Re: EU Competition?
Well, I *worked* on the Galileo program for a while. I managed the development of one of its network components, responsible for a team of about twenty engineers. So I probably know a lot more technically about this than you do.
Galileo is famously and horrifically *unreliable*. The average SISRE is Ok, but that isn’t very useful if you are one of the unlucky 0.1% driving over a cliff. Here’s the most recent official quarterly performance report
https://www.gsc-europa.eu/sites/default/files/sites/all/files/Galileo-OS-Quarterly-Performance_Report-Q3-2021.pdf
In September, average SISRE 2cm, 95% better than 35cm (hmmmm….’! However, the 99.9% point was….539 *metres*. You read that right. And 99.9% isn’t even that high as a metric. That’s *nearly an hour a month* that’s its off by half a Kilometer. That’s insane. But this is actually a huge *improvement*. Since “initial launch”, 2021 was the first year they actually managed to do better than 99% of better than 30 meters! Think about that for a second. Never mind landing a plane, if you were a lorry driver using Galileo alone in 2020, it would have placed you a city block away from true position up to *7 hours per month*.
Galileo “went live” in 2016, but it does *not* have Full Operational Capability. The main problem it has, and by no means the only one, is that it doesn’t have enough satellites. This makes its accuracy *usually* ok, but *sometimes* very bad. So when it tries to provide a guaranteed service, eg. relied upon to land an aeroplane, it fails badly. It’s *supposed* to have 30 satellites in each of 3 orbital planes. It *needs* 8 per orbital plane, and because of the way orbits work it’s no good having 7 in one plane and 10 in the other two. And it needs a spare in each orbital plane, so that if any one satellite fails, there are still sufficient to provide the stated accuracy. What it actually *has*, right now, is 22 usable satellites, and 2 under commissioning. They only have production capability to launch 4 per year. They won’t have the required 9 satellites in each of the 3 orbital planes until end 2024. Unfortunately, these satellites only have 12 year nominal life. The oldest satellites in the current constellation were launched in 2011. Hence, they are *expected* to start dying at the rate of 4 per year in about 2023/2024. Unless the satellites exceed their design life, Galileo will *never* have 9 operational satellites per orbital plane. Perhaps they will? Well at the moment, signs aren’t good. The onboard atomic clocks of the first ten launched turned out to have a design problem, causing them to degrade early. They have onboard redundancy, but frankly it’s a real race at the moment.
Galileo also suffers because the Ops Centre needs to uplink corrected orbits to every satellite every day, without fail. But there simply aren’t enough uplink stations to do that every day. Brexit hasn’t helped it, because they lost two uplink stations. Until quite recently, at least a couple of days a week, several of the satellites had quite badly degraded timing, which affects accuracy. In October 2020 they added more uplink capacity, but it still fails quite often. Notice how the public haven’t been told anything about this.
https://berthub.eu/articles/posts/galileos-ground-segment-problem/
There are also some really bad design failures in the way it handles the distribution of time. Basically, one network element depends on another element which depends on…..etc, with circular definition. This has the effect that once the ground network has the slightest failure, it can take up to *several days* for the timing state to converge back to nominal. The USA GPS people actually told Galileo program about this quite early in the system definition. When they designed theirs in the seventies they made the same mistake and wanted Galileo to learn from their experience. But Galileo project was too arrogant and ignored their advice. This was the root cause of the six-day outage in July 2019. And it simply can’t be fixed, because it is implicit in the network architecture. It absolutely will happen again identically with a six-day outage next time a single timing component fails.
And finally….even intrinsically on average SISRE, Galileo is no more than accurate than GPS. This was an official myth put out in the early days by people who didn’t understand it, and just repeated on paper what they had heard, or would like to have heard. The truth is, the only physics difference comes from satellite power. More power, more accuracy, that’s all. In 2000, the new Galileo system was spec’d to have a higher power than the then-current GPS Block II satellites. The GPS Block III satellites launched from 2018 onwards have much higher power, and 3x better accuracy. And now better accuracy than Galileo. Probably next block Galileo from 2030 onwards might have higher power still, but there really isn’t any intrinsic difference in performance between the two systems.
Biting the hand that feeds IT
