Starliner lands safely after failed orbital insertion
Capitalism in space:Boeing’s Starliner capsule successfully landed today in New Mexico, returning to Earth prematurely because of its failure to reach its proper orbit after launch two days ago.
The article quotes extensively from both NASA and Boeing officials touting the many successful achievements of this flight, while trying to minimize the failure that prevented the capsule from docking with ISS properly. And that failure?
The mission elapsed timer issue that cut short Starliner’s planned eight-day mission started before the spacecraft lifted off Friday from Cape Canaveral aboard a United Launch Alliance Atlas 5 rocket, according to Chilton. “Our spacecraft needs to reach down into the Atlas 5 and figure out what time it is, where the Atlas 5 is in its mission profile, and then we set the clock based on that,” Chilton said in a press conference Saturday. “Somehow we reached in there and grabbed the wrong (number). This doesn’t look like an Atlas problem. This looks like we reached in and grabbed the wrong coefficient.”
“As a result of starting the clock at the wrong time, the spacecraft upon reaching space, she thought she was later in the mission, and, being autonomous, started to behave that way,” Chilton said. “And so it wasn’t in the orbit we expected without the burn and it wasn’t in the attitude expected and was, in fact, adjusting that attitude.”
I read this and find myself appalled. While I agree that overall the mission proved the capsule capable of launching humans to ISS (which is why NASA is considering making the next Starliner mission manned despite this failure), this failure suggests a worrisome lack of quality control at Boeing. I can’t even imagine how the Starliner software could be mis-configured to “grab the wrong number.” This explanation makes no sense, and suggests they are spinning the failure to avoid telling us what they really did wrong.
Either way, I suspect that NASA will approve a manned launch for Starliner’s next orbital flight, but will do so only after dwelling on the problem for at least six months.
On Christmas Eve 1968 three Americans became the first humans to visit another world. What they did to celebrate was unexpected and profound, and will be remembered throughout all human history. Genesis: the Story of Apollo 8, Robert Zimmerman's classic history of humanity's first journey to another world, tells that story, and it is now available as both an ebook and an audiobook, both with a foreword by Valerie Anders and a new introduction by Robert Zimmerman.
The ebook is available everywhere for $5.99 (before discount) at amazon, or direct from my ebook publisher, ebookit. If you buy it from ebookit you don't support the big tech companies and the author gets a bigger cut much sooner.
The audiobook is also available at all these vendors, and is also free with a 30-day trial membership to Audible.
"Not simply about one mission, [Genesis] is also the history of America's quest for the moon... Zimmerman has done a masterful job of tying disparate events together into a solid account of one of America's greatest human triumphs."--San Antonio Express-News
Capitalism in space:Boeing’s Starliner capsule successfully landed today in New Mexico, returning to Earth prematurely because of its failure to reach its proper orbit after launch two days ago.
The article quotes extensively from both NASA and Boeing officials touting the many successful achievements of this flight, while trying to minimize the failure that prevented the capsule from docking with ISS properly. And that failure?
The mission elapsed timer issue that cut short Starliner’s planned eight-day mission started before the spacecraft lifted off Friday from Cape Canaveral aboard a United Launch Alliance Atlas 5 rocket, according to Chilton. “Our spacecraft needs to reach down into the Atlas 5 and figure out what time it is, where the Atlas 5 is in its mission profile, and then we set the clock based on that,” Chilton said in a press conference Saturday. “Somehow we reached in there and grabbed the wrong (number). This doesn’t look like an Atlas problem. This looks like we reached in and grabbed the wrong coefficient.”
“As a result of starting the clock at the wrong time, the spacecraft upon reaching space, she thought she was later in the mission, and, being autonomous, started to behave that way,” Chilton said. “And so it wasn’t in the orbit we expected without the burn and it wasn’t in the attitude expected and was, in fact, adjusting that attitude.”
I read this and find myself appalled. While I agree that overall the mission proved the capsule capable of launching humans to ISS (which is why NASA is considering making the next Starliner mission manned despite this failure), this failure suggests a worrisome lack of quality control at Boeing. I can’t even imagine how the Starliner software could be mis-configured to “grab the wrong number.” This explanation makes no sense, and suggests they are spinning the failure to avoid telling us what they really did wrong.
