Signs of a new scandel emerging. Three Years After Air France 447 Crash, A Hint of Scandal - Global - The Atlantic Wire
Perhaps CCTV should be fitted in rest areas.

• see
• think
• react

• input
• compute
• output

You can present the "triangle" in many forms, but you need to know conclusively two sides to ensure an understanding of the third.

The thousands of posts in many PPRuNe threads on the subject have traversed this fundamental triangle many times and from many angles, and all propositions inevitably involved guesstimates of at least one side and probably two. The BEA's final take on the same problem will most likely also fall into this great "unknown".

What bothers me most at this point, is the obvious disconnect between both co-pilots and their failure to observe SOPs in even an elementary form. So my conclusion is that this "disconnect" went beyond the confines of the cockpit and may well have involved all three of the flight deck crew.

Yes, just another triangle. :uhoh:

SM - Well, I wasn't referring to that, but if it helps solving the "problem", then use it.:}

SM
Age: 64
Posts: 683
Couth: 0
So long as you don't mind CCTV cameras being fitted to your private office, your home, your bedroom, your shower, and your lavatory ... not that anyone wants to watch you ... :p

@mm: well said, in re various triangles, but not the silly love triangle speculation dreamed up by some.

When I have 300 lives in my hands I'd have no objection.:ok:
No true proffesional would.

Hi,

Welcome back of the sleeping captain Dubois story .....

BEA report N°3 (plain english)
2 h 11 min 45 - 2 h 01 min 46
Do your maths .....
10 minutes .....
Call me whatever you want .. but my opinion is that the back of the captain never touched the bed sack and the captain never closed his eyes ......
ZZZZZzzzzzzzzz :=

From the "report"

"more than a minute" :eek: I'm shocked, I really am :uhoh:....

Think about it - everyone the most mighty skygod, requires a few seconds to wake up, especially if they've gone off into a deep sleep....and most of us don't sleep in the uniform ( creased uniform is bad for the company's image, as is running to the flight deck in your shreddies, don't you know :bored:) Most of the time on a flight with a "heavy" you're lucky to get a resting crewmember(s) back onto the flightdeck in under 5 minutes from initiating a wake up call. You can run all the love triangle and conspiracy theories you want but we all know (don't we?) that it takes time to wake up and be aware of one's surroundings - I can't see how putting TV cameras in the bunk area is going to solve that particular human factors' "problem", but nevertheless I see the usual suspect has declared him/herself as a fan of the idea .....

FWIW since I was utterly cream crackered somewhere over the Far East last night it took me less than ten minutes from seat to a very deep sleep j.c. , so my opinion is that yes, it certainly can be done.

Hi,

wiggy
You maybe .. what was your rest time before the flight .. how much time you had flying
Remember that captain Dubois had many many hours (days) for rest in Rio and he was not from looong time on flying AF447 on the return segment ....
Other:
I just finished watching the video of the Nightline program and Mr. Troadec (BEA chief) does not seem very at ease to answer certain questions or makes a dodge with a smile that speaks volumes
He was more at ease during his meetings in France
But journalists did not asked him the same questions

If he was attending to a call of nature, getting back to the cockpit in a minute from when called is actually quite quick.


Chris N

ChrisN

Agreed.


j.c.

It doesn't much rest he (or I for that matter) was rostered to have had. In an ideal world (the sort of world the regulators live in) after a multiday slip you will be well rested, but we all know that sometimes you're not at all rested. You've maybe had bad sleep, had your sleep disturbed by the party out by the hotel pool or the couple in the next room :ooh:.

In heavy crew ops as a captain, IMHO, you are entitled to think that your colleagues on the flight deck can action, at the very least, the initial items a non-normal checklist, and contain a problem long enough for you to get dressed and return to the flight deck. If (note, I say "if") that wasn't the case on AF447 then the problem lies with AF's procedures, training and checking of junior crewmembers, not with captain Dubois sleeping arrangements.

Turbine D
I would argue that what we are doing is very definitely not 'hindsight'. To my mind the question with hindsight is 'could the actions taken in the past have been avoided with the knowledge available at the time?' If the answer is yes then the discussion is 'looking back' not hindsight. If the answer is no then the discussion is hindsight.

Leaving aside the issue of Air France and it's part in this accident if we ask whether the crew could have avoided the accident using knowledge that they could have reasonably expected to possess or have access to we reach an answer of yes they could. Why they didn't is part of finding out why the accident happened. A lack of any evidence that the correct procedures were followed leads one to think that things went very badly wrong but this is not hindsight. As I understand Airline training and practice it is based on following these and thus one would expect such practices and procedures to be followed in all cases. Once again this isn't hindsight but just an expectation that the already designed processes will be followed.

