Mars Polar Lander (was NASA declines to protect the Planet Earth)

[Some irrelevant newsgroups dropped, and s.s.h added.]

In article ,
Jan Panteltje wrote:
: ...the Mars Polar Lander crash...

...I think the concensus was that some microswitch that was supposed
to flip when the lander touched ground and then swith off the engine, was
triggered early when the landing gear was put down...
They could reproduce that.

Correct. Through a series of mishaps, this problem got overlooked during
development. (Incidentally, while tight budget was a major contributing
factor in the loss of Mars Climate Orbiter, *this* problem would have
killed a billion-dollar megaproject just as easily.)

An onforgivable screw up, not only should the whole procedure have been
tested...

It was. But the test was difficult to set up, and when it hit some
problems in the middle, everyone agreed that it was preferable to stop,
fix the problems, and restart where it left off, rather than starting over
from the beginning. Unfortunately, that restart had the subtle side
effect of resetting the spurious "touchdown seen" flag, thus hiding the
problem.

but also the teams did have no communication, just got a
'this switch will come on if it touches ground' to work with, while the
other team just installed the switch?

Communication wasn't great, but in principle there was enough. Alas,
there were some little complications, like a key person having to miss a
major software review meeting. Some of this was just sheer bad luck.
--
spsystems.net is temporarily off the air; | Henry Spencer
mail to henry at zoo.utoronto.ca instead. |