Should Starliner be given another chance?

[DanClemmensen]

Pushing back a bit on the idea that NASA could cancel the Starliner contract 'for cause' with no termination penalty, CFT-1 returned without a crew at the request of NASA. Those would be good contract terms if NASA had gotten them: 'We can tell you not to fulfill your obligation so then we can cancel with no obligation to pay you.'

While I'm in on the "cancel the contract" side of the argument, I agree with your logic here re: termination for cause over CFT-1 return without a crew. I think you could argue that being as late as they are could potentially be cause. But even if not, I think it might be worth terminating even if they terminate only for convenience. But only if they reinvest any savings into starting one or more other groups on developing commercial crew capabilities. Redundancy is nice, but economically viable competition for post-ISS CLD operations is so much more important. IMO, YMMV, etc.

~Jon

The cause is not the failure of CFT. It's the failure to deliver an operational system by the contractually agreed-upon date of 2017. However, everything we have heard from NASA (and the silence from Boeing) makes it clear that if NASA no longer wanted Starliner, then all they need to do is decline to certify it until after a successful CFT that Boeing needs to pay for. NASA does not actually need to do anything.

[Sesquipedalian]

or even Lockheed Martin with Orion (hey, stop laughing!)

Isn't it still the law of the land that Orion is the official backup to the Commercial Crew Program?

[jongoff]

Pushing back a bit on the idea that NASA could cancel the Starliner contract 'for cause' with no termination penalty, CFT-1 returned without a crew at the request of NASA. Those would be good contract terms if NASA had gotten them: 'We can tell you not to fulfill your obligation so then we can cancel with no obligation to pay you.'

While I'm in on the "cancel the contract" side of the argument, I agree with your logic here re: termination for cause over CFT-1 return without a crew. I think you could argue that being as late as they are could potentially be cause. But even if not, I think it might be worth terminating even if they terminate only for convenience. But only if they reinvest any savings into starting one or more other groups on developing commercial crew capabilities. Redundancy is nice, but economically viable competition for post-ISS CLD operations is so much more important. IMO, YMMV, etc.

~Jon

The cause is not the failure of CFT. It's the failure to deliver an operational system by the contractually agreed-upon date of 2017. However, everything we have heard from NASA (and the silence from Boeing) makes it clear that if NASA no longer wanted Starliner, then all they need to do is decline to certify it until after a successful CFT that Boeing needs to pay for. NASA does not actually need to do anything.

I think everyone would be better off if the program was officially terminated -- NASA can't actually free up any remaining money to fund development of alternatives to Starliner or procure additional Dragon flights with the project not formally dead but basically pining for the fjords. I'd be fine with that being Boeing formally backing out, NASA formally terminating, or them mutually agreeing to part ways.

~Jon

[woods170]

Re-assigning Starliner pilot Michael Fincke to Dragons Crew-11 crew may be a subtle hint as to Boeing's future.

No it isn't.

It's NASA making sure that Fincke doesn't have to wait yet another full year before flying. It's similar to how original CFT crew member Nicole Mann, who was assigned to Starliner CFT in August 2018, was re-assigned by NASA in October 2021 to become the vehicle commander of the Crew-5 mission.

Fincke was assigned as the pilot of Starliner PCM-1, the first operational mission of Starliner, in September 2022. Similar to Nicole Mann, he was reassigned to a Crew Dragon mission (Crew-11) to prevent any further delays to him getting into orbit.

Subtle hint as to Boeing's future this reassignment is not therefore.

[DanClemmensen]

Pushing back a bit on the idea that NASA could cancel the Starliner contract 'for cause' with no termination penalty, CFT-1 returned without a crew at the request of NASA. Those would be good contract terms if NASA had gotten them: 'We can tell you not to fulfill your obligation so then we can cancel with no obligation to pay you.'

While I'm in on the "cancel the contract" side of the argument, I agree with your logic here re: termination for cause over CFT-1 return without a crew. I think you could argue that being as late as they are could potentially be cause. But even if not, I think it might be worth terminating even if they terminate only for convenience. But only if they reinvest any savings into starting one or more other groups on developing commercial crew capabilities. Redundancy is nice, but economically viable competition for post-ISS CLD operations is so much more important. IMO, YMMV, etc.

~Jon

The cause is not the failure of CFT. It's the failure to deliver an operational system by the contractually agreed-upon date of 2017. However, everything we have heard from NASA (and the silence from Boeing) makes it clear that if NASA no longer wanted Starliner, then all they need to do is decline to certify it until after a successful CFT that Boeing needs to pay for. NASA does not actually need to do anything.

I think everyone would be better off if the program was officially terminated -- NASA can't actually free up any remaining money to fund development of alternatives to Starliner or procure additional Dragon flights with the project not formally dead but basically pining for the fjords. I'd be fine with that being Boeing formally backing out, NASA formally terminating, or them mutually agreeing to part ways.