One can say that Air France should have expedited the replacement of the pitot tubes and this accident wouldn't have happened. This would be hindsight given that non of the incidents had led to anything resembling AF447. A situation existed which needed change but one could argue that this accident could not have been anticipated. However, if we were to say that the crew should have responded to this situation in a different way this is not hindsight. This is just an expectation based on a reasonable supposition that the crew was properly trained and conversant with the procedures for successfully operating a large transport aircraft. We trust airline professionals to reach a certain standard and ability. We may not expect them to be supermen/women but we do expect a certain level and the vast majority to conform to our expectations.
OC

OC,

Thanks for your response.

Your quote is exactly what I meant by hindsight. I limited my post only to the pitot tube situation, leaving what the AF447 crew did or didn't do up to those have flown or presently fly commercial jet aircraft.

OC, “if we ask whether the crew could have avoided the accident using knowledge that they could have reasonably expected to possess or have access to we reach an answer of yes they could.”

Do you think that they could have been reasonably expected to possess or have access to: manual flying expertise at FL350; and an appropriate UAS SOP at FL350?

(Not that these were the only factors in their inappropriate reaction to events – just some which seem to me and some others to exceed what might be expected, in view of the industry standard training in general from what I have deduced from these threads, and about AF training in particular.)

I think that SLF, regulators in principle, and MPs and the public, might have such reasonable expectations; but reading the difference between old-time pilots who grew up in a different era, and the modern 250-hour computer management “children of the magenta line”, do insiders believe it to be a reasonable expectation given the latter’s background?

chrisN
I am a general aviation pilot (biplanes, biplanes, biplanes) so I hope you will forgive me if my suppositions are wrong. I always understood it that modern civil airliners flew with a set of SOPs on board in the cockpit, immediately available for the crew to use. I also understood that airline training involved the use of SOPs and CRM to identify and troubleshoot issues that might come up. As for manual flying skills I agree with your last paragraph and I wasn't referencing these. I would suggest though that some form of manual flying skill is essential and to be expected - for any pilot.

OC re #1169. You appear to assume that because knowledge exits it can be recalled for use at any time, or that all situations will be understood (as you would understand them). Hindsight bias is within these assumptions.
Without detailed evidence we do not know what the crew thought – what was or was not recalled from memory, or how it was used in their assessment; similarly what they deduced about the situation and events.

The human mind has a natural tendency to ‘join-up the dots’ to create a familiar picture; – our understanding of the world as we would wish it to be. In order to learn from difficult accidents such as AF447 we have to restrain this tendency and only work with what we know factually.

There is much which we might learn from hypothetical speculation, by attempting to understand what the crew might have seen, thought, or decided. In this it is essential that our thinking is tightly controlled and that we do not form erroneous conclusions due to hindsight.
We might identify what physically happened; and with considered judgement it may be possible to identify contributing factors which could have influenced the crew, but without any assurance of proof.
In these circumstances the safety lessons to be learned come from questioning how we might see, think, or decide when faced by the contributing factors, e.g. as a start, would we have identified the situation as requiring reference to the UAS drill – why? Ask why 5 times, without knowledge of AF447 outcome.


Chris N, whilst it might be reasonable for MPs and the public (SLF) to have their expectations, those of the regulator and industry should be kerbed by the facts, and knowledge and application of human factors.
If, or when public judgement is required, then hopefully legal judgement would caution about hindsight bias. However, this is becoming a very disturbing area of aviation which requires both public (media) and industry restraint, supported with simple education of human factors to dampen expectations.

alf5071h
The issue here would be whether the knowledge available could be expected to be used by the individuals concerned and whether it was indeed provided for that purpose. Furthermore, one could ask whether the training is provided teach them about use of such knowledge. The answer to all these questions is yes. It is a reasonable expectation that a pilot of an airliner is familiar with his machine and knows how to access troubleshooting and analysis procedures. There are after all certain standards which must be maintained.

This is not hindsight or even linked with it. We know that the crew did not use the SOPs from the transcript of the CVR but we also know that airline training is based on use of these tools which are easily accessible. This also is not hindsight - the processes and systems are built on this usage. If (to hypothesise) the SOPs were not meant to be used by crew members then of course any statement that they should have used them would represent hindsight. This I do not think is the case - the SOPs are there to be used by flight crews and therefore to not use them is a demonstration of an incorrect response to the situation. This is fact - it is not hindsight.

One thing that demonstrably took much of PF's attention during the first 35 seconds after the AP drop was the very significant roll oscillation and the method used by PF to control the oscillation. (He had apparently never seen Alt2 law at altitude.)

