More on Air-Inter charges

From:         Dr Peter B Ladkin <pbl@compsci.stirling.ac.uk>
Date:         04 Feb 93 02:36:34 PST
Followups:    1
Next article
View raw article
  or MIME structure

In Le Monde for Wednesday Jan 20th p15, there are two articles on the Air
Inter crash. One states that the company president, M. Jean-Cyril Spinetta,
has demanded that he be indicted on the same charges as his safety officer, 
M. Jacques Rantet. It seems also that the prosecution is trying to blame
everything on the lack of a GPWS (specifically M. Rene' Pech, the public
prosecutor of the Re'publique de Colmar, where the case will be tried).  
The second is a review and commentary by Alain Faujas.

Blaming everything on the lack of a GPWS is an anticipable legal tactic, since
that is what the Air Inter executives are being tried on.  But it's hard to
see how it would stand up for two minutes in a US court. The NTSB would only
be able to call it a `contributory factor', since the agreed cause is
controlled flight into terrain (CFIT), and when all instruments are working
correctly, the only possible proximate cause of CFIT can be pilot error,
according to the `standard' classification (which has been questioned in these
columns, correctly in my view). I suspect a US court would therefore be bound
to conclude pilot error.

Further, since French law doesn't require the GPWS for internal flights, and
there is as yet no question that the airline was operating in accordance with
regulation, it's a mystery to me how the executives could be *criminally*
liable for any of this. But I don't claim to understand French law, or the
notions of responsibility enshrined therein.

A further little legal puzzle.  Air Inter has since been required to install
GPWS.  The article says that the transport minster at the time of the
accident, M. Paul Quile`s, instructed them to do so.  But they didn't change
the law.  

Does all this mean that flying without GPWS is legal, but people will try to
throw you in jail if you do?

I perceive similarities between the technical aspects of the case, and
those of the Viper case (which Devlin commented was an attempt to ask judges
to decide what is a mathematical proof). It's a little harder to put it in
writing. It's something like:

human factors engineering + software safety questions + system design 
questions + lack of established engineering practice or terminology + humans
in the loop ----> go to court to try to blame someone.

If this is to become the system engineering process model of the 90s, it's
going to put the field back 20 years.

Peter Ladkin

P.S. What on earth is the appropriate word for `anticipable'?