Re: RTO Decision made by the computer

Date:         07 Oct 98 02:49:18 
From:         "Jim Sokoloff" <>
Organization: D. E. Shaw & Co., L.P., but I speak only for myself.
References:   1
Next article
View raw article
  or MIME structure

Thomas Buro <> writes:

> On a conference about the design of human-machine systems,
> i.e. Cockpit design the issue of Rejected take off was discussed. A
> professor could mathematically prove that decisions shortly before
> V1 is reached are better made by a computer.

The computer's decision can only be "proven" better in the context of
the information it is being fed. Feed it bad data that a human would
correctly interpret, and I can "mathematically prove" that a human
will fly better. Humans are better at judgement than computers; we're
less sensitive to blindly following bad data.

> The scenario was: one engine failure, no further problems etc. In
> case of engine failure and a wrong decision of the pilot the
> computer would fly the aircraft.

> What do professional pilots think about that issue? Would you like
> such a system, which is taking control or would you like to have
> warnings from the computer what is the better decision ?

I'm a (private) pilot AND a computer programmer, and you can bet that
I would rather a human be making the top-level decisions about whether
or not to reject a takeoff. Computers are good at some things (fly by
wire, and doing math quickly), but give me a good old-fashioned
gray/bald head up front, please.. :-)