If you look at the periodicity of the oscillation, and of the roll control inputs, his control inputs accelerated (in an apparent attempt to get ahead of the oscillation.) His inputs did not cease although the aircraft momentarily stopped roll oscillating on approximately 3 occasions, and these continued inputs caused a phase shift in the roll oscillation-based upon my analysis. This type of control input indicates formation of a control strategy, and correction of the control strategy. Clearly, a significant portion of his attention was devoted to the early roll control problem.

OC;
Some observations, for what it's worth...

I couldnt' get a single F/O to hand-fly the A320 or the A340/A330 when I offered (and in some cases asked them to fly). Primarily they were afraid of the thrust levers, (disconnecting them, controlling thrust etc) but for whatever other reason they just didn't want to actually hand-fly. I always thought flying a visual was both challenging and about all we have left to us to practise that part of our craft that was being discouraged, but I was told by my F/Os that a lot of captains refused the request to hand-fly because they themselves were uncomfortable disconnecting and it increased the workload on the PM. The PM set all the autoflight windows...the headings/altitudes and programmed the FMC when needed (which was heads-down of course), and kept a watch (there was usually a third pilot, an RP, in the cockpit...hand-flying was discouraged by the company except in low traffic density airports...yeah, right...low density international destinations) and when there was an "event" associated with hand-flying, they tightened the rules even more. We used to do it all the time on the DC8's/9's B727's, Lockheeds even the B767's but airlines want the automation used and now navigation procedures (RNP, STARS, SIDS) make autoflight necessary. The loss of skills, both handling and thinking, is a vicious circle process.

Now before we assume too much in this, the actual hand-flying the airplane isn't much of a challenge...trim is done for you, thrust is automatic and the autopilot is (depending upon the "hands"), smoother. In fact while hand-flying isn't much of a pilot's challenge, its more of an interesting challenge to make the autoflight work well. What's lost is the thinking skills when you have your hands on the controls....it's different than programming and flying through the Mode Control Panel; You're "connected" to the airplane.

I always felt I had good reason to trust the guys up front when I went back for the break. Still, as a rule I didn't go back went during the ITCZ crossing, "just because". It wasn't because of the hand-flying issue...to be blunt, I just didn't know what kind of radar skills everyone had.

At the time, I'd never heard of such a thing as "UAS" and losing the airspeed data at altitude, (1999 - 2007 on the 340/330). It was just never discussed or demonstrated.

I don't know what the answer is. Automation has become necessary, not just nice to have. But handling skills (which form thinking skills and a proper scan in my books) need to be re-introduced, taught and reinforced with practise. It wasn't in the script, but I can recall one sim in particular where the check captain made us do climbing then descending S-turns while changing speed. Every skill was practised...including the instrument scan.

climbing then descending S-turns while changing speed.

Such or similar fun used to be a standard first endorsement sim exercise to give the trainee a feel for the machine.

My favourite was to start S&L on a heading and then enter a steady climbing turn to arrive on the same heading, 1000 ft higher while increasing steadily to 30kt faster ... and then do the same on the way back down. Then repeat, this time losing the speed increments.

Got the eyeballs and thinking cells active.

Then, interspersed through the routine session stuff of the endorsement, get the trainee up to being able to hand fly, raw data, an ILS in 0/0 to a stop on the centreline. Very much confidence building and the I/F stick and rudder skills skyrocketed.

This was standard fare when I went through as a student and much later when I instructed in the Navy advanced jet program. We called the exercises Basic Instrument exercises because they formed the foundation for the rest of instrument flying. Mostly we used the S patterns, primarily S-2 and S-3 patterns involving standard rate turns combined with standard climb/descent rates at constant airspeed.

If the training aircraft had an autopilot, they were not maintained/not used and it was strictly a manual exercise.

Are modern day pilots still being taught to hand fly such exercises at any stage in their training?

FD1 was continuously displayed and FD2 was missing for a split second. The low resolution of the published data do not allow to tell more.

I disagree on that.
What was displayed in the FMA as a vertical mode ?
Where is the FD/AP vertical mode trace ?

It will be a major blow if the FD behavior is not extensively covered in the coming BEA final report ...

It sure brings you to the stall warning. That's flying on the edge with no necessity. What's the point when usual pitch and thrust for cruise is the answer.
Airbus has it wrong on that one, even its chief pilot says differently now.

Machinbird;
No, not to my knowledge and I keep in touch as best I can.

John T.
And it needn't be a drawn-out thing. Both pilots can get some practise in, in half-an-hour and carry on with the rest of the script. It's not difficult, doesn't require Congress to be involved in any new FAA regs and as you say, builds tremendous confidence. It was the most frustrating thing, to watch guys give up the opportunity to fly no matter what I said I'd help with. It's not difficult stuff but you can lose the touch and overcontrol quite easily, (seen it in the sim). What happened with the roll isn't serious but perhaps the PF didn't know that. I think that's all that happened with the PF trying to get the little bit of roll under control - don't stir the porridge...just set the stick in one position and hold it, wait for the massive machine to settle down, then tiny movements...But if you don't have the touch due to lack of training and experience, confidence in your solutions in trying to bring the airplane back with the stick can erode very quickly if you see yourself making a hash of things. It isn't complicated - it's what we used to routinely do.