Either way, I suspect that NASA will approve a manned launch for Starliner’s next orbital flight, but will do so only after dwelling on the problem for at least six months.
On Christmas Eve 1968 three Americans became the first humans to visit another world. What they did to celebrate was unexpected and profound, and will be remembered throughout all human history. Genesis: the Story of Apollo 8, Robert Zimmerman's classic history of humanity's first journey to another world, tells that story, and it is now available as both an ebook and an audiobook, both with a foreword by Valerie Anders and a new introduction by Robert Zimmerman.
The ebook is available everywhere for $5.99 (before discount) at amazon, or direct from my ebook publisher, ebookit. If you buy it from ebookit you don't support the big tech companies and the author gets a bigger cut much sooner.
The audiobook is also available at all these vendors, and is also free with a 30-day trial membership to Audible.
"Not simply about one mission, [Genesis] is also the history of America's quest for the moon... Zimmerman has done a masterful job of tying disparate events together into a solid account of one of America's greatest human triumphs."--San Antonio Express-News
Another thing that I find curious about all the discussion. SpaceX launches Dragon capsules to ISS with an instantaneous launch window – no holds, no delays. Any issue, and they scrub for the day. I know that ISS has a high inclination orbit because of Russia’s launch sites being farther north, so I can understand that it’s a difficult launch from Florida.
And yet for Starliner, they keep insisting that if there had been astronauts on board, they could have “punched in the numbers” and gotten back on track for ISS. Really? Seat-of-the-pants flying? And they’re supposedly trained for this?
I’m not an orbital mechanic, but this doesn’t pass the sniff test, any more than the clock problem.
According from what I’ve read, the clock was off by 11 hours. Atlas V powers up 11 hours before launch. Why is this interesting? Its because ULA/Boeing tried something new: the booster & capsule ‘talk’ to each other and share data. The computer itself is fine but there might be a bug in….what for it….Boeing’s software.
Diane
They are trying to hint around that the astronauts would have just shut off the engines before all the fuel was wasted and waited for ground control to figure out the problem.
They don’t expect ‘seat of the pants flying’. NASA would just upload new software.
But I wonder if Boeing trusts their docking software over their flight software.
Considering computer communication speeds now available why are the “variables” like the clock not double checked by ground control seconds before launch?
Diane:
The ISS orbits the Earth in an orbit inclined
51.6 degrees to the equator. This orbit is
fixed in space as the earth rotates beneath it.
The instantaneous launch window is when the
vehicle sitting on the rotating earth rotates
into the plane of ISS’s obit. On this launch
the Atlas put the Starliner into the ISS’s o
orbital plane but with the perigee in Earth’s
upper atmosphere (as designed). Starliner was
then to fire its own OMACS engines to raise the
perigee out of the atmosphere. Had there been
test pilots aboard, they could have 1) recognized
the attitude control system was behaving erratically
2) shut it down, 3) troubleshot the problem
and tested manual control for a period of time, and
possibly 4) manually raised the perigee out
of the atmosphere. The resulting orbit would
have been in ISS’s orbital plane but lower
and thus faster, allowing Starliner to catch
up to the ISS.
Robert:
Failed orbital insertion? Really? The orbital
insertion by the Atlas was near perfect. The
problem was with Starliner’s orbital maneuvering.
And there is no need for conspiracy theories.
There is likely a logical explanation for this
screw-up. The zero value for the Mission
Elapsed Timer (MET) appears to coincide with
Atlas vehicle power-up, so the problem could
be as simple as the wrong parameter being placed
on the logical Interface Control Document (ICD).
However, I do suspect you’re right about one thing.
I supsect there will be a six-month delay but
not for NASA pondering. The delay will be while
Boeing conducts a root cause corrective action
to determine why the MET had the wrong value,
what process or procedure failures allowed that
to happen, and why it wasn’t caught during
software testing, hardware-in-the-loop testing,
or the Integrated Systems Test with the Atlas
vehicle before launch. Then, depending of the
findings of the RCCA, Boeing will have to
determine whether the root cause caused any other
issues, including those that may not have been
seen on this flight. It’s a time-consuming
process to do right, but if it’s done right, I
suspect NASA will allow the next flight to be manned.
