EASA NPA for Upset Prevention and Recovery training

[RAT 5]

Regarding the training philosophy these days: firstly I'm concerned that the basic CPL flight training has been reduced from 250 - 150hrs. Not every pilot will end up in an airline jet. There are Air taxis & biz-jest our there. OK, the amount of VFR nav-ex we used to do was fun, but with GPS these days perhaps that should be reduced, but 100hrs seems a lot. MPL courses with a lot of sim based training might be considered OK for airliners, but it must be at the expense of developing manual handling skills. Perhaps that is where the problem lies; there is an attitude that manual handling skills above 500' are not required......until they are. Consequently airlines with modern a/c do not want to spend time teaching beyond the basic base check and do not like it when the number of Go-rounds increase on sunny days. So they discourage such frivolous activities as manual visual approaches; and the downward spiral continues.
I think back to B732 training. It had the most simplest of autopilot and the nav system was the same as you saw in flight school: a HSI with VOR/LOC display, a DME and a basic T instrument panel. It did not take any extra training. So we first learnt to fly/handle the a/c; then we learnt operate the a/c; then we added some systems non-normal, and then we added some emergencies. Todays TR syllabi is biased towards the operation of the a/c under normal & non-normal conditions. Because of the new sophisticated systems there is more FBS time, but then the FFS hours are also absorbed by a/c operation with systems non-normals and much FMC programming and briefing and set-ups and use of automatics. There is little time for handling, except for those mandatory in the LST. The basic mandatory manoeuvres required are the same as B732 35 years ago. V1-cut, SE LS & G/A are still the same. There has been very little evolution of modern day testing in modern day a/c. It has been static. The amount of FFS time is much the same. One possibility would be to add 2 sessions to the TR for handling, but that's cost, and for whom? The XAA's say their job is to set a base standard of handling and the operators can set their in-house standards and train their crews as required. To me that should include the manual skills that the TR course does not have time for. Base training is not enough to achieve all that is required. It needs daily practice and exposure, and principally that means approaches. Every take off is manual and following the FD up to CRZ with constant thrust is hardly difficult. Manual flying from FL100 down is another matter, busy TMA's excepted. There is a time & place; but even there, from 5000', it should not scare the horses or small children.
How many airlines encourage this? What does your employer encourage? And I'm not thinking just of EU airlines, but worldwide. We, the pax, climb aboard partner airlines from very different cultures. Their standards might not be the same as the host airlines we bought the ticket from.
It might be dream to have a worldwide standard and similar philosophies, but the first step on an adventure is always the most difficult and the most important. Who will take it? FAA? EASA?

[Reverserbucket]

The basic training content is indeed part of the problem - too greater emphasis on completing the disproportionate solo cross-country element (typically navigating (by satellite) to the same destinations and turning-points over sparse, practically featureless landscapes) with far too little G/H in place beforehand only leads to practicing bad habits picked up in the early hours of training which cannot be corrected adequately as the instructors simply don't get a chance. This is no criticism of instructional standards - simply a case of insufficient time to teach the air exercises thoroughly enough before the solo components. By far the majority of the CPL course is spent engaged in straight and level and this is one of the root causes that needs to be addressed. MPL - even less basic handling. UPRT is a 'Bandaid' to a far greater issue which needs addressing properly, as RAT 5 suggests, and with the increasing emphasis on CDA and 4D trajectory based traffic management, it's one that needs addressing soon.


fokker1000 - TG still takes the C150A out for a 'spin' or two on Saturdays I understand

[Dan Winterland]

Appreciate some ATO's include an element of UPRT in the initial type rating stage but not much.

The problem with simulators is that they do a poor job of UPRT. The motion is not representative of the real manoeuvres and the UPRT I have done has been of poor value. There is no substitute for real hands on practice at the early stages.

As an RAF pilot, I was subjected to lots of such training at an early stage. Although some of the manoeuvres are not necessarily relevant to modern commercial operations, the basics are. I frequently surprised sitting next to some pilots that they have never seen extreme attitudes for real - 60 degrees angle of bank is their limit. My company trains a lot of pilots through a cadet scheme. They all get 5 hours in an aerobatic aircraft learning how to recover from extreme attitudes and are better pilots for it. This is a company initiative and not a regulatory one.

