Originally Posted by
bearfoil
IMAO, The one basic and glaring error in the "Philosophy" of fbw as practiced by the folks @ Airbus, is the following:
NORMAL is Rules based, and these rules are explicit. Flying is situational, and the computer doesn't have the capability to program for all of them.
To me, the one difference in FBW "philosophy" between A and B is hard vs soft limits / protections. "Rules" is not the difference - the rules are there
in both cases.
Approach the limits in A and the system starts ignoring your command, and/or takes other action to bring you back into the envelope - "I'm sorry, Dave. I'm afraid I can't do that"
.
Approach the limits in B and the system starts to fight you - more like "sorry Dave I don't want to do that, you'll have to pull harder to make me".
It's still a rule. Moreover, it still has the
potential to go wrong in very interesting ways if the system loses knowledge of how fast it's going or which way is up. Just like A, B will also in some cases take uncommanded action to keep you in the envelope - rules rules rules.
And rules can lead to wrong decisions, eg. after A/P and A/T turned off:
autothrottle activated and automatically advanced the thrust levers when it sensed a low-speed condition as a result of erroneous data being provided by the ADIRU
A or B ? Surely that's the A "rules" philosophy ? [Actually it's not.]
Rules and hard limits/protections have been part of aviation engineering before FBW and extend far beyond it. Want reverse thrust in flight ? No can do. Why ? It's a rule. Why ? It makes big smoking holes in the ground, so we made the rule. What if I need it ? We can't think why you ever would. But you can't think of every eventuality at design time...
And then of course the rule can end up going wrong - put the a/c down ever so lightly on a wet greasy runway, fail to trip to sensors and... no reverse. Oops.
The rules aren't perfect - fbw or other. They won't ever be. The aim is merely to cause
fewer big smoking holes in the ground overall - and even that is very difficult to assess in the stats when the accidents are so few. Every rule, every protection or limit, every safety feature
will kill someone, in some scenario, sooner or later. Car seatbelts and airbags kill people, reliably, every year - doesn't mean they are a bad idea.
A belief is a conclusion based on data. Faith is a format based based on Intuition, and "feel". When my beliefs are dashed, I rely on my Faith, which is in me, and my F/O.
The Wall of misunderstanding twixt the engineer and the Pilot will keep killing people until both understand the other. Please stop stubborn Pride from putting people at risk.
Faith is also belief in the absence of data/fact or contrary to it.
Interesting paper on A vs B approach (GPWS escape):
http://www.rvs.uni-bielefeld.de/publ...9_CFIT_FBW.pdf
Conclusions (as I read it):
- The A approach, hard limits, stops you getting that last bit of performance that might save you, sometimes
- But the A approach gives more reliable escape performance most of the time
- The B approach might be better overall given a perfect pilot
- But for real human pilots, the A approach will save your arse more often
Yet the test pilots preferred B.
Why ? "Faith" - in their abilities, even when contrary to the data ? Dare I say "stubborn pride" ?
And of course the final irony is that we are discussing this in the context of an accident where those rules and protections you so dislike didn't kick in, but rather packed up and left. The cyber rules engine, deprived of its senses, took itself out of the loop, and handed full unprotected control to the meat...