On a first unmanned launch of any new system, an unexpected glitch is acceptable…as long as it is corrected.
That’s what test launches are for..simulations are never exactly like the actual event. All in all, the hardware seemed to work well.
On a first unmanned launch of any new system, an unexpected glitch is acceptable…as long as it is corrected.
That’s what test launches are for..simulations are never exactly like the actual event. All in all, the hardware seemed to work well.
I noticed that during the landing webcast, Boeing’s host claimed that Starliner’s landing was “the first time an American made human rated capsule has landed on land.” Didn’t Blue Origin do this already? Several times? At least once with Mannequin Skywalker on board?
Diane Wilson,
Apollo astronauts also “punched the numbers” on multiple occasions during each flight. For them is was standard operating procedure, since their computers were too primitive to contain and do as much as modern computers can. Houston radioed up the numbers, by voice, and the astronauts punched the numbers into their DSKY keyboard. Rather than upload new software, as pzatchok wrote, astronauts likely would have changed either the timing or some form of navigational instruction.
mkent,
You are right about the timing of the orbital plane as the Earth rotates, but some timing delays can be corrected during launch by making adjustments during launch. A dog-leg maneuver can sometimes be employed to reach the correct orbital plane during launch, but only if the plane is slightly off from the normal launch. Upper stages employ a closed-loop guidance that helps reach the proper orbit.
SpaceX’s narrow window on launches tends to be more for fuel density considerations. If the rocket sits on the pad too long, the propellants warm up and become less dense, meaning that there is less propellant in the tanks at launch. It is also why SpaceX would prefer fueling after the astronaut are settled in the Dragon spacecraft.
Also, Robert did not say “failed orbital insertion” during launch but that Starliner had a “failure to reach its proper orbit after launch.” It was Starliner’s error after launch was over.
Robert,
You wrote: “this failure suggests a worrisome lack of quality control at Boeing. I can’t even imagine how the Starliner software could be mis-configured to ‘grab the wrong number.’ This explanation makes no sense, and suggests they are spinning the failure to avoid telling us what they really did wrong.”
These kinds of errors are shockingly common. Well, they are rare, but they occur more than we find acceptable. Lockheed Martin mismatched metric vs. “imperial units” (U.S. standard units), and lost a Mars mission. The Russians, a couple of years ago, gave one of their rockets the wrong launch pad location, and the upper stage (in closed loop mode) tried to navigate back to where it thought it should be, slowing down rather than speeding up into orbit. Ariane 5’s 97th mission veered off course, but its payloads managed to get into the correct orbits anyway.
Each time one of these errors occurs, we are appalled that they are still possible. After all, why haven’t we managed to find these types of errors in advance? How does a computer look into the wrong data cell of another computer, anyway? Isn’t there sufficient communication between groups, and isn’t there sufficient double checking to make this never happen? Or at least to happen rarely?
Well, it does happen, and it does happen rarely. The real question is: what is being done wrong that allows these to happen at all? Why haven’t previous corrective actions not prevented similar mistakes from being made? (Make that: two real questions.) Aren’t the different companies and space programs learning from each other’s mistakes? (Three questions — but maybe we should stop correcting my first error.)
There have been thousands of launches and thousands more spacecraft missions. Surely we should know how to make things work right by now.
Bill Whittle once pointed out that even in air travel, it took decades and millions of flights before we got to the point where the U.S. went a decade without a passenger fatality in a major airliner.
https://www.youtube.com/watch?v=GXbdJ3kyVyU (7 minutes)
“The reason that we have cheap, affordable, and safe air transportation today and no space transportation whatsoever is simply that we were serious about air travel — serious enough to pay the price in blood and money — and we’re not serious about space.” … “That’s the deal. That’s what it costs.” … “Part of the deal, you see, is that you pay in blood for progress. If there’s no progress, what’s the point?”
Hopefully, we have learned enough from the air travel industry that we won’t have to lose as many people in the space travel industry before we learn how to do it right. Hopefully, transitioning from government-run manned space programs to commercial space programs is progress worth the price of the deal.
