Cars also don’t move in easily predicted patterns for decades without turning from their predetermined path.

They do. We call those paths roads.

And cars never take a sudden unpredictable turn off the road?

They do. Now make them move ten times faster than an airplane. What happens?

Mind you I'm not the one who came up with the TERRIBLE analogy with cars. Millions of people a year die in car accidents, and this was somehow the example that's supposed to convince us that if "cars are safe" then satellites travelling 100x faster than a car over our heads (and 10x faster than an airplane) are also safe. The whole construct betrays extremely sketchy thought process.

They sort of do though.

Cars, for the most part, follow the road and the direction they're already going. They (usually) signal if they're about to make changes. And to be safe, we've built an arrangement that they only travel on these specific paths at specific speeds, modifying their speed to match the other cars they might hit.

The challenge with satellites is that steering is hard and very limited. And they don't have brakes they can slam to come to a stop if they need to suddenly avoid a collision.

It is well known that potential points of conflict are a huge issue. That's why a round about is safer then intersection.

> And to be safe

Tell that to the 40k people killed and the millions hurt by cars in the US alone.

Cars constantly turn and intersect routes with other cars without any planning, even if perfectly handled and most of time they are not.

> The challenge with satellites is that steering is hard and very limited. And they don't have brakes they can slam to come to a stop if they need to suddenly avoid a collision.

They don't need to break, they just need to know the position of speed of other objects and correct position ahead of time.

The only reason why this is even a program at all is because only the US has even minimally invested in monitoring. If each nation just did the same our accuracy would be far, far higher and things would be way better.

So, as of now, we have minimally invested in solving this problem and it isn't a big issue. Slightly more investment (that is justified if we 10-100-1000x the amount of sats) is required eventually.

But compared to what it takes to make cars not crash into each other its nothing.

(Unless its like US infrastructure design that seems to want to make cars crash into things, but I hope in Space we can avoid that)

Cool thing about satellites is if you're on the same path, you must also be going the same speed.

Collision avoidance in LEO is quite elegant. When two satellites are predicted to pass uncomfortably close to one another, one satellite will fire its station keeping thrusters to raise its orbit above the other. No brakes or swerving needed, just a slight push well ahead of time. Starlink sats do this over 1000 times per week.

Even better, energy used for avoidance reduces energy needed for station keeping on a 1:1 basis. Avoidance maneuvers are effectively free.

What powers those thrusters? Never really thought about it.

They are powered by solar electricity that is saved in batteries. But to actually move you need something to throw out. That can be many things, anything from normal water, metal or noble gases.

Starlink initially used Krypton for its thrusters. But they have now switched to Argon since its incredibly cheap and available. Xenon for a long time was the standard, but its to expensive and rare.

Some sats also have chemical power, often a Hypergolic propellant, like dinitrogen tetroxide plus hydrazine. Specially for larger sats this is the case.

- https://en.wikipedia.org/wiki/Ion_thruster

-

Cars move much more uncontrollably / unpredictably than sats.