Starlink is SpaceX’s satellite internet service that everybody is excited about. But… Satellite internet has been around for a long time. And now, there are several competitors working on similar projects. So let’s look at what’s so attractive about LEO satellite internet and how Starlink’s competitors stack up.
You know, sometimes I think we’re a little too hard on ourselves.You know, sometimes I think we’re a little too hard on ourselves.
Yes, the world is garbage right now, fascism is on the rise, nobody can agree on anything, we clearly are not very good at being an internet species.
But really… Why would we be?
This level of connectivity has never been experienced before in the 200,000 year history of our species.
I mean up until just a couple hundred years ago most people rarely even traveled one town over because the only way to get around was on a horse, and most people didn’t even have those.
Things had always been this way. They had never not been this way. Most people didn’t even know how to read because everything they needed to learn came from the hundred or so people they knew in the village around them.
And every technological leap in access to information has been followed by a period of social and political upheaval.
I mean, the printing press was invented in 1440 and 7 decades later the Christian church split in to protestants and catholics. And they’re still fighting about it.(on screen:Guttenberg’s printing press – 1440Martin Luther’s Ninety-Five Theses – 1517)
Today, we have these things in our pockets with the compendium of all human knowledge and the ability to connect with literally anybody in the world at the tap of a finger. And this is literally like 15 years old. And the internet that powers it is barely 30 years old.
I mean, of course we’re going to be really bad at this.
True story, I have been skiing once in my entire life. And it was a nightmare.
I was falling over left and right, literally couldn’t go more than 10 feet without beefing it, it was super embarrassing but the worst part about is was that for some reason… I thought I would be good at this.
I’d seen skiing on TV and in the movies, it looked like a lot of fun, and I’d played sports, I had good reaction speed, I rode a skateboard back in the day, I got this.
Turns out sliding down a mountain covered in snow, when you have spent your entire life in the flattest place on Earth, which almost never gets any snow… is terrifying.
It was like I was driving through San Francisco on black ice and bald tires.
And I kinda melted down at the whole thing. Because I had this expectation – a ridiculous one in hindsight – that I should be good at this.
I feel like that’s the whole world with the internet right now.
We’re really beating ourselves up for how badly we’re bungling this, I’ve done plenty of it myself. But… We are literally learning how to share the world with a digital superorganism. One that we created.
So maybe the expectation that we would be good at this is as ridiculous as me thinking I could ski.
And yet, we continue to lean into it. Because that’s what we’ve always done, and despite the initial disruptions to the fabric of society, over the long run our quality of life has gone up because of it.
And soon, if SpaceX has their way, the entire planet will be covered with fast, cheap satellite internet. That’s the Starlink project that we all know about but SpaceX isn’t the only company working on a satellite internet swarm.
There are several in the works actually. So let’s take a look at why this is such an attractive idea to these companies, which ones are likely to succeed, and what this means for the world as we know it.
History of Satellites
So, everybody’s pumped for satellite internet with Starlink. But here’s the thing — satellite internet has been around for a while.
In fact, Telstar 1, the first US communication satellite was launched in 1962, just 4 years after Explorer 1, which I talked about in my video about the Van Allen Belts last month.
It was designed at Bell Labs for AT&T, and it was the first satellite that beamed live TV to the US and Europe. It also carried the first satellite phone call.
And… it made the first satellite data transmission between two computers. In 1962.
So technically there was satellite internet… before there was an internet.
Unfortunately, Telstar 1 was functional for less than a year. And the reason for that can also be found in my Van Allen Belts video. I was talking about a nuclear test called Starfish Prime that was aimed at the Van Allen Belts: Yeah, this was one of those satellites.
Eighteen months later, a satellite named Syncom 3 became the first to launch into geostationary orbit.It provided live coverage of the Tokyo Olympic Games
And this was a big deal because geostationary orbit means that the satellite is in the same relative position over Earth at all times. It does this because it’s far enough away that the orbital speed matches the speed of the Earth’s rotation.
It’s really far away, like 35,786 kilometers (22,236 miles) far away, but it allows you to send and receive signals 24/7.
Whereas a satellite in low Earth orbit would be whizzing by overhead and could only send and receive signals in short windows.
Geostationary Earth Orbit
Plus you can cover a massive area from Geostationary orbit, in fact, it only takes 3 satellites to cover the entire planet.
So yeah, the communications industry has used Geostationary orbit ever since, increasing the power of their satellites over the years to accommodate the increased traffic.
Just a quick side story here, my grandparents when I was growing up lived on a ranch out in the middle of nowhere and they didn’t have access to cable so their only option was satellite TV.
But this was way before the direcTV dishes you see on the side of people’s houses, no their dish was like out in a field and it was like 8 feet wide.
I could have recreated the Contact poster with that thing, I’m not kidding.
