3 ms·

>Tesla was a place with low engineering quality SpaceX has had a somewhat similar reputation (at least early, it may have changed for the better). For example

by bumby 4y ago

>Tesla was a place with low engineering quality SpaceX has had a somewhat similar reputation (at least early, it may have changed for the better). For example, they had a spectacular failure related to bad material quality in a strut.[1] It's common practice in aerospace quality to have "coupons" kept, tested, and traceable for these types of critical components. Apparently, SpaceX wasn't follow this procedure and they changed their policies after the mishap to do so. There have been some other deviations from standard practices that may lead one to raise an eyebrow. I've sometimes wondered if these are examples of an attitude of "move fast and break things" creeping into other (generally more conservative) safety critical domains. [1]

pbronez 4y ago
Interesting. My impression is that SpaceX has delivered some extraordinary results by throwing out a lot of conventional wisdom and starting over from first principles. This anecdote suggests they’re humble enough to re-adopt standard practices when experience shows their value. Seems like a good approach… if the cost of failure is manageable.
- bumby 4y ago
  What, in that example, do you think is the first principle they were working from? It seems to me that they were missing the principle of quality of “trust but verify.” This is a risk to SpaceX that I don’t think gets talked about much. They are able to take risks and throw out conventional wisdom to streamline, but sometimes it’s easy to conflate being lucky with being good. But each time their luck goes bad, they lose a little bit of that edge by having to layer on bureaucratic process (like additional quality checks) to mitigate that previously unrecognized risk. Do that enough times and you start to look just like the dinosaurs they’re replacing. If the lose a human life, it may come even faster.

by bumby 4y ago

pbronez 4y ago

Interesting. My impression is that SpaceX has delivered some extraordinary results by throwing out a lot of conventional wisdom and starting over from first principles. This anecdote suggests they’re humble enough to re-adopt standard practices when experience shows their value. Seems like a good approach… if the cost of failure is manageable.

bumby 4y ago
What, in that example, do you think is the first principle they were working from? It seems to me that they were missing the principle of quality of “trust but verify.” This is a risk to SpaceX that I don’t think gets talked about much. They are able to take risks and throw out conventional wisdom to streamline, but sometimes it’s easy to conflate being lucky with being good. But each time their luck goes bad, they lose a little bit of that edge by having to layer on bureaucratic process (like additional quality checks) to mitigate that previously unrecognized risk. Do that enough times and you start to look just like the dinosaurs they’re replacing. If the lose a human life, it may come even faster.