Originally Posted by
srm.dashboard
Traditional aviation safety is built on 'Preservation'—preventing any failure at all costs. SpaceX, however, uses an 'Iterative Failure' model. From an SRM perspective, they aren't ignoring risks; they are simply moving the 'Acceptable Level of Safety' (ALoS) boundary. They treat a prototype explosion not as a safety breach, but as a high-speed data collection event.
The real challenge for the industry moving forward will be: how do we integrate this kind of 'Rapid Prototyping' safety culture into traditional operations without compromising public safety? The data they are gathering on structural resilience and engine reliability under extreme stress is going to rewrite the textbooks on 'Machine' factors in the 5M framework.
This is not a whole lot different than the way the US treated the Atlas program in the late 1950's-early 1960's. They were building Atlas missiles by the hundreds, launching one or two to see what went wrong, figure out how to fix it, then repeating.
I didn't know this at the time, but the failure rate of the Atlas booster during the 4 manned Mercury-Atlas launches was
still ~33%. Those astronauts really did have 'The Right Stuff'...