I have enjoyed remaining a silent lurker throughout this thread. Until now that is... so stand by for yet another non-pilot sounding off, albeit in the hope that he won't be knocked on his ass for doing so.
My field of study was Psychology, and whilst I don't presume to judge what occurred in this crash I do have a couple of observations to offer from my own perspective.
Human cognition (taken as the psychological result of perception and learning and reasoning) is complex and fascinating. At times it can be at least as counter-intuitive as any other complex system, even though it has been endlessly refined towards human survival. There is no more effective demonstration of what an inside knowledge of human cognition can achieve than to see a good up-close magician - through deliberate mis-cues and misdirection an apparently intuitive situation can become bizarre, with the impossible apparently occurring or the inevitable failing to occur. In such situations you arrive knowing someone will try to fool you, on your guard, but still end up bamboozled.
Effective cognitive misdirection is typically founded in extremely familiar scenarios or easy tasks, in which confusion is created between cause and effect. When the magician urges you to check very very carefully that the cup is empty and there is nothing in his pocket, it's because the ball is pinched between two fingers at the back of his hand. Yet the more carefully he gets you to check, the more amazed you are when the ball appears. A secondary part of this technique is momentum - the magician will stand very close, keep talking, showing, pointing, setting tasks and suggesting conclusions to draw. There is 'no time to think'.
It's not surprising that experienced pilots are incredulous that an experienced pilot could possibly fail in a safety-critical task which is elementary to every landing, especially given that the same crew had already landed successfully with a rev inop. I'm sure that certainty was shared by the unfortunate crew of the aircraft which crashed. Whatever truly happened, that's the kind of everyday certainty which both cognitive psychologists and magicians know to be the mainstay of misdirection. Without that certainty, the misdirection could not occur, it is what fuels it.
So here's a hypothetical from a psych viewpoint.
Crew learn rev inop. This is fairly unusual and thus gets some attention. Undertake flight, on approach there is the reminder 'rev inop'. Upon landing everything seems fine. It is fine. 'Rev inop' is no big deal.
Next flight. Rev inop. OK, mustn't forget but it's no longer news. The new task (rev inop) is becoming familiar and is being incorporated in a very very familiar task (TL retard at landing). This is where human cognition can offer pitfalls: making alterations to an everyday task which is genuinely, in itself, simple. If a momentary task is muddled through a complex blend of familiarity with unfamiliarity, it is the very simplicity and ease of the task in itself which contributes both to the error and to the difficulty of later diagnosing what has gone wrong.
If we allow that this could occur (not that it is likely to happen on any one flight but that it is conceivable to ever happen at all) we are now in a whole world of classic misdirection. The runway is flagged as slippery, so expectation of this possibility is conscious. Also known is that the runway is short. The TLs are not a concern, they are familiar and proven good, even with the rev inop. On landing, there is an immediate and alarming DECEL problem. Where will attention go? Slippery runway. Length of runway. Now systems are not triggering: spoilers, auto-braking. Why are they not deploying? TLs are in the past, before everything went crazy.
It seems that in this rare situation, auto-thrust is giving an extra push to a fast closing door: not only are braking systems not being triggered but one engine is spooling up. This is a consequence of the TL error which is, in the moment, entirely outside the crews sphere of awareness. Circumstances have misdirected them to lack of braking rather than over-thrust, both because of
1. where in the time line the problems occur
2. the warning of a slippery runway
3. the implicit belief that 'TLs are so easy and obvious that nobody gets them wrong'.
One Pontification (which I am unqualified to make but heck this is only the internet):
If the the above described human error did occur, the Airbus system exacerbated an error for which circumstances left little margin. That does not mean that Airbus are necessarily in error or inferior, but there is surely a case for examining whether the auto thrust / braking logic trees could be refined. Of course, many may believe that this apparent crew error did not occur...
Final Pontification (from a psych viewpoint of which I really am quite confident):
Being absolutely certain you would never ever make an elementary mistake is in itself dangerous.
Last edited by robdean; 13th Aug 2007 at 14:21.