[deptrai]

RAT5, MPL courses with a lot of sim based training might be considered OK for airliners, but it must be at the expense of developing manual handling skills.

A lot of the MPL sim training is "hand-flying". There are fresh MPL graduates with very good (procedural) manual handling skills, for the type they have been trained for, within limitations of a sim. For airline cadets, I can see how MPL make sense (no sarcasm). Acquiring 80 more hours of manual handling skills in a real aircraft (classic ATPL path) also has it's place, but frankly there are limits to how much of those skills are transferable from a much smaller, prop a/c.

Rather than 80 more hours of cross-country cruise in real aircraft (they'll get that quickly on the line), I think 5 hours of aerobatic training as Dan Winterland suggested is a better idea to improve cadet training.

[RAT 5]

Guys: I agree the time to correct this is at basic training stage. The students are enthusiastic pilots who are then not being allowed to fly. It is sad to meet some disillusioned souls stuck in RHS of an aluminium tube not being encouraged to fly. If the basic training issues can be fixed they will only bear fruit if the skills can be carried forward and kept sharp on a daily basis.
I wonder how many incidents were converted into accidents not because of lack of skills, specifically, but by boredom and complacency that allowed the situation to go undetected. I include in this the simple scenario where you are sucked in hot & high and do not react in time: and ultimately do not go-round.
The best lessons I ever learnt in my jet apprentice days was: by my first PNF G/A due not visual at DA; by my first PF G/A from an SRA (not visual); by a G/A as PNF from a visual circuit flown by an ace. A thermal tailwind kicked in <500' on a shortish runway, right in front of the terminal and all the pax. No compunction to toss it away and go again. Best lesson ever.
Now that would be an interesting practice. An a/c that becomes subtly unstable <500' with no windshear warning. PF has to decide. That is a +ve lesson, plus more practice at one of the most realistic possibilities - a normal all engine G/A - that is most often routinely messed up.
In the sim the mandatory G/A is of little value as you know it's coming. It's the same with RTO's. At the end of the detail you always finish with RTO & Pax Evac. There is no surprise element.
V1-cut. I once had a TRE give me the engine failure at TOGA from an NPA. Now that was a real +ve experience. Vref40+5 = V2F15 so it was classified as loss of thrust <V2. Ticked a box, but in a very educational manner.
More imagination vicar.

[BEagle]

Although Option 1 is fully in line with ICAO, Option 2 requires more than ICAO does, because the proposal also includes LAPL(A) and PPL(A).

EASA strikes again! Anything which EASA admits to requiring more than ICAO does should be rejected on principle.

[GlobalNav]

Questions.

The limitations and shortcomings of piloted simulation - a given. The potential educational value of flight training - in a relatively small trainer - a given.

How does any of these help a pilot understand how to skillfully handle a heavy transport at FL350, let alone provide a meaningful level of control proficiency? How many pilots ever manually maneuver a heavy at FL350 under normal conditions when all indications are normal, correct and expected?

Considering the surprise, suddeness, and confusion surrounding an upset at altitude, coupled initially with disbelief and the likelihood of missteps that compound the problem - what are the chances that a typical, "experienced and trained" line pilot will recover safely?

Would the amount and content of a mandated training campaign, that operators would accept, give time for and pay for - hence that authorities would actually mandate - be an effective solution?

Is something more than training needed?

[Lonewolf_50]

Originally Posted by 4Greens

The main problem in unusual attitudes is the effect of g forces. This needs to be a part of any serious training course.

That isn't what I saw when I was teaching it. (Granted, over two decades ago).

Did you find this to be true with people who have developed instrument scans, or those who haven't yet?

[Dan Winterland]

Questions

.

Answers.

How does any of these help a pilot understand how to skillfully handle a heavy transport at FL350, let alone provide a meaningful level of control proficiency? How many pilots ever manually maneuver a heavy at FL350 under normal conditions when all indications are normal, correct and expected?

It gives them the confidence to handle an upset and a good idea of how to recover.

Considering the surprise, suddeness, and confusion surrounding an upset at altitude, coupled initially with disbelief and the likelihood of missteps that compound the problem - what are the chances that a typical, "experienced and trained" line pilot will recover safely?

Much improved as the pilot will be better prepared and less likely to suffer any 'startle effect'. Also, see answer above.

