1. Unless Boeing decides themselves to back out of the contract, I don't really see Starliner "failing".<p>I do think that it will be quite expensive if they have to re-do the crew flight test - there'll be a substantial delay while they address the failures and Boeing will have to eat the cost of a full Starliner cycle (including a new service module) along with an Atlas V. I can't tell if the potential profit of the other operational missions is worth it or not in that case.<p>2. From what I understand, Boeing and NASA believe that the cause of the thrusters malfunctioning and the helium leaks is a combination of exposure to corrosive propellant (NTO/MMH) and the heating effect of the "doghouse" which is an insulated structure on the side of the service module that houses many thrusters and is insulated to protect the systems from low temperatures by being exposed to the vacuum of space.<p>3. IMO, most of Starliner's failures have stemmed from a reluctance from Boeing to embrace integrated hardware-in-the-loop testing. Instead they seem to rely on component level testing and analysis. The problem with that approach, of course, is that it's pretty bad at catching unknown unknowns.<p>If Boeing had opted for a in-flight abort test (like SpaceX) instead of doing a pad abort test, they would probably have caught the timer issue that plagued OFT-1.<p>Likewise, if they had fueled up Starliner and let it sit for an amount of time that is likely to occur pre-launch, they would have caught the valve corrosion issues that delayed OFT-2.<p>You can see some of the details about the doghouse here[1]. As you can see, it appears like all of the testing was done with the insulation off and many of the systems stripped out of the doghouse. It seems likely that this overheating issue would have been caught much earlier if they had conducted more integrated testing.<p>---<p>1. <a href="https://www.reddit.com/r/Starliner/comments/1eiggns/boeing_cst100_starliner_crewed_flight_test_cft/" rel="nofollow">https://www.reddit.com/r/Starliner/comments/1eiggns/boeing_c...</a>