ESA: Mars lander crash caused by 1-second inertial measurement error

[gary]

In an article today (http://spacenews.com/esa-mars-lander-crash-caused-by-1-second-inertial-measurement-error/) at SpaceNews.com, Peter B. de Selding reports that -



Article here -
http://spacenews.com/esa-mars-lander-crash-caused-by-1-second-inertial-measurement-error/#sthash.MqNwH2sq.dpuf

[pluto]

Thanks for posting Gary.

[julianh72]

I'm sure there are good reasons why they programmed the lander such that a single instrument feed was used to trigger the parachute release and thruster firing, so that when that single data feed became saturated, the landing sequence was triggered prematurely. However, it does seem to be a contravention of long-standing design practice in the aerospace industry that you use multiple redundant feeds / processors, with independently designed systems and sensors for critical functions, and in the event of a discrepancy, you use a 2:1 voting majority to decide what to do next. (E.g. if one sensor says "We've landed!", but the other two say "No we haven't!", you conclude that you probably haven't yet reached the ground.) It's not exactly rocket science. (Well, actually, in this case it IS rocket science!)

As ESA Says - the main objectives of the lander project were to acquire understanding for future lander projects - one of those learnings should be to not forget the learnings of the past!