Couldn’t they have used a simple accelerometer to tell the capsule when the launch started and the capsule could then start counting from there (T + whatever)? It already knows the mission profile, so when the clock says to do a certain maneuver, and the profile says that maneuver should happen at T + n, but the time-since-launch counter says it is not yet T + n, something is obviously wrong and the capsule should contact mission control for instructions.
It seems to me that it greatly tempts fate for Starliner to rely on mission elapsed time for thruster firing decisions, especially when that data source is not under Boeing’s control. Just as with Boeing management’s decision to move the 737 Max’s engine placement in a manner that makes the aircraft less stable, this is a design decision that replaces fundamental engineering soundness by requiring absolute perfection of complex software. I’d love to know the details of how these decisions came about but as a software developer for over thirty years I can imagine the scenarios very well. There is too much confidence in the ability of software to compensate for fundamentally bad design decisions made upstream. I hate to criticize with the benefit of hindsight but there is a pattern here and it is not unique to Boeing.
I really don’t like the entire concept of basing spacecraft operations, and indeed, entire modes of flight around a CLOCK of all things. Sure, time should always be a factor in navigation; but not in basic operations.
Using a set of variables–even just simple flags–would have prevented this whole mess. Don’t assume that everything in a launch will go exactly according to the schedule every time.
Andi asked: “Couldn’t they have used a simple accelerometer to tell the capsule when the launch started and the capsule could then start counting from there (T + whatever)?”
The time is unlikely to be the only information that the spacecraft got from the launch vehicle. It probably also got the release point, the orbit that the spacecraft was in during release, and other relevant information that would be needed to navigate and properly continue its mission.
Why ask the launch vehicle? The answer to that was included in one of the articles that Robert linked in another post.
https://www.nasaspaceflight.com/2019/12/starliner-mission-shortening-failure-successful-launch/
“When mission controllers realized the issue, they sent manual commands to Starliner to perform an Orbit Insertion Burn in a backup window that came roughly eight minutes after the planned maneuver. However, a known and brief gap in NASA satellite communications caused a further delay.”
Gaps in communications means that the ground is not a completely reliable source of information. Until release, the launch vehicle is a good source that is available even when the ground information is not.
Michael Mangold wrote: “There is too much confidence in the ability of software to compensate for fundamentally bad design decisions made upstream.”
I agree. In addition, software can be adaptable or updated after release, but it is difficult to write it to account for all possible scenarios. Humans are built to adapt to new situations as they arise, and this is why Boeing said that if a crew had been aboard they would have taken corrective actions soon after noticing the anomaly. This is one of the reasons why I believe that self-driving cars are not in our near future. It is nice to dream about, but just as with the flying car, it is difficult to safely implement.
To my uninformed eye, it would seem that the difficult part of the mission would be navigating and docking with ISS. If they give Boeing a pass this will not be tested.
Steve C wrote: “To my uninformed eye, it would seem that the difficult part of the mission would be navigating and docking with ISS. If they give Boeing a pass this will not be tested.”
There have been plenty of dockings at the ISS and other space stations. That should be a no brainer, by now.
Oh, wait. Boeing is one of those without any experience docking anything to anything else, as Boeing’s ISS modules were carried by the Space Shuttle and assembled, not autonomously docked. Plus, there is far, far more experience separating spacecraft from launch vehicles, even by Boeing’s own commercial communication satellite division. So, how could Boeing have gotten that wrong?
On the other hand, if the software goes awry on an unmanned spacecraft, would an errant spacecraft be safe around the space station? This is roughly what happened in the 1990s, when a docking test went badly wrong, and a Progress resupply cargo spacecraft struck and depressurized the Spektr module of the Mir space station. It was under local cosmonaut control, not automatic control, but the video feedback was so poor that the cosmonaut could not make out the Mir, much less the docking port.
Maybe, if we cannot trust the software, we want onboard human control during docking, or the ability for an onboard human to take control if things go wrong again.
The main part of this Starliner test was to make sure that it is safe for humans. It can launch safely and it can return safely. That is the important part. All that is really missing is whether it can perform its mission, and as long as the astronauts can take control of the Reaction Control System then it should do that safely, too.