Would the amount and content of a mandated training campaign, that operators would accept, give time for and pay for - hence that authorities would actually mandate - be an effective solution?

Probably yes, but we don't know until someone turns around and says "thank God I did my upset training in that CAP10". Until then, it's best guess.

Is something more than training needed?

What do you suggest - a power-point brief? From my experience as a military QFI, you can't brief a severe upset. the student only gains confidence by experiencing the recovery.

[alf5071h]

This NPA is a sad reflection of the state of safety regulation in Europe.
EASA has a bureaucratic style, aligning safety documentation more with the EU legal processes than those aspects required by operators; how can a 135 page NPA be justified to introduce small changes in training requirements.

The NPA fails to provide a convincing argument of what the nature of the safety problem is; LOC is the result of previous activity or lack of it, surmised to be deficiencies in knowledge or training. Yet with the same logic, the vast, overwhelming majority of European operators who apparently do not suffer LOC incidents, have satisfactory LOC prevention training. The industry has yet to understand what ‘good’ operators do well in training in order to avoid LOC situations.
If these operations represent an improved method of training then why not identify and share these aspects, perhaps avoiding the risk that additional requirements might detract from training which is now well executed.

The safety process represented in the NPA is reactive; it is not a wide ranging human centred view, but instead it is adversely human specific to the point of blame and train.

What is the root issue; it is difficult to be specific in any operation involving human activity. The NPA reflects this uncertainty and the resultant is a shotgun approach which attempts to cover a wide range of aspects, none of which appears to provide the certainty of improving avoidance of a LOC accident.

There is little if any correlation with the primary factors in the referenced accidents (3). The accidents involved technical weaknesses, but all of the incidents used technical protections to mitigate weak human performance; perhaps the critical issue is with the availability of technical protections in the event of poor awareness. Thus perhaps EASA should be looking beyond human activity for the ‘root issues’ which might provide better focussed regulatory action.

[GlobalNav]

"What do you suggest - a power-point brief? From my experience as a military QFI, you can't brief a severe upset. the student only gains confidence by experiencing the recovery."

Hardly. I'm no more a fan of Powerpoint solutions than I sense you seem to be.

This is a case where education, alone, is insufficient, a change in behavior and performance/proficiency is required, in my judgement. Training is good, if there is enough of it, carried out in the right venue, with the proper syllabus and training objectives. While I would not totally discount some of the items that have been mentioned, I just don't see them as truly meeting the need. Unfortunately the frequency of occurrence is probably to low to convince the operators to invest the time and resources needed for sufficiently effective training.

I'm no inventor, but it seems to me that today's modern transports should have a flight guidance mode that provides the command guidance to detect and recover from the upset. We need to shorten the interval between occurrence and recognition/belief that it is happening, accurate identification of the condition and immediate correct intervention. The conditions are too critical to bear the delay of recognition, misinterpretation and wrong actions. I think the FGS needs to tell the crew what has happened, and unequivocal cues what to do and how.

[Judd]

What does your employer encourage? And I'm not thinking just of EU airlines, but worldwide

A general question calls for a general answer and that is, it is a good bet the world's airlines demand the pilot engages the automatics a few seconds after lift off and stay engaged until short final.

[Tee Emm]

The problem with simulators is that they do a poor job of UPRT

Must disagree with you on that point, Dan. Having taught unusual attitudes in 737 simulators from fully inverted after simulated wake turbulence encounter, to a 60 degree vertical pitch up on initial take off climb (unload and roll to the nearest horizon as per Boeing FCTM recommendations,) my experience is the flight instrument indications work exactly as one would expect in a real aircraft. OK the "G" forces are not there of course.

Nearly all the well publicised loss of control situations have been at night or in IMC where prompt instrument interpretation and manual flying skills would have saved the aircraft. I have flown in many 737 simulators and control forces (not "G" ) and flight instrument indications, have been satisfactory whether in inverted or steep nose high or nose low attitudes.

After 15 minutes of this training in IMC, every student I have seen trained comes out of those simulators with a smile on their face and armed with the confidence and technical skill how to manually fly out of trouble in IMC. Really, it is no big deal although there are those in the Regulator who make a big song and dance over UA training as if it is something new. As long as the ADI's in the simulator are capable of rolling through 36 degrees (which most are) then UA recoveries on instruments simply require average handling skills.

