PPRuNe Forums - View Single Post - TAM A320 crash at Congonhas, Brazil
View Single Post
Old 6th Aug 2007, 10:23
  #1214 (permalink)  
SoaringTheSkies
 
Join Date: Aug 2005
Location: Cloudbase
Posts: 149
Likes: 0
Received 0 Likes on 0 Posts
ELAC,

yes, I was polemic with this post.

To err is human, it's a fact that we both seem to agree on.

Most human-human communication has enough redundancy to allow for errors.

If I, as an engineer, am building a system which communicates with a human, I need to obey three rules that I have outlined twice before:

1. avoid error as far as possible (by human centric interface design)
2. tolerate error where possible (make sure there's enough communication redundancy so that a single human error cannot render the whole system clueless)
3. fail gracefully if #1 and #2 did not work

I do not want to fingerpoint every human error to the manufacturers, or, more generally speaking, to the designers of a certain system.
But I do want to point out that the AB system has imho failed on all three levels.

1. avoid error: due to the fact that the thrust levers are stationary, there's a chance that the PF does not move them, however outlandish and improbable that may be, it has happened here and it has happened before.

2. tolerate error: the "thrust levers in or near idle" clause is a hard inhibitor to ground spoilers deployment and thus indirectly to autobrake. A second clue to the system could be beneficial here (this has been called a "red button" or "override function" in this thread, but technically more reasonable would be, for example, to use manual brakes as an override clue for the gs logic)

3. fail gracefully: this is a difficult bit. #1 and #2 try to catch things that you can envision. #3 has to deal with things you have not thought of and do so without jeopardizing other scenarios - thought of or not.
I won't make a guess what that would be, but I think we can agree that the system has failed far from gracefully in this case.

So please excuse my polemics, but I just can't take the implied "bad airmanship" as the sole answer to this, especially not in the light that it's not the first time this same "bad airmanship" error has occurred.

I hope that anyone who posts here finds himself in a forgiving environment when he makes the next mistake.

pj
SoaringTheSkies is offline