes: “When it comes to recovery from a loss of control or ‘upset’, an understanding of the principles involved is valuable to any pilot. However, current transport aircraft types can be divided into two contrasting groups in this respect. Those modern aircraft types which have fly-by-wire primary control systems also have built-in protection which will ensure that recovery occurs without a manual response from flight crew. For all other aircraft, successful recovery from any significant loss of control is likely to be a challenging aircraft handling experience for which both some theoretical understanding and simulator practice will be needed. This article is just a start.”
http://www.skybrary.aero/index.php/Recovery_from_Unusual_Aircraft_Attitudes?utm_source=SKYbrary&utm_campaign=18dee888dd-SKYbrary_Highlight_26_03_2012&utm_medium=email
“Airbus is going to train pilots for its A350XWB differently. The first three days in the A350 simulator will be about letting the pilots find out that it is ‘just another aeroplane’. Without using any of the sophisticated flight guidance systems…”
View the entire entry from David Learmount at:
http://www.flightglobal.com/blogs/learmount/2012/09/airbus-takes-pilots-back-to-ba.html
LOC-I is a still unmatched challenge.
Acrobatic airplanes and airliner simulators are not able provide a realistic training environment. Transport category simulators are technically not able to provide what is needed. Acrobatic airplanes can only be safely manoeuvered during visual day condition.
It is the complexity of factors that creates situations that pilots are not prepared for. Regular training does not account for it. Regular training devices are not cabable of filling the need.
AeronautX is soon to offer a training scheme to bridge the gap between techs and no-techs. Practical training will not only cover all known physiological phenomena that a pilot might encounter but also prevention and recovery techniques in IMC and night condition.
Test-runs with have shown incredible results. Final work on our training modules will be finished in March. We will provide more details soon.
Joerg