[DirtyProp]

Well, amazingly I agree with easa this time (but I still think its training sucks).
It's also my opinion that the syllabus should include minimum mandatory training in a glider.

PS: am I the only one who's getting fed-up with all those acronyms?

[RAT 5]

alf507iH: Excellent post. The solution will always be unsatisfactory if the problem has been mis-identified. More research vicar.

Tee Emm: After 15 minutes of this training in IMC, every student I have seen trained comes out of those simulators with a smile on their face and armed with the confidence and technical skill how to manually fly out of trouble in IMC.

Oh that this was a common training scenario. In the 3 year cycle of box ticking I've been subjected to the sim data based stored UA's for 2 manoeuvres, perhaps 3. My oppo also had 3 attempts. Total time <10mins total for 2 pilots. Box ticked. Move on to more 3.4/3.6 items and tick the boxes. AGH!

[fdr]

MCC/upset training

Centaurus, concur with your observation, but the problem goes well beyond CASA's following of the nonsense being exported from EASA and caught by FAR 61.

On a side note, as one who is relatively frequently stalling jet transports, i assure the people assuming that the simulator has validity of the fact they are basically mistaken. At the point of a g break, being a substantial distance in front of the cg becomes much more interesting than any simulator. The control resposes are correct when the aircraft gets unloaded, but otherwise are different enough to merit thorough briefing.

The planes still recover nicely, if the finally incorporated non involvement of thrust application is adhered to. Otherwise, my large jets behave similarly to my learjet through the early stall. Stalling smaller military jets does not give much of the control issues that the swept wing transport has inherently.

Training recovery in some small/mid size jets may be reasonable training, but otherwise, the limitations are that the simulator does not accurately model the handling, dynamics, vertical and longitudinal loads, or lateral loads that are experienced at the cockpit. The intensity if the buffett in the aircraft relative to the simulator is enough to give pause.

Personally, i think that upset in a Pitts while lots of fun, has general confidence but limited direct applicability( inhave changed my view on that) but, any aircraft or device that reinforces the universal recovery technique of neutralising the controls, reducing thrust, attaining a wings level attiude and recovering from the upright or inverted dive remains valid in our B737, 777. And even Airbus aircraft if loss of control is identified.

The biggest single issue IMHO, is the failure of recognition of loss of control, which was evident very early on in 447, but not recognised. Aircraft unless impacted by a change in dynamics (National etc) want to fly, they are failing as flight crew are not recgnising the loss of control occurring, and then do not follow very simple recovery techniques. These do need to be ingrained, as in the real case, the loads can become quite a factor towards cognitive overload.

MCC itself is a bureaucratic irrelevancy to the pernicious issue of lack of fundamental flying skills. Additional issues are the trends towards prescriptive solutions towards compliance that reduce the skill ,aintenance of our flight crews. My flight crews fly mulltiple jet transports in R&D, as well as maintaining skills on various aircraft with varied configurations and handling qualities. Loss of control is a briefed item prior to every test flight, and is practiced routinely in the sim sessions we undertake for proficiency training. We do around 4 times the sim sessions the airlines do, but all training in unusual manoeuvers is prefaced by the understanding that the simulator is limited in validity.

Any program assuming that the existing and proposed standards will give solutions to issues of operational safety may be ill prepared for the realities of stochastic system behavior... Compliance is not safety per se. Loss of SA at all levels remains the primary problem, as in 447, at I, II, (and level III on occasion)

Good luck out there.

[Linktrained]

Perhaps all stalling training is done with power off or at Flight Idle. And well below F/L 20.0 as well.

If a pilot has never hand-flown above this, then he may not be aware just how sensitive the controls may feel.. Does a Simulator give a sufficiently accurate a "FEEL" in some fairly ordinary S/L flight at height ?

( I have thought about G and cannot see how provide enough in each direction even in an accellerometer! It will just have to be briefed.)

At some moment on 447, Flight Idle was selected and the nose lowered, but that was as it would tend to do with the engines mounted under the wings.

I do not recall what comment was made (if any) before TOGA was re-selected. Or afterwards...

The Captain returned to the Flight Deck. I do not know whether he had heard a change in the engine power or not. ( It might be thought to be unusual at cruising level and might have helped to hastened his return.)