A similar maneuver, but with changing speeds.

Lazy 8

@PJ2 Your relevation is frankly speaking "shocking".

RetiredF4;
Well, this is just one pilot's/captain's experience - others can speak out if they wish. For me, I really, really encouraged guys/gals to hand-fly including manual thrust levers and there were almost 100% no takers. When doing line indoctrination I taught manual flight, what to be aware of when disconnecting the autothrust, how Bangalore happened (the software's changed since then) and so on. I've only been a captain on Airbus aircraft. I hand-flew all the other types I was on but so did the captains, because we were better than the automation and it was, frankly, (serious) fun.

I think the Airbus is as easy and straightforward to hand-fly as any other type I've been on. The source of the "mystique" and therefore the fear of disconnecting and hand-flying is, in my view, twofold - 1) the airplane is complex with many modes including the autothrust, and 2) transition training and recurrent training focus competence and facility with the autoflight system and took competency at hand-flying (including thrust management) for granted.

The first time I did a proficiency check and IFR ride on the A320 it was entirely on the autopilot to demonstrate knowledge and understanding of the system. It felt really strange and it felt like I was somehow cheating and not demonstrating what I knew, (because with work, the autoflight systems aren't difficult to understand, remember and use).

Subsequent recurrent training sessions worked in a manually-flown approach on raw data and we'd also do steep turns. I never saw the "S-turns" exercise again and I'll bet these days most would find it very difficult to complete.

In steep turns, autothrust was left on and we used the bird, (FPV symbol) for steep turns. It was a piece of cake...for the autoflight. All that was necessary was to keep the FPV on the horizon. Looking back, with autotrim it wasn't really a test for handling skills in my opinion. Cognitive connection to the machine reduced because of this, I believe.

I hand-flew all previous types and so did the captains I flew with. For the A320 I often flew from top-of-descent to touchdown and the airplane was a joy to fly manually. I did the same for the A330/A340 into terminals such as Hong Kong, Narita, Sydney, London, (after we were handed off to the final approach controller and out of the hold!), Frankfurt on the downwind, Honolulu, the Caribbean.

Towards the end I gradually stopped hand-flying this way, not because the SIDS and STARS got too demanding but because I began to realize that if anything were to occur during hand-flying, the first thing they'd look at would be the fact that the automation wasn't engaged and I'd have some explaining to do. I think that is the wrong approach, but...before we come to a black-and-white conclusion that this is all bad, I think there is good reason on the part of airline managements to require the use of automation and to teach/train/instruct thoroughly on its use, abuse and failure modes. They are doing what they think is best in terms of risk management in an increasingly busy airspace and terminal environment.

That said, when an approach incident occurred, hand-flying was increasingly discouraged Formal policies which provide guidance as to when one can do it, (low traffic volume/low work load for the other pilot etc) helped but the policies effectively prohibited hand-flying because entering non-busy terminals, especially on international routes, simply doesn't happen.

I think the vicious circle has been complete for about a decade now where discouraging hand-flying has indeed resulted in a loss of those important but invisible skills: instrument scan while busy, smooth, anticipatory manual handling of the aircraft and engine thrust...in short, "energy management" (and therefore fuel cost management) and crew coordination during manual flying conditions...the change is subtle but material to effectively and safely piloting the airplane.

The routine was established in the 90's - the autoflight was engaged right after the last flap/slat retraction and disconnected at about 400ft on approach at destination. There is NO opportunity for practise under such conditions but guys realized that the airline wanted the autoflight engaged and most didn't argue. I think they should have, as pilot associations should have, but that is a personal view.

This isn't a sudden, unexpected, surprising state of affairs. In one of the many, many AOM changes we experienced came the admonition that "the autoflight WILL be engaged right after takeoff and disengaged on the landing roll". Aside from the fact that many of our approaches in the A320 were NPAs and couldn't be an autoland, the short-sightedness of such an airline policy was fought very hard and we won the freedom to hand-fly the airplane under an "automation policy". It was a step in the right direction but there was no formal acknowledgement that training was required and so the focus on autoflight comptetency remained and one hand-flew if one wished but it wasn't supported. I don't think we were unusual. We may be surprised and shocked but that's the way it was and, I suspect, is today.

