Starling satellites have motors, and regularly boost their altitude to help overcome atmospheric drag. That figure is from when the satellite either runs out of fuel, or shuts down.

[–] Thorry84@feddit.nl 4 points 11 months ago* (last edited 11 months ago) (1 children)

Yes this helps with positioning and orbital decay. Almost every satellite has this, it isn't special to Starlink stuff. I know Elon makes it sound like they've invented the wheel here, but much of what they do has been done in one way or another.

There are a couple of factors which impact the lifespan of these satellites:

Technological progression. As they refine the technology and techniques they need to update the satellites with the latest and greatest. This means of course removing the old satellite and replacing it with a new one. Especially in the early days (now) this is a huge factor in replacing their stuff.
Failure rates. This is mostly due to radiation, but may also be due to other factors. The network is only as good as its nodes, so failing nodes need to be replaced fast. Radiation hardening is expensive and usually adds weight. This is a trade off between launch costs, the number of satellites they can fit in a Falcon 9 and lifespan of the stuff. Things like solar storms can have a huge impact, as Starlink found out the hard way.
Fuel consumption. Exact positioning is important for Starlink and with their VLEO orbits drag is a big factor. The satellite have very cool engines that help them stay in place, but only a limited fuel supply. There is a safe minimum fuel as regulation requires them to de-orbit safely, which takes a lot of fuel. So just running it till it's empty is a no go, they need a good safety margin. They also don't want to start any kind of Kessler syndrome kind of deal, so old spots need to be cleared out before new stuff can go in.

There are other factors, but these are the big ones. Starlink say they are aiming for a total replacement every 5 years, but in practice it's more like 3 years. This is mainly due to the first batches being more prototype like, getting nearer to a final design recently.

With the proposed 11.000 unit constellation and the 5 year replacement rate, they would need more than 1 Falcon 9 launch each week. The costs are literally astronomical and the revenue has been only a fraction of what Elon sold the investors. I would be surprised if the plug is pulled on the whole Starlink thing.

People seem to think Starlink is the first and only one to try this, but it has been tried for decades and almost all have failed. The only success is with companies targeting niches, where there is little to no competition and premium rates can be had. For example reporters in the field broadcasting from a van to a satellite to be live on TV was a big niche. So far Starlink hasn't delivered on a lot of the promises made by Elon and is destined to fail unless something big changes.

[–] Ilovethebomb@lemm.ee 1 points 11 months ago (1 children)

I didn't realize the sats had such a short lifespan, I thought it was closer to eight years.

Although, there are hundreds of millions of people around the world who are potential customers, and I've spoken to a few people who either are, or plan to be, a customer. I do think the market exists.

[–] Thorry84@feddit.nl 3 points 11 months ago

Hundreds of millions would be a lot. I think you overestimate the demand for something like Starlink a lot. People who can afford to pay and would consider paying for Starlink tend to live in well developed countries. These countries typically have internet connections which are better than wat Starlink offers. Statistically most people live in cities, which also typically have good internet. People who live in lesser developed countries and don't live in cities tend to not be able to afford or willing to pay for Starlink. Usually there are other cheaper options available, even though they would offer less bandwidth than Starlink. So the total market would not be hundreds of millions.

Starlink also offers poor bandwidth and latencies compared to local solutions. People who just use things like Facebook would rather have a low latency and low bandwidth solution than a high inconsistent latency and high bandwidth solution. Starlink is getting better, but the latency, especially in regions with few base stations (which is their best use case) will be inherently poor compared to wired or local wireless solutions.

Starlink themselves thought they would have 20 million subscribers in 2022. In reality they managed just about 1.5 million. (It's not clear how accurate these numbers are and if they include non paying customers) They could get more people on board if they lower pricing, but then they need more customers to get the same revenue. Since the costs of building and launching the satellites, managing them and maintaining the ground part of the system are fixed and high, they need to generate a lot of revenue to turn a profit.

There may be large parts of underdeveloped areas in the US for example where people have the need for high bandwidth internet and are able to afford it and local solutions are lacking. But you end up with only 50 potential customers for one area of which maybe 5-10 people actually sign up. As soon as you hit something like a town, local wired and wireless internet solutions will outcompete Starlink easily. In a poorer country there may be more people to be found in rural areas, but if you only make the equivalent of $5000 a year, you probably won't spend more than $1000 for Starlink. For those people the budget they have for internet would be more like $50 a year max.

And remember even if Starlink starts to operate at a profit, they aren't out of the woods yet. They have had huge upfront starting costs, much more than they expected. Those costs need to be covered before investers actually get anything.

All the while they are competing with local internet solutions which are being rolled out fast all around the world. Something like 5G is rapidly cutting into the need for something like Starlink. As soon as subscriber count starts dropping instead of rising, it's all over.