And every time you changed the channel, you could look out the window and see it… turning…
It was a lot of work just to watch Tales From The Crypt, I gotta say.
So yeah, these satellites provided live events, and phone calls and movies young boys shouldn’t be watching, but as soon as the internet became a thing, that was on there too.
DirecPC to HughesNet
In 1996 Hughes Network Systems launched the first satellite internet service called DirecPC. I mentioned DirecTV a second ago, well that was Hughes Network systems, they created DirecTV in 1994, that streamed TV, DirecPC carried internet. You get it.
Hughes Network Systems by the way, they were a real pioneer in early satellite communication – that SynCom 3 satellite I just talked about, the first one in geostationary orbit, that was them. Or… it was their parent company, Hughes Aircraft Company. And if that name sounds familiar, that’s because it was founded by Howard Hughes. Yes. THE Howard Hughes. Who Leonardo DiCaprio played in The Aviator, the guy from the Rocketeer, the Spruce Goose guy.
You did not think this was gonna go all the way back to Howard Hughes, did you?
Anyway, DirecPC launched to mixed reviews, but it was faster than dial-up internet in those days.
But as terrestrial services like cable and DSL caught on, satellite fell behind. For two main reasons.
The first is that satellite systems are super expensive to upgrade.
Case in point, HughesNet — which is what Hughes Network Systems goes by now — they have satellites capable of over 100 Megabits per second download speed. But, they only advertise up to 25 Mbps.
That’s because they don’t have enough satellites to give every customer full-speed, so they have to restrict the amount each customer can get.– has a breakdown on bandwidth vs speed
They hope to fix this with a launch this year, but this single satellite cost the company $400 million. And that doesn’t include launch costs.
Launch Costs to GEO
Another company, Viasat, is planning two new satellite launches for the same reason, they’re expecting to pay between 1.2 and 1.4 billion dollars for the two of them.
That’s a lot of money for these companies to invest, and keep in mind that will only get their download speeds up to 100Mbps. Which is way better, yes, but doesn’t really blow your hair back compared to cable and fiber internet services.
The Latency Problem
So the expense of it all is the first problem, the second big problem is latency.
And this is simply not something that can be fixed from geostationary orbit. Not without breaking the laws of physics.
Latency refers how long it takes for a user to make a request from an internet provider’s network and get a response
You can think of it as the time between clicking a link and loading a web page
When the nearest node in a provider’s network is 35,786 kilometers above your head, getting a response will take some time
One way of looking at this is if you want to send a message to someone just a mile away via GEO satellite, the distance that signal has to travel to get there and back is basically the same as sending the message all the way around the planet. Just to go one mile.
The absolute minimum latency for a GEO round trip is about 240 milliseconds. HughesNet and Viasat average over 600 milliseconds. In more real-life terms that means a little more than half a second.
For comparison, terrestrial internet can go as low as 10 milliseconds. In practice, anything under 100 is considered fine for everything except gaming.
LEO in the 1990s
Half a second may not sound like a big deal, but satellite internet customers have been complaining about latency since the early days of HughesNet.
It’s just clunkier, and especially as more work has gone online, more communication has gone online, this starts to matter even more, and by the way, in some financial applications it can actually be a detriment.
And yeah, the only way to fix this problem is to bring the satellites closer to Earth. But to do the same thing in LEO, you need a lot of satellites.
Where GEO satellites can hit 1/3 of the planet from 36,000 kilometers up (22,000 miles), something from lower than 2000 km in LEO can only hit a small area, plus there’s the fact that in order to stay in orbit in LEO, that small area is moving across the surface of the Earth at 25,000 kilometers per hour. So the only way to cover the entire circumference of the Earth is to have a bunch of satellites that can relay back and forth with each other.
Passing the signal from one spot on the ground to the other like a game of hot potato. This drastically reduces the distance the signal has to travel, but again, it takes a LOT of satellites to do this, especially considering the Earth is a sphere.
In the interest of overcommunication, there are some orbits in between GEO and LEO like the Molinya (mole-nya) and Tundra orbits that can provide full coverage with just a handful of satellites, these are mostly used for things like GPS though.
But the idea of an LEO constellation of satellites to provide internet service goes all the way back to the 90s.
Teledesic was formed in 1994, backed by Bill Gates and they planned to build a constellation of 840 satellites. This later got scaled back to 288, and then to none. The company folded in 2002.
Several other companies have tried this as well and they all failed. The cost of launching that many satellites is just too high. The idea is solid, most people agree that it would work and probably be better than geostationary satellites. If anybody can afford to get them up there.
Which brings us to Starlink.
SpaceX, with their reusable rockets have brought the cost of launches down low enough that this crazy idea is actually kinda possible.
