I'm not sure that SpaceX's excuse that it wasn't deliberate really helps their image here. Deliberately ignoring it, for whatever reason, could be fixed with a firm "you really can't do that", backed by lawsuits or complaints to the US government (FCC? Airforce?) etc.
On the other hand, a culture of oversight that fails to adequately test its emergency response procedures seems a more systemic problem.
The article mentions that at the time they said that, the probability of collision wasn't that high. So, it looks totally OK to say they didn't have plans to move it, then, since it means they didn't believe there really was any risk of collision if they both stayed in their current paths. At that time, ESA's query was probably to ensure they didn't have any plans for maneuvers that would raise the probability.
> “They said at that point in time they had no plans,” says Merz. SpaceX confirmed that it exchanged the initial email with ESA. At that time, the probability of collision was about 1 in 50,000, according to SpaceX, which is too low to require any preventive action.
SpaceX might have even interpreted it as ESA asking because they might have their own plans for maneuvers and they just wanted to make sure that SpaceX's movements stayed predictable.
It doesn't need to be that SpaceX was being negligent here.
At the time they said that, yes, probability of impact was low. But then the probability got much, much, much higher. 10x the probability at which maneuvers are typically made. And we can't chalk it up to what or how SpaceX might have interpreted it. We know interpretation was not the problem. They freely admit that they never saw further messages because of a bug in their process, so we don't need to imagine their interpretation. The simply never knew the risk was there. Even if we were going to that extent though, clearly such an interpretation doesn't comport itself to how these things are resolved within the community of satellite operators. Which, if SpaceX made a misinterpretation, also shows a negligent disregard for how emergency resolution functions here.
I hope their ubiquitous constellation of broadband-providing satellites works well. I couldn't care less about the negligible impact it might have on visual observations of the sky-- the society enriching effects of cheap ubiquitous networks outweigh it. But they have to get basic things like safety protocols right, not iterating without such things from a minimum viable product.
> They freely admit that they never saw further messaged because of a bug in their process, so we don't need to imagine their interpretation.
No, what I said about interpretation was about the first query when it was a 1 in 50,000 chance. For the message they missed, we don't need to imagine their interpretation because they can't interpret a message they never received.
> Even if we were going to that extent though, clearly such an interpretation doesn't comport itself to how these things are resolved within the community of satellite operators.
At the level of the community, I don't see a problem here. ESA tried to contact SpaceX to agree on a course of action; SpaceX couldn't be contacted because of a bug on their end, so ESA acted on its own to resolve the situation from its end.
I can see a problem if they plan to be a significant portion of all space traffic and they can never be relied on to be contactable in case of problems arising from their satellites. However, this happening once doesn't seem like negligence.
Their response to the first query was fine. We agree on that.
My criticism is with them having a system that wasn't tested adequately enough to realize they would be blind to further messages. This forced ESA to act on its own when standard protocol, the responsible way to behave, is to work together. The ESA's ability to act was a luxury not all satellites have. SpaceX's inadequate testing could have had catastrophic consequences for both parties. You seem to be using the fact that it all worked this time as evidence that there was no real problem here. But the fact that one responsible party avoided disaster doesn't excuse the irresponsible party's problem with inadequate testing. It would be a bit like having a car drive on the wrong side of the road, nearly hit another car, but that other car moved out of the way, so really everything is okay. That doesn't quite parse for me. I don't see how someone can come away from this without saying that SpaceX needs to fix something, because SpaceX themselves admit to that.
If it happens again, that's one thing, but happening once is bound to happen as SpaceX doesn't have experience managing this type of constellation. Not to mention it's being done by SpaceX Seattle rather than SpaceX Hawthorne, which has much less experience in this type of thing. Better it happen with a near miss (unlikely the collision would have actually occurred even if ESA had done nothing) than with an actual collision.
Just a few months ago SpaceX fanboys were telling us those satellites were no problem, because "space is so large". The problem is, they’re operated by overworked, delusional people.