~Jon

Yes, some sort of formal and immediate termination is very much preferred, and a mutual agreement would be the least embarrassing end of this fiasco. As of now, NASA continues to encourage and enable Boeing's continuation of Starliner.

My uninformed opinion: NASA should quietly tell Boeing that NASA can no longer support Starliner. Stich can use DOGE as his excuse. Boeing should agree to a joint announcement of a termination using the most positive wording that the PR teams can invent. No further money should change hands in either direction: no further progress payments, but no clawbacks either.

[Comga]

What’s with the “another chance”?
Boeing had grievous problems and errors on OFT(1).
Lots of work and diagnosis and tests
Boeing said they had it all corrected.
They are given “another chance”.
Boeing had new problems on OFT-2(1).
Lots of diagnosis and redesigns and tests
Boeing said NOW they have it all corrected.
Boeing had more problems on OFT-2(2).
Lots more of tests and redesigns.
Boeing says now they REALLY have it all corrected.
CFT, with astronauts on board, has more problems AGAIN, proving that they are unable to “Test like you fly”.
NASA pulls the astronauts off the return flight, engendering a host of reshuffling.
After developing and uploading the autonomous deorbit software, which Boeing had long spoken of as implemented, Starliner lands successfully but not safely.  (After the point of no return, the untested capsule RCS suffers loss of redundancy, dropping to single string, one identical to its quickly failed twin.)

Now, nine months after launch, Boeing is still testing, modifying, and redesigning.
But NASA goes ahead and certifies Statliner.
Sounds like Boeing has been given an undeserved FOURTH chance.

Obviously I am with those wanting Starliner to be abandoned.

[sdsds]

When looking at the on-going viability of a system, cost over-runs should be considered 'sunk' and so too should past under-performance. The assertion implicitly (and sometimes explicitly) made is that past cost over-runs and under-performance predict future cost over-runs and under-performance. That's reasonable, particularly when the approach to resolving cost and performance discrepancies can be characterized as 'whack-a-mole.'

Is there evidence that new faults in Starliner are being discovered faster than known faults are being fixed?

(Disclaimer: I have no horse in the race; just curious.)

[DanClemmensen]

When looking at the on-going viability of a system, cost over-runs should be considered 'sunk' and so too should past under-performance. The assertion implicitly (and sometimes explicitly) made is that past cost over-runs and under-performance predict future cost over-runs and under-performance. That's reasonable, particularly when the approach to resolving cost and performance discrepancies can be characterized as 'whack-a-mole.'

Is there evidence that new faults in Starliner are being discovered faster than known faults are being fixed?

(Disclaimer: I have no horse in the race; just curious.)

Boeing's Starliner program has demonstrated failures in costs, schedules, and reliability. Past program performance is the best predictor of future program performance, unless there is reason to believe Boeing is making effective changes in the way the program operates.

[BN]

In another universe very similar to this one, those astronauts are dead and the answer to this question is more obvious.

[sdsds]

In another universe very similar to this one, those astronauts are dead and the answer to this question is more obvious.

That universe may be very similar, but its Starliner is very different.

[SoftwareDude]

In another universe very similar to this one, those astronauts are dead and the answer to this question is more obvious.

That universe may be very similar, but its Starliner is very different.

Why do you say that? They were lucky to get a third axis of rotation back. It could have gone either way.

[sdsds]

In another universe very similar to this one, those astronauts are dead and the answer to this question is more obvious.

That universe may be very similar, but its Starliner is very different.

Why do you say that? They were lucky to get a third axis of rotation back. It could have gone either way.

They never lost 6DOF control. They did become zero-fault tolerant. Five of the six jets were re-enabled. After the test they reported verbally the sixth did also fire.

At the time they didn't know what would happen if those jets were used when there were indications of overheating — indeed it isn't even clear they knew overheating was the cause — and the correct choice was made. It takes a lot of analysis to get any degree of certainty and there was no time for that.

[armchairfan]

They never lost 6DOF control. They did become zero-fault tolerant. Five of the six jets were re-enabled. After the test they reported verbally the sixth did also fire.

At the time they didn't know what would happen if those jets were used when there were indications of overheating — indeed it isn't even clear they knew overheating was the cause — and the correct choice was made. It takes a lot of analysis to get any degree of certainty and there was no time for that.

https://arstechnica.com/space/2025/04/the-harrowing-story-of-what-flying-starliner-was-like-when-its-thrusters-failed/

And this is the part I'm sure you haven't heard. We lost the fourth thruster. Now we've lost 6DOF control. We can't maneuver forward. I still have control, supposedly, on all the other axes.