Especially combined with the fact that satellite technology is smaller than ever before with cubesats becoming popular, you can launch hundreds of these things at a time.
So yeah, if you’re SpaceX, it’s kind-of a no-brainer.
Assuming it’s that much better than traditional satellite internet.
Right now, Starlink has around 2000 Starlink satellites at 550-570 kilometers up and over 145,000 customers so far. And according to Ookla, they’re averaging at around 97.23 megabits per second download speeds. For context, HughesNet’s Geostationary satellites are averaging 19.3 Mbps.(Disclaimer: 1468 of the satellites are currently active, others were prototypes, have gone offline, or been deorbited)
As for latency, a camper on Pikes Peak recently recorded between 34 and 36 milliseconds. But… he was on a mountain.
SpaceX is promising speeds will eventually reach 300Mbps, but as more customers use it, it’s actually starting to slow down. But considering they still have like 40,000 satellites to go, it looks promising.
No wonder Viasat has tried to get their FCC license revoked.
Last year they were part of a group that filed papers saying the Starlink launches would be bad for the environment. No conflict of interest there. Actually, it gets funnier because those two super expensive satellites they’re scheduled to launch? They’re going up on a Falcon Heavy.
Kiss the Sky Goodbye
Conflict of interest, maybe… But are they wrong?
42,000 satellites is a LOT of satellites. It’s actually eight times more than all the satellites in orbit when they started launching Starlink satellites in May 2019.
The reason for so many is because they want to create multiple shells of coverage, with constellations at different altitudes.
Current plans call for 5 shells, all with 20 vertical miles between them.
All right, now might be time to talk about the elephant in the room.
From almost the very first Starlink launch, concerns have been floated around about the effect Starlink would have on ground-based astronomy.
Especially in some of the earliest prototypes, they reflected a lot of sunlight and had a lot of people worried about what that would look like with a sky full of them.
They have made some progress since then to add less reflective coatings and “visors” that help reduce the glare, but it’s not perfect.
Honestly I never knew how to feel about that argument because anything Elon Musk related is going to face a ton of criticism because he’s so polarizing, sometimes it’s hard to separate the signal from the noise.
Like I know these telescopes have powerful algorithms to account for atmospheric disturbances, I just assumed they would be able to account for satellites as well, I mean those are nothing new, they’ve been up there for a long time.
So how much of a problem is this really? I decided to ask someone who knows far more about this than myself, so recently, I had Dr. Becky Smethurst on my podcast where we spent most of the time talking about the James Webb Space Telescope, but while she was there, I asked her about the Starlink problem. Here’s what she had to say:
So it’s not an insurmountable problem. But it is a problem.
And while it might feel satisfying to point fingers at SpaceX and Elon, the fact of the matter is, it might be an inevitable problem.
Because people have been wanting to do this for a long time, since before SpaceX even existed. And there are multiple companies working on similar constellations right now. Let’s take a look at some of them.
Project Kuiper was first announced in April 2019, with the goal of launching 3,236 satellites to an orbit of between 590 to 630 kilometers (370 to 390 miles), slightly higher than Starlink.
But they’ll be using more robust satellites that can support up to 400 Mbps download speed, which would be faster than Starlink.
So far, Amazon has invested more than $10 billion in Project Kuiper, and there have been some speed bumps along the way, but they plan on launching two prototype satellites in the 4th quarter of this year.
As for who will launch these satellites, for now they’re using ULA’s Atlas V and the ABL Space Systems new RS1 rocket but you know once Blue Origin gets the New Glenn off the ground – at this rate sometime in 2040 – they’ll be using that to get them up there.
Regardless of what they launch on, their license with the FCC requires them to launch half the constellation by 2026, so expect to hear a lot about this in the next few years.
Another company working on a satellite constellation is OneWeb.
OneWeb is based out of the UK and at one point, they were planning on a constellation of nearly 50,000 satellites.
Then they went through funding issues, reportedly due to the pandemic, this led to a bankruptcy, and once they came out of bankruptcy, their plans slimmed down to 6372 satellites.
Of which so far, they’ve launched 394.
These are in a much higher orbit than Starlink and Kuiper, at 1200 kilometers (745 miles), but so far they seem to be doing really well. A test in 2019 showed download speeds of 400 Mbps with a 32 millisecond latency.
A more recent test in 2021 got 165 Mbps and 45 ms latency, so it’s clearly fluctuating but it’s competitive.
Now one thing that differentiates OneWeb from Starlink and Kuiper is they’re targeting commercial uses instead of residential consumers.
Hughes Network Systems, which I mentioned before, is an investor in OneWeb and according to their press release, they’re targeting “enterprise, government, commercial aviation and maritime, cellular backhaul, and community Wi-Fi hotspots” to quote the press release.
Man, Hughes really wants to do this constellation internet thing.