People are going to die, because hype is more fun than security.
Probably because the created debris would be practically impossible to clean up later, and if we ever have consumer space travel, that debris might pose a significant risk.
> SpaceX CEO Elon Musk claims that each Starlink satellite is programmed with its own autonomous space debris tracking system, which it will use to move out of the way of potential debris if necessary.
Uh huh. Yup, and teslas already fully self drive. Do not believe Musk's claim that anything he works on does X until you see it do X with your own two eyes!
If you have one mission critical system, all your systems are mission critical.
It's usually the link you don't expect or don't know about that screws you over. Like an email server, DNS, networking. At any level, a complicated house of cards can easily come down.
Thankfully, in rocket engines, simplicity is a design goal.
Are there 'give way to the right' or 'both alter heading to starboard' type rules for orbit? You wouldn't want both satellites to alter their orbits in the same direction.
No. The only thing codified in law is that the country of launch bears the liability for the satellite. If there were a collision, the liability to ESA would be due from the US government given that the satellites launched from US soil.
One point that's missing here, is this occurred on Labor Day weekend, a Federal US Holiday. Everyone involved would have been off of work and they were relying on the pager system. If a bug occurred in that then there would have been no one to reach.
19 comments
[ 251 ms ] story [ 1203 ms ] threadOn the other hand, a culture of oversight that fails to adequately test its emergency response procedures seems a more systemic problem.
> “They said at that point in time they had no plans,” says Merz. SpaceX confirmed that it exchanged the initial email with ESA. At that time, the probability of collision was about 1 in 50,000, according to SpaceX, which is too low to require any preventive action.
SpaceX might have even interpreted it as ESA asking because they might have their own plans for maneuvers and they just wanted to make sure that SpaceX's movements stayed predictable.
It doesn't need to be that SpaceX was being negligent here.
I hope their ubiquitous constellation of broadband-providing satellites works well. I couldn't care less about the negligible impact it might have on visual observations of the sky-- the society enriching effects of cheap ubiquitous networks outweigh it. But they have to get basic things like safety protocols right, not iterating without such things from a minimum viable product.
No, what I said about interpretation was about the first query when it was a 1 in 50,000 chance. For the message they missed, we don't need to imagine their interpretation because they can't interpret a message they never received.
> Even if we were going to that extent though, clearly such an interpretation doesn't comport itself to how these things are resolved within the community of satellite operators.
At the level of the community, I don't see a problem here. ESA tried to contact SpaceX to agree on a course of action; SpaceX couldn't be contacted because of a bug on their end, so ESA acted on its own to resolve the situation from its end.
I can see a problem if they plan to be a significant portion of all space traffic and they can never be relied on to be contactable in case of problems arising from their satellites. However, this happening once doesn't seem like negligence.
My criticism is with them having a system that wasn't tested adequately enough to realize they would be blind to further messages. This forced ESA to act on its own when standard protocol, the responsible way to behave, is to work together. The ESA's ability to act was a luxury not all satellites have. SpaceX's inadequate testing could have had catastrophic consequences for both parties. You seem to be using the fact that it all worked this time as evidence that there was no real problem here. But the fact that one responsible party avoided disaster doesn't excuse the irresponsible party's problem with inadequate testing. It would be a bit like having a car drive on the wrong side of the road, nearly hit another car, but that other car moved out of the way, so really everything is okay. That doesn't quite parse for me. I don't see how someone can come away from this without saying that SpaceX needs to fix something, because SpaceX themselves admit to that.
My theory: someone must've turned off their pagerduty and didn't hear the alarm.
People are going to die, because hype is more fun than security.
https://www.youtube.com/watch?v=heESAW2addo
Uh huh. Yup, and teslas already fully self drive. Do not believe Musk's claim that anything he works on does X until you see it do X with your own two eyes!
It's usually the link you don't expect or don't know about that screws you over. Like an email server, DNS, networking. At any level, a complicated house of cards can easily come down.
Thankfully, in rocket engines, simplicity is a design goal.