Dr Charles Camarda, a former NASA astronaut and engineer, put the Artemis II catastrophic failure probability at 1 in 20 before launch and called the modified trajectory "playing Russian roulette" 1 2. He told NBC News the heat shield was "deviant" and that NASA did not understand what caused the Artemis I failure. The mission succeeded, which means Camarda was wrong about the outcome.
He may still be right about the underlying engineering: NASA has already ordered a redesigned Artemis III shield with altered billet loading and greater Avcoat permeability 3. Howard Hu set a public benchmark the day before splashdown that the post-mission press conference did not address. The post-mission silence leaves Camarda's assessment neither confirmed nor refuted by data, only by mission survival. The OIG audit (IG-24-011) shows survival is an insufficient test: the bolt melt scenario it documented would produce a fatal re-entry orientation without triggering a loss-of-signal alarm.
The disclosure that would settle the question sits in a Kennedy Space Centre laboratory. Until KSC publishes its heat shield findings, the 1-in-20 estimate belongs in the category of unchallenged pre-mission critique. That is a category NASA has not previously had to manage: a named engineer, with relevant credentials, on the record before launch, whose concern the agency cannot yet address with data.
