One wishes to learn other people's best practices. One also wishes to avoid
repeating other people's mistakes. The problem is that until they were
identified as mistakes, the people who made those mistakes identified them
as part of their best practices.
How can you tell which of other people's practices are best practices, and
which are mistakes? You read their mishap reports. How can you tell in
advance of a mishap? Aye, there's the rub. When you add in cultural
mismatches and nomenclature problems, you find that - given how robust it
makes your safety culture - it's cheaper to make the smaller mistakes
yourself than to try to adapt someone else's fixes to your system.
First make it safe.
Then make it work.
Then make it work better.
-R
On Friday, September 2, 2016, Henry Spencer <hspencer@xxxxxxxxxxxxx> wrote:
On Thu, 1 Sep 2016, David Weinshenker wrote:
But yes, this is a class of accident that is pretty much obsolete,
How many rockets had NASA & Co. (wasn't the Atlas more of an "Air Force
project" than a NASA one at that point) fired by the time they had their
last such event? How many has SpaceX done so far? I 'spect they'll be
climbing the learning-curve a bit faster.
Possibly, but the question is whether it was part of the curve that they
should have climbed already. :-) To some extent, that early experience was
as much "learning to handle sizable rockets" as it was "learning to handle
Atlas", and the former part should be transferrable to other rockets. And
in fact has been...
Bear in mind that almost certainly, nobody who saw the Atlas explosions of
1959 was still active for the commissioning of Atlas V -- which has
essentially nothing in common with a 1959 Atlas, despite the name -- and
nevertheless there has been a marked lack of similar events for Atlas V.
Nor do the organizations that did Atlas V really have much in common with
those that did 1959 Atlas, despite some other similarities of name; the
extent to which expertise rests in an organization, as opposed to its
people, is much exaggerated. (In fact, such "we're the organization that
did Apollo" mythology can lead to overconfidence and complacency.)
So it's comforting to think that SpaceX is at least making its mistakes
faster, but it's still reasonable to ask whether those mistakes should have
been avoided by learning from the experience of others.
Henry
P.S. Yes, in fact, in 1959 Atlas was *entirely* an Air Force project. NASA
was just starting to get peripherally involved because of preparations for
Project Mercury.