Someone must know. Perhaps it is still in the FDR.

A POWER ON STALL at or above cruising level would have been a bad start.
LT

[RAT 5]

The sim stall training I've experienced is always at idle power, low level, and anticipated. I've never experienced a power ON stall at high altitude. Yet real life scenarios of stall have often been in that environment. Pitot/probe heaters off at FL 270CLB = stall: an unmonitored inadvertent auto election of V/S stalled the a/c at high FL in climb. These are real life power on stalls. Never trained in my 35 year life time.
We should include UA upset that results in a stall warning. This is what we should be training in the sim. Real life is not the time to wonder what the heck to do. That is what sims are for. IMHO sims have become mandatory or recurrent box tickers. I admit some of my operators have introduced some very worthwhile scenarios to experience as a LOFT. Excellent, but then the rest of the session was tick in the box items.
All I'm saying is IMHO the use of simulator time is often not used to its best value. Many airlines have sim time to the absolute mandatory annual minimum time. That tells me all I need to know bout the attitude of the airline training dept. It should also tell the relevant XAA the same. But, sadly, the mandatory minimum is sufficient for the XAA to sign off, and with airlines being run by profit motivated accountants that is what you are going to get.

[misd-agin]

There is no fear in a simulator. Nothing can simulate the actual experience being in an airplane can give you. And yes, that includes stress and fear. In a sim you can always decide to walk out. That is NEVER an option in the airplane. That is a night and day difference.


Rat5 - you said there was no relevance between a Piper Cub and an airliner. Currently there is no relevance between stalling a sim and stalling a jetliner. Aviation Week&Space Technology had several articles, to include interviews with Boeing and Airbus, on this issue. Boeing and Airbus have agreed to a generic post stall model for n/b a/c. They've said that w/b post stall behavior is different. Why a generic model? IDK. But can you imagine the internet fire storm, and marketing nightmare, if it people found out that the A or B product had worse post stall handling qualities?


Anything you'd experience in a simulator right now would be the software impression of a TRE/CKA that gives the simulator it's annual(?) fidelity check. Is that how the airplane performs? Tweak and certify. Next question - how many stalls does the TRE/CKA have in the airplane? At high altitude? See the problem? A guy with no experience is tweaking the simulator to what he thinks it will do.


A friend said that he'd done a w/b stall in the simulator at altitude and it was pretty docile. If it was pretty "docile" was does the AF 447 report show cyclic bank angle changes of up to 30-40 degrees from left to right? Company had a high altitude upset. Guy involved in the investigation asked me what bank angles I thought they experienced. I guessed 30 degrees. He said "not even close".

[RAT 5]

Rat5 - you said there was no relevance between a Piper Cub and an airliner

That's a little out of context, and I can't remember the exact discussion. I fly both types and enjoy both, in their different way. I think basic CPL training should be rethought. Whether light a/c recurrency during an airline career is a necessity is another point.

About stalling: airline stall training in a sim is always at stick shaker. It is not stall recovery, it is not even incipient stall recovery. It's a manoeuvre accomplished BEFORE the stall is reached and to avoid getting into the stall. When I had freedom how to teach B737NG TR I used to get the students to hold into the stall and feel the stick nudger. I get them to let go of the elevator and see what happened. I'd get them to stabilise at SS and waggle the wings. I'd get them to stall with A/P and ALT HLD and idle thrust and recover with A/P remaining engaged. In manual flight I'd get them to recover at SS with level flight and with descending flight and different power applications. They'd do SS exercises in turns, increasing bank to cause SS or pulling more g. And I always said reduce attitude/load BEFORE adding thrust, by a split second. FCTM said apply thrust and reduce attitude. Now it has been reversed with greater advise about allowing SS to cancel 1st. The truth of aerodynamics was rediscovered.
The point is the stall training was an in-depth training session to learn about the a/c and at different altitudes, configurations and speeds. We even included stick shaker caused by UA at low speed and low thrust, i.e. mishandling. It might take 30-45 minutes per student. Now it's about 10 mins and very low-depth.
Pitifully minimal.
It should have been covered extensively in basic flight school training. It should have been continued into TR sim training. It hasn't been. Would small a/c recurrency experience fill the void? I'm not sure.