People forget that on Apollo 13s return there was a legitimate concern that the explosion had damaged the heat shield and the fear was it would fail and the craft would melt and break up entirely during deorbit.
Due to their nature they're sensitive and easy to compromise. Which is also why NASA knew from day 1 that icing was going to be a huge issue on the shuttle. They were retrofitting the vehicle, the launch platform and even the software to reduce heat shield risks very early on.
Some of their earliest flights included EVA experiments where the astronauts translated themselves to the thermal tiles, did inspections, and even did mock repairs to test the feasibility of the idea and of the quality of the adhesives in near total vacuum.
After the final accident they started doing something they could have and should have been doing since the beginning. That was simply taking up a camera that could be attached to the Canadarm, swung "underneath" the shuttle, and used to take a comprehensive survey of the thermal management system immediately on orbit.
In any case, point is, human rated space flight will always require this level of attention to detail and ongoing effort to derisk every possible aspect of every mission performed. NASA management did an outright terrible job at this. From incentivizing the wrong behavior, falling in love with paper targets, and completely failing to audit their own internal risk estimations for errors.
> they started doing something they could have and should have been doing since the beginning. That was simply taking up a camera that could be attached to the Canadarm,
First shuttle launch, Columbia STS-1 12 April 1981.[0]
The Canadarm was first tested in orbit in 1981, on Space Shuttle Columbia's STS-2 mission [1] (12 November 1981)
It was always intended to be on the shuttle. It's not as if they conceived, designed, and then created it between STS-1 and STS-2. It was baked into the software and into the rear flight deck controls. It was late.
Launching a prototype rocket with the expectation that it will probably blow up and then having it blow up isn't a failure, especially when the goal is to see what happens.
You said "unexpected catastrophic explosive failures". But
1. They were not unexpected. They very very clearly communicated months ahead.
2. "Catastrophic" is a bit much too, as they were indeed expected and planned for. In fact, the biggest failure in the Starship development was that the rocket did NOT explode fast enough once.
3. "Failures". Well.. no. These are prototypes intended to learn from. Experiments if you will. A scientist that never has a negative result is a fraud. Same here.
Then it killed 7 astronauts.
Then it worked again they said.
Then 7 more died.
> Also, are we all forgetting that within the past year SpaceX launches have had multiple unexpected catastrophic explosive failures?
Well that's just a straight up lie.