That's the (honest) question I have. Articles like this pop up from time to time on HN, but it's hard to tell how big a problem this is. What's the reasonable upper bounds for LEO satellites? How do we determine what is reasonable? Who gets to decide and enforce that?
We can probably do a back of the napkin estimate on this.
The radius of the earth is 6563 km, and LEO is 2,000 km or less.
If we use 2,000 km for LEO and round the radius of the earth to 6,500 km, the "radius of LEO" is 8,500km.
Surface area is 4 pi r^2... so that's a surface area of about 908 million square kilometers.
There's a couple ways we could try to guesstimate the upper limit of satellites.
How many satellites per square kilometer? Maybe 1?
If so, that means 908 million satellites, which is far more than the 114,000 the Canadian company wants to launch. But if we start seeing many deployments in the millions of satellites... that starts to add up pretty quickly.
Given the speed of the satellites, is there a certain amount of distance travelled per second that we need to ensure is kept between them, or significant gaps to leave windows for other launches to pass through safely?
LEO velocity is about 25,000km/hr, so about 7km/second. If we want a gap of 1 second between satellites (I am not sure if that's excessive or insufficient, just a ballpark) but assuming the orbits are mostly "aligned" so we can keep 1km distance in one axis and 7km distance in the other axis, we can round that to about 1 satellite per 10 square kilometers, so 90.8 million satellites as an upper bound.
On the other hand, maybe clusters of launches can be handled more easily by launch planning software by treating them as a group instead of an individual object to be tracked, and maybe for some reason that makes the process easier.
Also, the above estimates are probably wild underestimates, because these are based on the surface area of a spherical shell, disregarding all the shells are different altitudes of the sphere... for example, LEO was taken to be 2,000km, but there are nearly equivalent shells at 1,999km, 1,9998km, etc. Even if we assume 10km between shells for our back of the napkin estimate, if we say anything from 1,500km to 2,000km is basically equivalent, we have 50x more capacity than estimated above.
Going in the reverse direction, if we assume that for some unknown reason these estimates are off by 2 orders of magnitude, 100x, after accounting for all the equivalent "shells", we're back at roughly the original capacity estimate of 90-900 million satellites.
Good point! So we would want to calculate the surface area excluding the poles?
We could calculate it more accurately, but the surface area excluding the poles is definitely more than 50%, probably closer to 60-80% of the total surface area, right? So the estimate is still mostly reasonable.
I think each great circle has its own poles orthogonal to its inclination and earth's poles has naught to do with it.
Without station keeping, LEO orbits (thankfully) degrade so the outter shells will be gradually bleeding into the lower shells as well.
Another big factor to consider is how obsolete satellites are de-commissioned. Can they be made to re-enter the atmosphere when no longer needed, or is a permanent piece of debris at that point?
The big unknown in the back of the envelope calculation is the degree of control over each satellite. What percentage of the satellites are uncontrolled? For instance, of Starlink's 1600 satellites, over 100 of them have become non-functional, most of them the unused v0.9 test satellites. However, most of these failed satellites have been actively de-orbited. A few failures are non-responsive and are de-orbiting naturally, which takes about a year.
This factor is much less important at the Starlink of 550km because air pressure deorbits those satellites relatively quickly.
If failure rates stay below 0.1% and orbits below 500km and proper international "traffic control" is implemented, billions of satellites would not pose a Kessler risk, IIRC. At higher altitudes that number decreases very rapidly. Sorry I can't find a link to those calculations.
Space is big. Any orbital plane has a diameter of, roughly, 13000km. That gives a sphere with a surface area of about 5.3 * 10^8 square kilometers. We could give every satellite in a single plane 100 square kilometers and still put millions of satellites in it. From the perspective of a rocket launch through a plane that's pretty safe.
But orbital mechanics dictate that satellites on different orbits have their orbits intersect in two points, obviously.
So we need to coordinate Management of orbital planes such that the satellites' orbits in a single plane don't intersect with each other. That means, inside a plane we would have to create "subplanes", instead of a plane one would assign a shell.
Capacity then depends a lot on how much space we want to leave. Say we want 1km "vertically" and 10km "horizontally" and a shell has a thickness of 10km. Then you can still put about 4000 satellites on the same orbit and have 40000 satellites in that shell at most.
So there is a lot of space in space. But coordination will be crucial.
We may as well get it over with. Our launch strategy as a species is so bad that we will inevitably lose access to our orbital region due to the high velocity garbage eventually. If we get to that stage early on, we can stop pretending that governments and companies are capable of acting like adults
Also, has anyone studied the debris? If we go to hundreds of millions of little satellites, that is a significant mass. what is that debris made of, is it reactive, what happens to it?
Is this really reasible to scale for a company without their own launch platform though?
As a Canadian living in a "rural" area, 15km from a municipality of 160,000 people with no access to wired broadband internet, I'd rather see whatever funding gets thrown at this startup into actually developing Canadian terrestrial infrastructure.
