[Normal_gaussian]

I think that "Africans should not have access to high-speed satellite Internet" is something you've just made up; the article you link talks about African people turning to Starlink because of local infrastructure issues, and the original article notes that the current satellite count is currently around 14k satellites. 100k is more than enough satellites to provide high-speed satellite internet globally.

The article makes mention of specific endeavours, like the night-time mirror satellites, which are particularly disruptive to astronomy, and the general risks of high numbers of satellites.

The ability to do Earth based astronomy is something that is of value to all the peoples of Earth, and is mainly funded by the western nations because of their current position as the people with more money.

[JumpCrisscross]

The article doesn’t consider that in a world with a million satellites in orbit, launching space-based telescopes—including into deep space—becomes an order of magnitude cheaper.

[acdha]

Does it? My understanding was that it’s less helpful for anything which isn’t in low-earth orbit because the commercial launch engineers are optimizing for the lucrative satellite business, not larger and higher payloads.

[JumpCrisscross]

> commercial launch engineers are optimizing for the lucrative satellite business, not larger and higher payloads

Commercial satellites are getting bigger and heavier. Launch that can put big and heavy in LEO can put big and slightly less heavy higher up. Add to that things like in-orbit propellant transfer and there is a good chance astronomy sees a golden age in the coming decades (in countries with space access).

I’m not dismissing the problem. Just this analysis as meriting any conclusions. It’s a start. But it’s only part of a full model of how these changes would affect astronomy.

[Teever]

I get where you're coming from but we haven't really seen any sort of space based telescope designs that take advantage of the Falcon launch paradigm of cheap and reliable launches.

Some sort of modular telescope array that could be launched in pieces and self-assemble in orbit. Something that improves in capacity as more pieces are added.

Everything seems to have stalled in this field, as if it's just waiting for a Starship which may never come.

[m4rtink]

There are hundreds of Dove Satellites with telescopes in the Flock constellations from Planet Lab: https://en.wikipedia.org/wiki/Planet_Labs

Most of them launched on the Falcon 9. They just point down to the Earth by default. ;-)

And there are other cubesats testing stuff for x-ray astronomy or gamma ray burst detection, such as:

https://en.wikipedia.org/wiki/VZLUSAT-2

[JumpCrisscross]

> we haven't really seen any sort of space based telescope designs

We’re only starting to truly mass manufacture satellites. A world with millions of satellites means one with lots of satellite production and design economies of scale. (Same for all manner of sensors and optics.)

> as if it's just waiting for a Starship which may never come

Or it may. We’ll know in a couple years. Building a scaling production system for Falcon right now would be silly.

And if Starship never works out, we probably don’t see millions of satellites. It’s a fundamentally tied problem, which is why I say the analysis is incomplete.

[Tuna-Fish]

Larger rockets are inherently more efficient, which is why all the commercial providers are moving towards them. And while yes, most of the providers are targeting primarily for LEO, if you have high payload capacity to LEO you can solve your issue of getting anywhere by packing in a kick stage. And cheap third-party kick stages are available and more are in development.

[onion2k]

Google says that James Webb telescope cost a total of $10bn. That's not a lot of money in the grand scheme of things. Private citizens could afford to put similar things into space if they chose to. We don't need them to be cheaper.

[JumpCrisscross]

> James Webb telescope cost a total of $10bn

I’d love to see an estimate of what a JWST-class telescope would cost to design, build and launch in a world maintaining a million-satellite fleet. My guess is less than $2bn.

[HPsquared]

A big part of the complexity of JWST was the way everything had to fold and fit into a small launch vehicle. With larger vehicles (e.g. Starship) the JWST mirror could have been built in one piece, which is much simpler. The whole thing was engineered to the max to fit within tight constraints, which is very expensive.

[JumpCrisscross]

> With larger vehicles (e.g. Starship) the JWST mirror could have been built in one piece, which is much simpler

Not just the complexity of design, but also cost and complexity of cryo-vacuum testing hundreds of deployment mechanisms any one of whose failure critcially endangers the project. The mirror could also be conventionally manufactured versus requiring gold-plated beryllium [1].

