Hacker News Reader

Comment by aeternum | Hacker News Reader

>Camarda is an outlier. The engineers at NASA believe it is safe. The astronauts believe it is safe. Former astronaut Danny Olivas was initially skeptical of the heat shield but came around.

How do you explain so many people believing it is safe?

The problem is risks are far too easy to brainstorm, anyone can come up with endless risks that it takes endless time to mitigate.

If I were the manager for challenger, I would have run the o-ring experiment as soon as it was brought up as a concern. Put the fuel pumps in a freezer, test if they leak. Feynman famously demonstrated it with a glass of icewater. Experiment is what separates made up risks from real risks, I would have definitely told the engineers to take a hike and would have hit launch if they couldn't provide experimental evidence of o-ring failure in cold temps. (Spoiler alert: in that case they easily could have)

View on HN

No. That famous demonstration only touched on the real failure mode--the rings were covering up other failure and in the cold could not do so.

The real test was creating a full-scale test of ignition, an engine containing mostly inert filler (to occupy the fuel volume) and just enough fuel to reach stable burning.

> How do you explain so many people believing it is safe?

The article itself answers this question: institutional incentives leading to heavy social pressure to agree with the groupthink and declare something is safe when it is not. And we know that the scenario it lays out is highly possible, because it has already destroyed two Space Shuttles. Now that this has happened twice, the burden of proof is on the people saying it's not happening again, especially when the OIG's report directly contradicted what NASA had been saying about the heat shield up to that point (indicating they were lying and had to hastily retcon their story).

>the burden of proof is on the people saying it's not happening again

This specifically I take issue with. You had a bug in your software before so now the burden is on you to formally prove your software is bug-free.

The burden of proof should remain on the naysayers. Take a plasma torch to the heatshield pock marks and see how long it takes to burn through. Do experiments just as Feynman did with the o-rings. Let the outcome of the experiment, not office politics decide.

I'd say when two conditions are true:

1) you have an established pattern of behavior of ignoring safety concerns (Challenger, Columbia), and

2) people are alleging that you are doing the same thing now, with independent auditing from the OIG backing them up,

that's sufficient to shift the burden of proof back onto you.

Your attempt at a gotcha with the heatshield is just ridiculous: everyone already agrees the heatshield works in small-scale testing. That's the entire problem! It failed on the actual mission and NASA couldn't explain why, so instead they pivoted to trying to explain why the failures don't matter.

(EDIT: As an addendum, I'll also add that you don't even need to go back to Columbia to find an example of NASA lying about safety to protect reputations. Remember when they insisted for months that the Starliner mission was going just fine, and then eventually they said the astronauts weren't coming back on it, and then it landed and the final report was that there multiple failures leaving it on the knife edge of total catastrophe? And remember how that was less than two years ago? You're a maniac if you take the safety claims of this organization at face value)

Reply on news.ycombinator.com