One more quick thing, because they’re targeting commercial customers, they can charge more for their terminals, so their terminals are going to be between 1000 and 1500 dollars, as opposed to Starlink’s terminals which go for $500.
It should be noted however that SpaceX is selling Starlink terminals at a loss for now, they apparently cost $1300 to make. But that’s their gamble to get early adopters on board.
Last, but not least, we come to the Canadian company Telesat.
Telesat has been in satellite communications since 1969, and they were founded as a crown-owned company, meaning the government of Canada holds a significant interest.
A deal in 2021 injected 1.44 billion Canadian dollars into the company. Canadian dollars are just like American dollars, except they’re all stuck together from the maple syrup.
Most of that investment was to fund Telesat’s “Lightspeed” satellite constellation
Much like some others on this list, they’ve changed the proposed size of their constellation multiple times.
In 2016 they announced plans for 117 satellites in LEO, this was later raised to 209, then 298, which is the current plan, BUT… They applied to launch a total of 1671 satellites, just to ensure they can meet future demand.
78 were slated to go up this year in 2022, but dates have slipped to 2023. The reason for that… Is because they’re supposed to go up on the New Glenn. (shrug)
But that’s okay, they may need the extra time, they’ve apparently had trouble getting satellites built because of global supply chain issues.
Like OneWeb, Lightspeed satellites will orbit at a higher altitude, around 1000 kilometers, and they’re expected to have latencies in the 30-50 millisecond range
Also like OneWeb, Telesat plans to target businesses and governments, as well as existing satellite customers like airlines and cruise ships.
But here’s the twist: One of the conditions of the government funding is that they also provide services to indigenous communities. Which they will do a better job of because they’ll be flying in a hybrid orbit that sweeps over the poles, so they’ll cover areas that Starlink won’t.
A Thousand Points of Light
So that’s what we can expect in the coming years but before we wrap this up, there is one more concern regarding these satellite swarms that’s worth talking about.
We already talked about the astronomy problem. But there’s also the problem of… what happens when there’s that much stuff in space?
I’ve talked on here before about the possibility of Kessler syndrome, which is when collisions in space create space debris which cause more collisions and on and on until the planet is trapped inside a shell of metal debris circling the Earth 15 times faster than a speeding bullet that will shred anything that tries to go through it.
Not only would it trap us here on Earth, it would remove our ability to use satellites at all, which would be a massive setback for our species.
There’s a lot we totally take for granted that are only made possible by the satellites circling overhead.
And obviously doubling or quadrupling or octupling the number of satellites in orbit only increases the potential of something like that happening. So is this a bad idea?
It’s something to be concerned about for sure but one could make the argument that these Low Earth Orbit swarms might be safer in the long run than launching stuff up to geostationary orbits.
Because satellites in low Earth orbits decay really quickly. There’s actually tiny amounts of atmosphere going up hundreds of miles that creates minute drag forces on LEO satellites that over time slow them down enough that they do eventually burn up in the atmosphere.
Satellites in the general orbit of Starlink would come down in a few years, maybe even less than a year.
Whereas satellites in geostationary orbit are essentially there forever, like thousands of years.
And for that, there is a company worth mentioning and that’s Privateer Space.
Just launched last year and back by Apple Co-Founder Steve Wozniak, Privateer Space aims to be focused on “space environmentalism”
Their plan is to monetize the removal and recycling of space junk – how exactly they plan to do this has not been fully announced but if you want to know more about it, their chief scientific adviser Moriba Jah talked about it on Startalk, I’ll put a link to that down below.
It’s probably way too soon to tell if their plan is the ultimate fix for space junk, but it’ll be interesting to see how that comes along.
So, what’s the verdict on these satellite internet swarms? I’ll be the first to admit I haven’t been that bullish on Starlink personally, I don’t think it’s a replacement for terrestrial internet, and like I said before, satellite internet has been around for a long time, is it really worth clouding our skies for a split-second less latency?
It just always seemed like more of a niche application to me but that’s really easy for me to say sitting where I am in the middle of a city with gigabit internet service, if I lived out in the middle of nowhere or was in an underserved indigenous community, something like this might be a godsend.
Having access to cheap, fast internet anywhere in the world is probably a good thing. And with the workforce becoming more mobile these days, which has been completely accelerated by Covid, yeah, something like this might be a game changer.
I’ve become kinda fascinated by the whole van life movement lately and the fact that you can live and work effortlessly anywhere you want. I have friends that are living like that and they send pictures of their “office” that day and it’s like working in a freaking postcard, it’s crazy.
But I’m interested to hear what you think, is this a service you could get some use out of? Are you a Starlink beta tester, and if so what has been your experience with that? Or, do you think it’s a dangerously ill-conceived idea that’s going to ruin astronomy and clutter up the skies? Discuss.