These notions have been expressed since the mid/late eighties almost exclusively from pilots transitioning to fully-automated aircraft. The original reason pilots didn't want to hand-fly was because, "What's it doing now?" was a real question in the early 90's. "Click-click", (autopilot/autothrust OFF) was the solution until one sorted oneself out but pilots are primarily problem-solvers and sometimes will try to fix the problem (how do I get the OFFSET Page?...how do I get a hold entered again?), instead of changing horses and disconnecting while sorting it out. It's a cognitive thing, not a technical thing. It's why I keep saying that the UAS item shouldn't have been a problem...it's just straight manual flight, keeping it level and keeping thrust while the other guy/gal gets out the books. No big deal. Really.

So now we have to sort out why this one became a big deal and ended in an accident and in the eight or nine threads we've had a good go at it. Those here who are pilots know that this kind of thinking isn't unique, it's "what is". Nor am I alone in knowing that the solutions are comfort with the machine in all its normal and abnormal regimes and phases, and that the job of training (the company) and the job of learning (the individual pilot) isn't finished until that comfort is there. It isn't about notions like efficiency, cost-control or shortened training footprints, it's about pilots being familiar and therefore comfortable in their machine, no matter what it costs or how long it extends the training footprint. And it doesn't take tens of thousands of hours in a career, or weeks added onto the normal training footprint to achieve this comfort - it takes work, mainly on the part of the pilot, but also on the part of the airline in providing a supportive, comprehending management approach to foster this level of comfort.

In my view, the answers to AF447 are in one way not complicated. But these notions certainly are not "provable" in the traditional ways we are accustomed to accepting "evidence". There is no recorder that can record confusion, fear, competency or lack thereof. We must come to those conclusions, if possible, obviously by other means and these days, when Cartesian thinking invisibly rules (and narrows) our assessments of "evidence", we can miss some processes that may be relevant to the accident. The above "stream" tries to deal with this.

Getting this off my chest...sorry for the thread drift.

PJ2,
Thanks (again) for your erudite and informative post.

IMHO, if THE FLYING PUBLIC, not only those "in the know," were to be made fully aware of this situation (never an easy proposition), the effect on the industry would be devastating. 'Twould be a huge scandal.

Speaking as a self-elected representative for SLF everywhere, being flown by "pilots" who CAN'T FLY THE AIRPLANE is absolutely, unconditionally UNACCEPTABLE. As discussed earlier, having pilots who can hand-fly the pane in a pickle is not only a reasonable expectation of every passenger, but must be mandatory.

I, too am shocked that this is even controversial at all. I don't care HOW much it costs the airlines-- this has to change, in the grim and bloody light of AF 447. If it doesn't, there will be more tragedies of this kind until it is fixed.

:eek:

Organfreak;

Caution is advised in interpreting these remarks.

The industry's record speaks for itself and it is a superb record overall.

It is critical to an understanding of these remarks that they emerge from and dwell within a context of significant and even spectacular success. They do not portray an industry "coming apart at the seams" about which we must then ride off in all directions regarding skies and the falling thereof. Rather, if I may, these remarks represent a distillation of ongoing issues which have been "in process" for many years and with which most pilots and certainly all safety specialists are familiar. We know that the character of accidents is changing. Many interested parties will interpret these changes in terms of their own specialties, which is a good thing because nobody is capable of seeing and then communicating well, the whole picture, while keeping in mind the industry's successful record and extremely safe state of affairs.

This is pretty open stuff. Our industry is, because of its high-risk nature, pretty open itself because we value learning and prevention of untoward trends and events, above all.

No other industry or endeavour demonstrates this willingness quite so strongly and so such frankness requires a reserved and contemplative approach. These are not sudden trends or sudden events. We should be even more willing to discuss some of the less attractive aspects of the business but over-reaction is what stops many. Yes, there are always politics involved; -we need only take a look at what is happening in Nigeria* at the moment to know that there remains an enormous misunderstanding about how our industry makes itself so safe.

These kinds of things discussed in the post must be taken with some forebearance and perspective - millions and millions of hours of safe and unremarkable passenger transport, decade after decade. Such events as AF447 are vanishingly-rare. Whether perfection is achievable or not is not the quest - the attempt is what examination of these aspects of our industry is about.

Such frankness and openess can result in a form of "autoimmune" disease, if you will. The very characteristics of frankness and a willingness to look at the nasty bits, all of which make our aviation transportation system safe, (our "immune system") also at the very same moment in time, has the capacity to damage or even destroy that which makes it so safe.

In short, that which makes us successful also equally has the power to harm. The key to our industry's health (and therefore its continuing high levels of safety) is in how "the immune system" is treated.

PJ2
 

Thank you very much for your open words.

A friend of mine, also a retired Phantom driver, flew the A320 until final retirement 2 years ago. On the few occasions we were able to talk about flying the bus he was happy with the manual handling qualities, he always told me "franzl, you wouls have loved that plane". In the same moments he was telling the stories about FO´s unwilling to handfly the plane except for T/O and Land, and in more than one occasion he was reported up the chain for doing so.