As a rant, in the 8 years since living here, I've seen cellular data prices double. Yes, double.
Yes, I won't disagree that running cable or fiber is expensive, but I'd like to see a push to run cable or fiber to places like ours which are on a major highway, and only a few kilometers out of town. Somehow in these discussions running a few kilometers of infrastructure to outlying areas around major cities gets painted with the same brush as running thousands of kilometers of cable to hit every outpost in Nunavut.
The next would be more cellular towers with realistic data plans to service rural data. As an example these Telus plans are fairly representative of what's available. $90 for 20GB, $135 for 50GB. https://www.telus.com/en/mobility/mobile-internet?linktype=g...
So you're OK if it's run a few kilometers to your place, but you don't have a solution for Nunavut?
Alternatively they can spend the limited funds to address the market that includes you and Nunavut and elsewhere around the world since it needs to orbit.
It remains to be seen whether they can really raise the money to do this.
I am a bit skeptical that a double-6U-cubesat will have enough power and RF link budget possible to compete with starlink. I'm saying this as a starlink beta customer for a year now.
If their intended target market is more like a lower-bandwidth competitor to Iridium and Inmarsat's L-band based services for mobile terminals/offshore/industrial/aviation and similar markets, maybe it can find a market niche, but it will have to be a whole lot less costly in dollar per MB.
From the article it looks like they are trying to act as an uplink/downlink service for other satellites rather than anything terrestrial.
>Kepler CEO and co-founder Mina Mitry said the six-year-old privately-held Toronto company only plans to launch 200 of its own small satellites to establish its internet-of-the-sky service called AEther. The rest of the six-figure sum of flying objects would actually be launched by its customers – such as Earth observation services, space tourism operators, space agencies and defence departments – which would affix a cellphone-sized 220-gram terminal provided by Kepler to their own satellites. The Kepler box would function like a SIM card and enable customer satellites to connect to the larger constellation “and any other space-borne assets” in LEO, via the always-on, always-available AEther network.
So their business plan is to buy launches from SpaceX to build their own fleet. Then SpaceX has the booster paid for and can re-use it up to 4 times to launch their StarLink satellites for a fraction of the cost.
Seems like it would be simpler for them to just write a check to Musk...
This might as well be a joke. I am willing to bet anyone here that this will never happen. It would be like trying to start a new company to compete with AMZN selling the same things they sell online. Yeah, good luck with that.
Why is it no one wanted to put up a fleet of Satellites to provide Internet access until SpaceX did it? What makes these companies/countries think they can do it when the expense will be (?) times greater than what SpaceX can do it for?
They are competing with Amazon (the AWS groundstation product)[0], not SpaceX Starlink.
>Kepler CEO and co-founder Mina Mitry said the six-year-old privately-held Toronto company only plans to launch 200 of its own small satellites to establish its internet-of-the-sky service called AEther. The rest of the six-figure sum of flying objects would actually be launched by its customers – such as Earth observation services, space tourism operators, space agencies and defence departments – which would affix a cellphone-sized 220-gram terminal provided by Kepler to their own satellites. The Kepler box would function like a SIM card and enable customer satellites to connect to the larger constellation “and any other space-borne assets” in LEO, via the always-on, always-available AEther network.
At what point will this be cheap enough that we could crowdfund enough satellites to build an Internet that can't be taken down with conventional means?
What rights does my satellite have? Is it like ships, that they must be registered to a nation-state? Or is it total anarchy? e.g. If I am ISIS can I launch a satellite filled with recruitment videos and have it essentially outside the jurisdiction of any country?
> If I am ISIS can I launch a satellite filled with recruitment videos and have it essentially outside the jurisdiction of any country?
sure, in this highly unlikely scenario, if you take into account that the likely result would be the USA air-striking your launch facility into rubble, which is big and fragile.
developing a launch vehicle that can send a 200-300kg payload into low earth orbit on short notice is pretty much the same thing as developing an ICBM, which the nuclear power nation states of the world tend to frown upon.
I don't know enough orbital mechanics to understand the feasibility, but it's fun to consider that putting a shade over some of the hotter cities of the world - Phoenix for example, could have a big effect by reducing the need for air conditioning.
A very cheap and easily implemented idea to help with climate change. That could be done in huge swaths of the world where a lot of people currently live. It probably needs some government help to stir people into action or make them aware of it. A tax break maybe for painting your roof white in a hot country/state?
In northern latitudes it would be nice to be able to change the albedo of the roof to reflect energy in the summer and absorb it in the winter (although the effect in winter will be muted in places where snow accumulates on the roof.) I wonder if there exists anything like that.
This won't just cause issues for earth through collisions. Imagine a mass coronal ejection that would fry the electronics in all orbiting satellites. The more that are up there the more that can come down very quickly and randomly.
That's the (honest) question I have. Articles like this pop up from time to time on HN, but it's hard to tell how big a problem this is. What's the reasonable upper bounds for LEO satellites? How do we determine what is reasonable? Who gets to decide and enforce that?