Originally Posted by
RetiredF4
Would it get more unsafe in your opinion with feedback? Why not add feedback and an AOA gauge for aditional safety?
AOA gauge is there
as an option already (I believe) - airlines just don't order it. An extra gauge will make no difference without training to use it - it might as well display phase of the moon. AF apparently couldn't be bothered to train pilots to manually fly in cruise
at all (A/P goes off - learn to fly a new a/c, fast) - what makes you think they would have trained AOA ?
Feedback - well, it might improve things or it might not. Intuitively, yes - more feedback through more channels = better. In pratice, aviation history is littered with the smoking holes left by those who have ignored and overridden stick shakers
and pushers, all the way to the ground.
The 'bus designers didn't just decide to do something different from shaker / pusher, they went for "better" with active protections ["limits" for Gums - but that's just semantics and audience because I sure wouldn't want to be the designer that tells a fighter pilot I'm going to "protect" him by limiting his control authority!].
Unfortunately, in this case (and perpignan) the protections were lost due to technical faults, leaving the warning system that is
maybe less good than the old stick shaker. So question should be, is the overall system - active protections degrading to aural warning at 1 in 10k flight hours - better or worse than "stick shaker only" ?
I'd say they made it better. Could it be improved ? - certainly.
Should they go
back ? - not IMO.
That is again a thinking in statistics and probabilities, i wont accept. Any near accident is too much, any accident is a waste. Why not improve things some more despite the relative high safety? Money? Pride? Neglecence?
By the way, its not A vs. B, its make things safer when you know how.
[...]
That is not a training issue alone, it is a problem to tune in the pilot into the system and to keep him in the loop from normal operation to the biggest f****up possible.
And this is also an engineering task, wether you like it or not.
That's agreed. The tricky bit is how to know what makes things safer. There are so many changing variables and so little accident signal buried in the statistical noise that you could find support for almost any change.
Human factors, and particularly human-machine interface is a complex area, and the little involvment I've had with it has given a clear impression of how
counterintuitive it is.
Users of a system (pilots in this case) will happily tell you how they use it and how it needs to work - but record their usage with eye-trackers etc. and you'll get completely different answers. I would expect that to apply even to highly trained pilots and insturment scan - I bet that they weren't looking at what/where they thought they were when it all hit the fan in that cockpit.
The most interesting link posted on these threads (more than once)
was the Nasa study of A vs B control systems for CFIT escape. Covering sidestick, laws protections the lot. Result:
- the pilots overwhelmingly thought B was the better system
- the actual outcome was that the A system saved your ass more often
So which set of designers got it right ? Not easy.