No place for assumption

One always assumes that the people at NASA know how to do rocket science. Goodness knows they have achieved some marvellous things in my lifetime alone. Man into space, man on the moon, probes all over the galaxy and comet intercepts. So when seven atronauts are killed and some insulation hitting the shuttle on launch is suspected as a cause, you'd assume the analysis they did at the time to discount that as a risk was done with a least a decent amount of scientific rigor. Sadly the latest reports say it was done with less of a computational model, and more of an Excel spreadsheet from Boeing.

Now I've seen some pretty amazing things done with spreadsheets and they are perfectly capable of some quiet complicated numerical calculations. But really, when you're analysing something with life or death impact you'd expect the parties involved to have done a little more than punch a few numbers, which in any case turned out to be the wrong numbers, into a spreadsheet and give a thumbs up for landing.

You can give some of the people involved with the decision credit for requesting further investigation, including actually looking at the shuttle from the ground to see if any damage could be observed. That was done before, the first time debris was observed hitting the orbiter on launch. Tragically this request was denied, largely on the weight of the spreadsheet evidence and an assumption that if it happened before and there was no damage then it will probably be okay this time.

I sincerely hope the cause of the shuttle disintigration turns out to be something else, and that whoever made such a bad judgement call can breath sigh of relief. We know there are many more predictable modes of failure that will bring the shuttle down without assumption, the mother of all fuck ups, lending a hand. Given the past experience with O-rings there really should be no room for assumption in a place like NASA any more, just rocket science.