From: (Robert Dorsett)
Organization: Netcom Online Communications Services (408-241-9760 login: guest)
Date:         22 Jun 95 03:07:37 
References:   1 2
Next article
View raw article
  or MIME structure

In article <airliners.1995.830@ohare.Chicago.COM> Airam J Preto <> writes:
>I think this thread is getting filled with misconcepts and prejudices.
>The only type of airplane which is TOTALLY under pilot's control is the
>"fly-by-cable" type :).
>The rest is a question of good training and maintenance.

That is an oversimplification of the issue.  Even with a hydraulic-based
system providing artificial feel, there is an innate set of physical
principles at work.  There is nothing even remotely approaching the
complexity of the modalities introduced by digital flight control.

As a trivial example, a few years ago, a Fokker F-100 landed at ORD.
Their brakes weren't working.  Their thrust reversers weren't working.
Their ground spoilers weren't working.  The crew shut down an engine, and
waited for the headwind to slow them down.  They did.  They exited the
runway via a high-speed exit, managed a U-turn up the parallel taxiway, and
started to pick up speed.  They took another exit, and entered a perpendicular
runway, which was in use at the time.

When it finally rolled to a stop, they determined that the FMS was stuck
in the "air" mode.  Upon subsequent inspection, each air/ground switch was
covered in ice, stuck in the "air" decision.

Their luck, and the skill of the crew, stopped this from adding to USAir's
dismal safety record over the last five years.

The digital logic involved had turned off normal ground braking functions--
restricting pilot authority to brake the airplane.

This in itself is not too unusual: spoilers are regularly inhibited depending
upon the position of A/G switches in conventional airplanes, as are thrust
reversers.  But in this case, the brakes were, too.  All A/G functions were
trivially encapsulated, and disabled with a trivial conditional.

*Electronic* flight control systems are rife with these modalities.  There
is no open pilot loop, which provides natural, honest, and consistent
feedback to the pilot throughout the flight regime.

So, the point here is that new technology introduces new problems, including
unforeseen or "extremely unlikely" occurrances, such as the F-100's little

So the question, then, is whether you, the engineer, feel you have a good
enough handle on the problem definition--which, recall, is in a highly
hostile operating environment--to restrict pilot authority?  Or do you
provide guidance and feedback to the pilot, without overriding his authority?

Boeing chose the second approach, which I would call a versatile and tolerant
one.  Airbus chose the first one, which I would call restrictive and author-

Robert Dorsett                         Moderator, sci.aeronautics.simulation