Well, i didn´t believe everything he said, and your post sheds a different view to his narratives.

My nephew got accepted by LH to become a pilot last month, he is very happy about it. I myself was and am not sure, wether i should slap his shoulder for well done or wether i should caution him on being alert not to get a kid of the magenta line.

Not a thread drift at all, but I am surprised you did not mention the FD at all ... !?

Manually flying the 330 with A/THR + FD has nothing to do with manually flying.
It's only when FD is missing on the PFD that one can realize how much mental work is needed to keep things straight.

You told us how little manual flying is done these days, how much of that little is done without FD guidance ...

For AF447, as FD were not selected OFF, they keep coming back, the PF did not doubt them and was probably too happy to put his faith in them.

Hi everyone,

Interesting posts on the Humbling sim experience thread...

Well said PJ, training to confidence is what it should be..
Unfortunately the bean counters (think reduced training time) & the regulators think testing testing is the way to go, but then they will not be in the smoking hole and can always go and get another job.
:mad:

Actually, I was of the same opinion as I misread the 3rd interim; I believed that drill in QRH/FCOM stipulates memory attitudes while "if safety of the flight is affected" decision point is mentioned only in FCTM. Guess what: I was wrong.

By the time AF447 made its final flight I already left A320 and I have never worked for Air France. Our unreliable airspeed on 19/20 was very, very similar to AF on 30 with one very significant difference: there was no option to consider whether to apply memory items or to maintain cruise thrust/pitch. Legally, if I ever got to UAS, I had to apply 5° pitch above FL100 while PNF gets the weight/power/pitch tables out. Outcome-wise, it would not make a lot of difference if I would have set it or kept the things as they were.

Perhaps because those who wrote the checklist did not believe the pilots' capability to remember the typical cruise attitude and N1 for different weights? 5° with CLB works for any weight until level-off values are read from table and set.

Let me list some of the ways in which this statement is wrong. 1. pulling can lead to high-mach-low-threshold transient stall warning if aeroplane is jerked into climb, reasonably smooth pull from about 1.5 to 5° will avoid it 2. when settled at 5° pitch, AoA will be near cruise pitch and will gently increase towards 5° as speed is bled off. By the time it gets there, you won't be at high mach anymore. 3. If no valid mach, stall warning reverts to low threshold 4. stall warning is not stall itself - a fact cheerfully ignored by those unable to tell the difference between "approach to stall recovery" and "stall recovery". 5. when arguing about Habsheim showoff, you repeatedly claimed that alpha prot prevented aeroplane from achieving higher lift at even higher AoA (backside of the power curve, anyone?) and now all of a sudden, stall warning is considered to be the edge? Are you having it both ways or are you about to make a breakthrough in aerodynamics - discovery of the area of fantastic aerodynamic performance between alpha max and stall warning? Besides, it would add some credibility to the statement of yours if you could provide quote of Airbus chief pilot.

Don't you think if he was really trying to go for 15° pitch, he would have gone for it all of the time speed was not available? By the time second stall warning went off, left speed readout was normal. By the time aeroplane departed envelope, it was consistent with ISIS. No one checked it.

From CVR we know he never announced what he was doing. Form DFDR we know we know his actions were totally inconsistent with UAS procedure. What else do we need to know?

When I'm talking about Yuri Gagarin succumbing to spiral dive in cloud or Charles Basset and Elliot See perishing in controlled flight into building, I'm talking about astronauts, not some random airline pilots.

Good example of "Invulnerability" risky attitude of which every CRM course warns one about, even the bad ones.

Pilots must have far, far better emotional stability than general public, however it can not be absolute. Emotions do affect the pilots' performance and everyone can break down, given enough pressure. Trick is having the breaking point in the area which is extremely unlikely to be encountered in flight. Can this be achieved through training? I don't know.

Of course. Going emotional while flying will make one revert to basic instincts developed during millenia of terrestrial existence, which are sure to be wrong in the air and most probably fatal. Air is a rewarding mistress yet she is very demanding and impartial. If one doesn't abide by her laws every millisecond he's airborne, he'll get rejected promptly no matter if he's a newbie or an old acquaintance. Unlike pigeons, we don't have the flying skills in our DNA, it takes ability and devotion to etch them on the very surface of our cerebral cortex - the first thing to shut down when emotions run high.

Not just that, lot of them were not even detected until AF447 post-mortem uncovered them by sifting through QAR data. Pilots were so unfazed that they didn't even make reports.

