I just watched the Nat Geo reconstruction of the THY AMS stall crash. There was a comment from an eminent lady in USA who is an expert on human behaviour in automated environments. She declared that humans are very bad monitors of automation yet that is what a modern airline pilot needs to be. The solution is that NASA is researching new designs of automation to make the human interface more intuitive and easy. Better displays that are easier to understand; more accessible data/info, i.e. not having to page through CDU's to find the piece of info you want, but at a single touch of a button; voice activated equipment etc. etc.
This was in response to "why did the crew not realise their airspeed was decaying?" They did not know that the RA fault would cause the thrust levers to go to idle and they expected them to do that anyway as they were attacking the G.S. from above and trying to slow down. They were in monitoring mode, but not on top of it.
Well. if I'm attacking the G.S. from above and close to the runway, and trying to configure I can assure you that I am not passive and relaxed watching the needles & dials do their thing, but I am in a heightened state of alertness and making damn sure the autos are doing what I want and the a/c following suit. If it ain't then I know what to do about it and if all else fails I will disconnect and fly the sucker.
Once again the solution from the technocrats is to try and make a better human automatic interface; introduce more warning systems; make the automatics more intuitive etc. etc. No-one has suggested better flying training; better understanding of the a/c envelope; better understanding the capabilities of the automatics - their limitations and traps; better hands on skills to put the pilot back in touch with the a/c. The solution to pilots being lousy monitors and watching the automatics crash the a/c is to address the automatics rather than the pilot.
I'm sure the solution lies in a combination of both. What is certain is that there should be a lot more training of how to best use the automatics then we have now. The understanding of them which I witnessed on the line in many companies was very scant. The TQ courses and LT is focused on using the automatics to stay well inside the SOP envelope. "If you do that you are safe," is the mantra preached every prof check. Incidents happen when people are outside their comfort zone which is bounded by their knowledge envelope. A survivable incident can become an accident because the human intervened in the wrong way due to ignorance. Either ignorance of what is going on, or ignorance of what to do about it. You can make all the fancy gizmos you want to give the pilot lots of info, but they still have to know what is going on and what to do about it. SOP's can not cover all eventualities. Back to basics has to be an option.
We are debating this due to a few well published crashes. A disturbing number in such a short period; and from various cultures in various different airlines and on different types. In other words the common thread is the behaviour of the pilot. What needs to be known, to assess the extent of the problem, is how many similar incidents have been near misses and we never hear about them. Just how deep is the problem? I suspect worse than we dare imagine.
With regard to THY stall at AMS. The program said Boeing had received >2000 incident reports of faulty RA's. Numerous reports of a sudden thrust to idle. No other crashes; the pilots intervened and took over. There was no comment about why the difference. Could it have been different training qualities? Whatever; should it really mean that 1 accident out of 100's of incidents needs a rethink the whole automatic/human interface rather than address the issue purely from the human angle? The technocrats love to do it that way, perhaps a few 'old farts' in the training departments could achieve what is desired.