Space

Aerospace industry engineer here:

We try to identify failure modes and use tools like Failure Mode Effect Analysis (FMEA) and fishbone analysis to track down failures and how they cascade to understand system behaviors. However, the more you increase the complexity of the system, the more difficult it is to fully think through all the possible ways things can go wrong and it's not unheard of for things to slip through review.

Starliner has consistently been plagued by program management issues where they think they've caught the failure modes and implemented appropriate mitigations. They do an analysis, run some tests to prove those assumptions are correct, and fly it. In this case there was a design flaw in the thrusters that they saw on a different test flight, thought they fixed it, and flew again not knowing that they didn't actually fix the problem.

False sense of security is a dangerous place to be when it comes to fault scenarios, but the alternative is extreme paranoia where you trust nothing. In fairness to Boeing, taking some level of risk is necessary in the space industry but I think it's pretty obvious that they were not paranoid enough and were too trusting that they did their job right

repeating, rehashing and rephrasing

Remember when writing was Prose and poetry had rhythm?

It's okay to say words, a few times and in different order, for effect.

There was a lot more to the article than this. I've sat on console during launches,and reading their exchange gave me some anxiety. Trying to live troubleshoot thrusters issues would be a nightmare on an unscrewed satellite, let a lone one where it's human rated, and the people flying it could die if you are unable to recover fast enough.

You train for this crap excessively, so everyone knows what to do, but that doesn't make it any less nerve wracking in the moment.