Yes, if we apply selective and very narrow hindsight. Every aeroplane type has dozens if not hundreds technical issues being investigated simultaneously. How can you determine which is minor, which major and which will turn out to be lethal if left unchecked long enough? Before AF447 it seemed that UAS is somewhere in the middle of the seriousness scale as pilots have successfully coped with it. There was no rush to change the offending pitots.

There will certainly be proposals for improvement of flight recorders, however, what his eyes saw will be pretty straightforward to decipher. What he believed he was seeing, not so.

Nil. Even worse: four crews that passed through ordeal did not recognize they have unreliable airspeed at all. Two were undecided whether it was UAS or not.

Roll oscillation before stall was of low frequency, low and decreasing amplitude. It spells: insignificant and irrelevant. After stall all bets regarding roll control are off, unless you happen to fly some extremely aerobatic aeroplane, which A330 is not. Even if notion that preoccupation with roll precluded control in pitch were true, pilot unable to control the aeroplane around two axes simultaneously is severely incapacitated.

Fully correct. However, the fact that many more crews handled the situation in which aeroplane was thrown in their laps at altitude than not, will be used by those whose agenda involves playing the blame game by moving the focus from organizations to dead pilots.

I was lucky that my head of training, while I was working on my CPL, was a fellow who made it from gliding through cropdusting and night mail to position of chief pilot on DC10 fleet, only to have his career cut short by the flag carrier he worked in going down in flames, together with the country it served. He understood instrument flying very well, we started I/F training on FNPT with basic attitude+power, eventually progressing to VDF approaches, PAR approaches, ILS to stop on runway, UA recovery on partial panel, culminating on recovery with just vario and magnetic compass working. He was very keen to make us understand that while what we learnt might save us one day when situation gets really desperate, we are not supposed to get overconfident and paint ourselves in corner e.g. just because you can hold perfect raw data ILS below 200 ft QFE doesn't mean you should.

I don't think it would be much of use. It might uncover a pilot or two too lazy to know memory items or recognize what procedure to apply when the ECAM crutch gets broken but a pilot who is fully aware "it's only a simulator" and only on the dark, stormy night is fully hit with the realization that air is not a friend and that his life is at stake every time wheels leave the ground would slip under the radar.

Hindsight or just human behaviour
 

OC, re #1174, I agree with your thoughts on expectation, but any conclusion that the crew should have followed a specific drill (SOP), because it was the obvious course of action, involves hindsight bias.
It is not the physical presence of a drill which is important; it is the mental process which decides to use it. We can establish the presence of a drill and associated training, but not what the crew thought, or what might have influenced their thinking – the effectiveness of the drill / training.

“…the SOPs are there to be used by flight crews …” yes, but when the crew has understood the situation as requiring that specific drill, and thus the drill is selected. Inferring that this is obvious judges the crew’s awareness and decision making after the fact – this is hindsight bias.
“… and therefore to not use them is a demonstration of an incorrect response to the situation”, yes this is one possible hypothesis, another is that the crew did not understand the situation and thus ‘chose’ not to use the drill – it never occurred to them (see ref).
We can construct many hypotheses from the information gathered after the fact, which can be used as valuable tools for investigation and safety response, but instantly a hypothesis is taken as ‘fact’ without evidence, this involves hindsight bias.

In this instance, the bias is our assumption that the crew understood the situation as involving UAS, and thus did not follow the appropriate drill.
It is just as plausible that the crew’s actions were entirely consistent with the situation as they saw it; but we don’t know what that was and currently can only speculate on a rage of alternatives.

Unfortunately, the vast majority of the thread discussion is based on a biased starting point, although to be fair some contributions have qualified their points as hypothetical.
There may be some significant safety issues arising from these hypothetical explorations, particularly in the process of awareness in the first few seconds of the event (not the subsequent and more salient stall condition – which also the crew may have not been aware of).

Some significant contributors have been identified:-
UAS drill formulation; is the title relevant, will the crew know when to select it. These are issues of procedure formulation and industry wide communication.
How to identify UAS situations (involving system malfunction): what are the key features, were these explained in training and associated with the range of recovery procedures. These are training issues, but also aspects of memory and recall in context.
Human behaviour in sudden and surprising events; what might effect perception and choice of action, - Human Factors.
Are these a source of the problem or a solution?

Errors in Aviation Decision Making, Orasanu & Martin.

Hi,

clandestino
This air chief pilot quote is in a video (in french) posted earlier in this thread
clandestino
Do you think this behavior normal ?
This is in contradiction with the rules
clandestino
August 2001:
The DGAC mandates the replacement of the sensor Rosemount probes by the
Goodrich and Thales AA by AD 2001-354 (B) (Appendix 3).

July 2002:
In the OIT 999.0068/02/VHR SE (Annex 4), Airbus made ​​the observation of defects
probe Thales (formerly Sextant) P / N C16195AA.

