I don’t understand how satellites work. But Space-X is working on reducing the cost to put things into orbit. One of their goals with Starlink is to put something like 11,000 satellites into orbit to create everpresent 1Gbps internet speeds.
Here is my issue. Yeah, 1Gbps is extremely fast now. But 56.6 kbps was extremely fast 30 years ago. Isn’t there a risk 1Gbps will be laughably slow by 2050? If we need 1Tbps speeds in 2060, can the speeds on the satellites be upgraded? Wouldn’t that require changing the hardware, which wouldn’t be easy?
Can the satellites be changed to upgrade internet speeds? What if some new technology comes about that makes the satellites worthless. Space-X’s plans will add more satellites than have been added in the last 60 years. That increases the risks of collisions which causes parts to fly off, creating more collisions.
So can someone who knows more than me explain why I’m wrong? I’m worried that putting 11,000 satellites into orbit will increase the risk of satellite collisions, and whatever technology Space-X puts up there will probably be obsolete in 20 years anyway.
This seems like a recipe for putting a bunch of satellites into space that’ll be obsolete within a few decades, and will increase the risk that other satellites experience collisions.
Part of the appeal of SpaceX’s proposal is that they’ll be willing to de-orbit malfunctioning satellites because they’re so small and cheap (comparatively, at least for satellites).
Also, worth noting, SpaceX isn’t the only proposal along these lines. There are several other companies planning on competing with LEO-satellite-based Internet access.
I haven’t really looked into the details of the system, but I would strongly suspect that the satellites would have a mean mission duration of around about 2 or 3 years, after which their orbit would decay and burn up on re-entry to Earth’s atmosphere.
There’s no risk of these smallsats being obsolete in 2050 or whatever, because the constellation will have been reconstituted a dozen times.
But you are right, 11,000 smallsats is a metric fuckload of satellites. IIRC, the current catalog of satellites orbiting in space is somewhere around 10,000. Add in space junk, and I believe you’d more than double that number.
11,000 is a lot, but it’s not crazy. There are 500,000 tracked objects in orbit today (most of them are small, but a 1-inch hunk of metal can take out another satellite as easily as a 2-foot piece).
The constellation isn’t for urban areas; it’s for areas where it’s impractical or uneconomic to run fiber optics. Ships and airplanes are obvious targets, as are rural areas.
As others said, the satellites simply won’t last that long in orbit. They’ll be mass produced, which will make them *more *reliable than the current bespoke method of producing satellites.
They will deorbit themselves when obsolete with a dual-redundant system. And if that fails (which it probably will for a small fraction), they are in a low enough orbit that they’ll come down on their own after a couple of decades.
They are only intended to last 5-7 years. The whole idea is that if launches are cheap enough, and mass production makes the satellites themselves cheap, then they can just continually refresh the constellation so that it never becomes obsolete.
This isn’t some far-out idea: Planet Labs is doing pretty much this with their Earth imaging constellation, albeit at a smaller scale. The satellites are tiny and cheap, and they refresh them every few months. SpaceX is just doing the same thing at a much larger scale and for comms instead of imaging.
That makes more sense. If Space-X can intentionally crash the satellites into the ocean anytime they want then yeah it may not be a bad idea.
My understanding is Starlink wants to add more satellites than have been added since we started adding satellites until the present day. Plus each collision creates hundreds of pieces of junk which can cause more crashes.
It’s true that this would be more satellites than have been produced to date, but it’s really not a big deal. Space is big: 11,000 satellites over the surface of the Earth gives each one ~18,000 square miles. Of course this is a bit of a silly metric since the satellites are constantly moving around, but in practice that just makes it easier; the satellites will be in a set of slightly different orbits that won’t intersect. So really we’re talking about each satellite getting its own enormous volume of space.
The control systems, the tracking systems, and basically everything else are superior now to what we’ve had before. Caring at all about what happens at the end of a satellite’s life is more than what satellites have gotten for most of their history (this isn’t just SpaceX’s doing, of course, but it is a relatively recent development).
There’s plenty of room for millions of satellites if desired. And one day we’ll be able to just scoop up the dead ones, though that’s probably a little ways off.
Space is REALLY big. More specifically, the space of Low Earth Orbit is REALLY big. 11,000 is absolutely nothing. Let’s put that in perspective:
There are an estimated 200,000 leaves on your average large tree. A leaf is certainly smaller than a satellite - but don’t you think if we spread the leaves of just one large tree across the entire globe, there’d be a vanishingly small chance of two of them running into each other when the wind picks up?
To be clear, my point wasn’t that space is so big that we can ignore the possibility of collisions–as you note, they can and do happen. Just that as long as we are careful–maintaining active control, using avoidance maneuvers, deorbiting obsolete units, and so on–there is plenty of room for lots of satellites to coexist. The short intended lifetime of the SpaceX constellation is a boon here since it there is less time for the deorbiting systems to malfunction.
The bigger threat is the uncontrolled junk already up there (like the Russian satellite that collided with the Iridium one). We need to do a better job at producing less junk up there in the first place.
Apparently, inaccurate NORAD conjunction reports were a factor in the Cosmos 2251/Iridium 33 collision. It may be that this needs to be improved before putting up large constellations, but I don’t think there is any reason that can’t happen.
Well, sure, SSA can be improved as long as someone else pays the bill. I’m pretty sure the first new Air Force Space Fence radar is on track to become operational fairly soon, and there are thoughts of building a second site (though I’ve lost the bubble on those discussions). If someone wants to cough up about another $700 million for a second site to better track the huge number of commercial satellites planned to be put in LEO, I’m sure the JSpOC would love to have that data.
The leaves aren’t moving in potentially intersecting orbits at 10,000 mph. But your point is still valid. It would be difficult to collide two of these satellites intentionally.
Sounds like a good investment to me. If only there were a way for individuals and organizations to pool their resources to prevent “tragedy of the commons” situations like this one…
Great. The handful of companies that are looking to start this industry can start writing checks to he US Government so it can better address the issue they are creating. Problem solved!
The issue of course is that although it’s in everyone’s interest to pay for resource management, no particular organization wants to pay for it unless everyone’s paying for it.
Which probably means that every US satellite–and ideally, those from cooperating nations as well–needs to pay a tracking fee to support these services. That’s fine, but it’s not likely to be driven from the bottom up. Someone (from NORAD, or wherever) needs to make the argument that future space development depends on effective management and the cost of these services should be passed onto the appropriate parties.
Then again, it might be simpler to just treat it as a basic government service, like GPS, which is already in the government’s interest to pay for, and which costs nothing extra to extend to the public. $700M is pretty cheap compared to the public benefit from Starlink and other constellations.
So are 11,000 satellites required because of the low Earth orbit? And the latency would be a lot better than Hughesnet or whatever? Did Hughesnet go with fewer satellites higher up for cost/simplicity? Because current satellite offerings suck for both latency and allowed bandwidth.