Emphasis mine.

[BN]

In another universe very similar to this one, those astronauts are dead and the answer to this question is more obvious.

That universe may be very similar, but its Starliner is very different.

Why do you say that? They were lucky to get a third axis of rotation back. It could have gone either way.

They never lost 6DOF control. They did become zero-fault tolerant. Five of the six jets were re-enabled. After the test they reported verbally the sixth did also fire.

At the time they didn't know what would happen if those jets were used when there were indications of overheating — indeed it isn't even clear they knew overheating was the cause — and the correct choice was made. It takes a lot of analysis to get any degree of certainty and there was no time for that.

zero-fault tolerance means the difference between those universes was one fault.

and as you said, there wasn't time for certainty, so different choices could have been made.

[sdsds]

Emphasis mine.

I'll repeat my request. Of the thrusters lost at one point or another, the only combination leading to loss of 6DOF control would have been b1a3, b2a3, s2a2 and s1a1. When were those ever lost at the some time? Listening to the mission control audio I assert they never were, despite how Wilmore recalls it.

If I transcribed Stich's comments correctly at one time or another Starliner disabled each of: starboard 2a2, bottom 1a3, bottom 2a3, starboard 1a1 and top 2a2, all aft-facing. Please name the thrusters which — all at the same time — were disabled, which in combination led to loss of 6DOF control.

[meekGee]

I think it's more nuanced than that.

We just say 6-DoF, but actually each DoF requires two directions, and thrusters are push-only. Trivially, you need 12 symmetrical thrusters.

I can imagine a 9 thruster combination that gives 6-DoF control, but they need to be prepositioned correctly.  Not any 9 will do. Theory says a special arrangement of 8 is also possible. But they'll be coupled of course.

But, it's not clear to me how much coupled control Wilmore could handle.  The F-18 comment makes it clear that it wasn't too much. Which is reasonable, especially when it's proximity ops.

Also, aren't some thrusters additionally inhibited (by necessity or preference) when very close to the station?

[BN]

I think it's more nuanced than that.

We just say 6-DoF, but actually each DoF requires two directions, and thrusters are push-only. Trivially, you need 12 symmetrical thrusters.

I can imagine a 9 thruster combination that gives 6-DoF control, but they need to be prepositioned correctly.  Not any 9 will do. Theory says a special arrangement of 8 is also possible. But they'll be coupled of course.

But, it's not clear to me how much coupled control Wilmore could handle.  The F-18 comment makes it clear that it wasn't too much. Which is reasonable, especially when it's proximity ops.

Also, aren't some thrusters additionally inhibited (by necessity or preference) when very close to the station?

I think only OMAC is restricted, which leaves the 28 RCS.

[DanClemmensen]

I can imagine a 9 thruster combination that gives 6-DoF control, but they need to be prepositioned correctly.  Not any 9 will do. Theory says a special arrangement of 8 is also possible. But they'll be coupled of course.

But, it's not clear to me how much coupled control Wilmore could handle.  The F-18 comment makes it clear that it wasn't too much. Which is reasonable, especially when it's proximity ops.

I don't think the pilot uses separate control of the individual thrusters, so this coupling is done in software. Wilmore commented that the control became less precise as thrusters dropped out, and this would be consistent with software-mediated coupling.

[jarmumd]

I can imagine a 9 thruster combination that gives 6-DoF control, but they need to be prepositioned correctly.  Not any 9 will do. Theory says a special arrangement of 8 is also possible. But they'll be coupled of course.

But, it's not clear to me how much coupled control Wilmore could handle.  The F-18 comment makes it clear that it wasn't too much. Which is reasonable, especially when it's proximity ops.

I don't think the pilot uses separate control of the individual thrusters, so this coupling is done in software. Wilmore commented that the control became less precise as thrusters dropped out, and this would be consistent with software-mediated coupling.

Software control is how others do it.  Even in manual piloting, the pilot is saying "go that way", and the computer figures out the impulse needed from the various thrusters to eventually go that way.

[meekGee]

I can imagine a 9 thruster combination that gives 6-DoF control, but they need to be prepositioned correctly.  Not any 9 will do. Theory says a special arrangement of 8 is also possible. But they'll be coupled of course.

But, it's not clear to me how much coupled control Wilmore could handle.  The F-18 comment makes it clear that it wasn't too much. Which is reasonable, especially when it's proximity ops.

I don't think the pilot uses separate control of the individual thrusters, so this coupling is done in software. Wilmore commented that the control became less precise as thrusters dropped out, and this would be consistent with software-mediated coupling.

Software control is how others do it.  Even in manual piloting, the pilot is saying "go that way", and the computer figures out the impulse needed from the various thrusters to eventually go that way.

So when Wilmore says "now we can't go forward", how do you read this?