December 2002:
The FAA mandates the replacement of the sensor Rosemount probes by the
Goodrich and Thales AA , specifying the hazard out of the flight and that is the answer to an "unsafe condition" (Annex 39)

January 2005 : Thales launch the project « ADELINE » (annexe 5). Actual air data equipment is composed of a large number of individual probes and pressure sensors. This equipment delivers vital parameters for the safety of the aircraft’s flight such as air speed, angle of attack and altitude. The loss of these data can cause aircraft crashes especially in case of probe icing.

Pilots were so unfazed that they didn't even make reports.
....Clandestino

This is iPad Facto from the releases, and not a clear finding, certainly not expressed in this way.

If this statement is casually accurate, the fault is with the regulator, via the airline. The airline is responsible, in an agency, for the regulations. The Pilots, as representatives of the line, are the visible ones, but to understand "unfazed" the culture at AF is on the hook.


The decision maker acts according to his/her understanding of the situation, and the source of error is in the decision maker's knowledge base or in the process of reaching a decision. ....Orasanu et al

With respect, this is incomplete. It is a poor judge who lays off a decision on knowledge alone. Including "process" does not help.

One can have excellent knowledge base; without experience, it is not only partial, it is dangerous to believe data can fly. Untrained, U/A recovery is a crapshoot, and UAS recovery the same. Machinbird has demonstrated this fully, as others have, and does one doubt the lack of training will remain unaddressed in the report?

Has anyone here had a singular experience in the sim, without help, in UAS recovery, though conversant in the platform's operation? Dozy?

ALF
I see your point but can't find myself in agreement with it. We do not have any evidence that the crew actually followed any procedure to identify what the cause of the initial situation was. This would seem to go against their training. Hindsight bias would appertain if for example they chose an SOP other than UAS and we state well they should have followed UAS instead. However, to say that they should have followed a course of action to analyse the situation which fits in with what we might reasonably expect was trained does not show hindsight at all. One should have expectations that trained behaviour should be followed and in this situation we can't say that this was the case. Note I am not saying that they should have followed the UAS drill but that they should have followed an analysis procedure which allowed them to arrive at an appropriate conclusion about the situation they were in. This is what airline pilots are supposedly trained to do. I cannot see any hindsight bias in this.

We do not have any evidence that the crew actually followed any procedure to identify what the cause of the initial situation was.

We also have not seen a procedure to follow.

ALF posted a link to a very appropriate paper. The gist of it is in 3 key paragraphss.

I can see AF447 as Situation Assessment error leading to a Course of Action error (Creative). The initial roll oscillation figures into this.

Quote:
Originally Posted by Machinbird
Hey Clandestino, we are talking about trained pilots, not random guys you pick up on the street.

Clandestino
Where did you get that about Yuri Gagarin? The Russians to this day do not really know what happened to his aircraft other than both guys rode it in.

As for Elliot See flying into the building--how many circling approaches do you think he had made in his life? (I'll bet very few.) His initial approach as a section of two aircraft was FUBAR and his wingy (Gene Cernan) lost sight during the subsequent circling approach and wisely pulled up into the clag for another go on his own. It seems as if visibility was worse than given to the crews and was not suitable. The whole circling thing was improvised on the spot from all appearances. Is it any wonder it didn't turn out well?

Quote:
Originally Posted by Machinbird
One thing that demonstrably took much of PF's attention during the first 35 seconds after the AP drop was the very significant roll oscillation and the method used by PF to control the oscillation.

This looks like a Clandestino opinion piece. Do you have any facts to back it up? Have you tried it in a simulator? Have you mapped out and analyzed the roll oscillation?

Everything I've read about PIO events indicates that it is like the aircraft suddenly seems to switch from a Dr. Jekyll to a Mr. Hyde personality. What frequencies do you think PIO events tend to occur at? They are relatively low frequencies where the pilot is able to make inputs into the cycle. The only problem is that the inputs contribute to the oscillation.

Read the CVR breakdown.

The comments and such indicate lack of situational awareness not lack of hand flying skills. FO tried to correct, left seater took over, he comments that he can't figure it out, and by this time the captain, probably with his head in the middle comments he can't figure it out either.

Honestly I can't believe no one else is seeing this...but if there was a loss altitude from FL3700 to sea level...someone would be reading off a VSI and decreasing altitudes...nope...comments indicate that the only indication of an impending crash was the Ground prox (pull up pull up).

If I were to guess, and it's just that...lighting strike took out their avionics and fly by wire. AP disconnects as it's not getting reliable information...from where you say? The computer...pilots are now handflying bland tubes, maybe rebooting...they have not attitude indication, much less AS or alt...

Anyone here with Airbus experience flown the standby gyro with no AS or Alt?

Good point. The rote is trained, the thinking is not.