Jean-Raymond Abrial talks to us about the use of B at the RATP (excerpt from his lecture at the Collège de France) /1
During the 80s, I had contacts with the RATP, which planned a semi-automatic
metro on the RER line A and which had developed a very important embedded system.
But at the last minute, the managers had to make the final decision.
I was contacted by Claude Hennebert, RATP engineer who is an extremely
interesting individual because he was at the RATP and he was neither an ENA
nor a polytechnician, he was blocked. It was someone who was high enough
in the hierarchy so close enough to the taps but at the same time who had kept
a very good technical mind, so he contacted me and he told me "We would like
to do a technical audit". Lasted three weeks. I met some very interesting
people and had to answer questions.
Are the means used to ensure that the product does it correspond to its
initial specification?
So I do my study and then at a big meeting, there were people from the RATP
and also the manufacturers who manufactured this system and I said:
"I can't answer because I haven't seen the specifications, I haven't
seen anything before." People were mad with rage, mad and they told
me "it's not true". But Show it to me! And even I didn't see a design
document, a conception document, i.e. that people had programmed in the end.
All right, I wasn't wearing a judgement on what had been done, but I could
not answer the question. In the end, the project was delayed by a year,
probably not only because of what I said, but for the RATP to delay a project
for a year was a tragedy because the RATP is considered the best metro in
the world. which is true. [...] "
"There was no specification, so give us a Specification Course". So then I
was a little shaky because I had to I built a specification course that
worked so well that in the mid-80s the RATP decided to embark on a driverless
metro, you all took it, the line 14 and they were convinced by B to develop
the critical parts of the software And they also put their hands in the dough
and money to develop a tool" Atelier B", entirely financed by the RATP and
it is the tool that will make it possible to to do all the refinements and
all his proofs. And the RATP decides, and there again I think that this
Mr. Hennebert was really an extraordinary guy.
But at the last minute, the managers had to make the final decision.
I was contacted by Claude Hennebert, RATP engineer who is an extremely
interesting individual because he was at the RATP and he was neither an ENA
nor a polytechnician, he was blocked. It was someone who was high enough
in the hierarchy so close enough to the taps but at the same time who had kept
a very good technical mind, so he contacted me and he told me "We would like
to do a technical audit". Lasted three weeks. I met some very interesting
people and had to answer questions.
Are the means used to ensure that the product does it correspond to its
initial specification?
So I do my study and then at a big meeting, there were people from the RATP
and also the manufacturers who manufactured this system and I said:
"I can't answer because I haven't seen the specifications, I haven't
seen anything before." People were mad with rage, mad and they told
me "it's not true". But Show it to me! And even I didn't see a design
document, a conception document, i.e. that people had programmed in the end.
All right, I wasn't wearing a judgement on what had been done, but I could
not answer the question. In the end, the project was delayed by a year,
probably not only because of what I said, but for the RATP to delay a project
for a year was a tragedy because the RATP is considered the best metro in
the world. which is true. [...] "
"There was no specification, so give us a Specification Course". So then I
was a little shaky because I had to I built a specification course that
worked so well that in the mid-80s the RATP decided to embark on a driverless
metro, you all took it, the line 14 and they were convinced by B to develop
the critical parts of the software And they also put their hands in the dough
and money to develop a tool" Atelier B", entirely financed by the RATP and
it is the tool that will make it possible to to do all the refinements and
all his proofs. And the RATP decides, and there again I think that this
Mr. Hennebert was really an extraordinary guy.
Comments
Post a Comment