[1] https://en.wikipedia.org/wiki/Optical_Telescope_Element

[m4rtink]

Not to mention it would be much cheaper and quicker to launch if it could have any form of robotic or human servicing/tuning available.

Designing it to not fail at the start & then work without any help for decades must have also made it much more expensive.

[ben_w]

Currently writing a draft blog post on all the issues (and non-issues) with these things, it is now long enough (7k words) I'm slightly wondering if it's less "a blog post" and more "one section of a decent sized book on why we can't have nice things".

Here's a visual to consider the implications of things you can do with actually one million satellites of the kind of size scale being discussed:

https://raw.githubusercontent.com/BenWheatley/blog/refs/head...

[JumpCrisscross]

> Yes, they really would be this closely spaced: Earth's circumference is 40 million meters

Satellites don’t orbit on the ground, which makes the 40m spacing nonsense. And nobody proposes putting a million 120 kW satellites in a single orbit.

They really would never be that closely spaced. To approach those densities in a single orbital shell you’d need hundreds of billions of birds in orbit. Spread across all of LEO (and only LEO) we’re talking orders of magnitudes more satellites (like, quadrillions).

[ben_w]

> Satellites don’t orbit on the ground, which makes the 40m spacing nonsense.

Hence why the horizontal scale bar says "40 m to 43 m": Going to 500 km doesn't add much to the orbit's circumference.

> And nobody proposes putting a million 120 kW satellites in a single orbit.

One of my tentative conclusions is that it would be an improvement if they did. But also, there's better contiguous structures to build if you could put that much up.

It's in my blog post because I'm considering all possible arrangements of ways to do this. Current list:

  • Spread them out by altitude while still keeping them in sun-synchronous low earth orbit like SpaceX plan
  • Put them all of them in a single sun-synchronous low earth orbit so none of them can hit each other
  • Spread them out like Starlink currently is
  • Have swarms, where each group has many satellites significantly closer to each other than the usual safety separation, like Google's Project Suncatcher
  • Have fewer, bigger satellites, like Starcloud

> To approach those densities in a single orbital shell you’d need hundreds of billions of birds in orbit. Spread across all of LEO (and only LEO) we’re talking orders of magnitudes more satellites (like, quadrillions).

Matters less than I expected when I started writing. How much so depends on what I end up adding by treating gaps in "full" (up to the safety margin) orbits as the thing of interest and seeing if someone's done a version of this on spherical geometry and also add a time component: https://math.stackexchange.com/questions/270937/how-can-you-...

(That I'm researching such questions should explain both why I'm at 7k words and why I've not published it yet).

(One big surprise to me while researching this: along-track, cross-track, and radial positional error per unit time are all much higher than I would have guessed: https://digitalcommons.usu.edu/smallsat/2015/all2015/58/).

[HPsquared]

Can all orbits be completely filled at once, though? They'll intersect at some point (I had originally said poles but that's only polar orbits..) ... I suppose you have phase, altitude and inclination (and eccentricity which adds another couple of variables). But they do intersect, don't they?

[JumpCrisscross]

> Can all orbits be completely filled at once, though? They’ll intersect

Correct. I wasn’t proposing a realistic configuration. Just showing why OP’s visual doesn’t work for the numbers it gives. (It 1D space fills. I expand that to 2 and 3D.)

Millions of satellites is currently accepted as the maximum carrying capacity of LEO before collisions becomes a PITA.

[ben_w]

Can't take advantage of the 3D-ness when you want them all to be in terminator-following SSOs. Best you get is 2D, and with current safety margins, even that fills up fast.

However, this was not the point of the illustration, which is more "if you can do this at all, you can do better".

[rcxdude]

Not to mention starlink is not a solution for internet for everyone on the planet: it cannot serve everyone in a densely populated area, no matter how many satellites they have in their constellation. It's a useful piece of infrastructure, but it's far from the panacea people seem to think it is.

[ethin]

Starlink is already over-congested and physics says that it is not possible to scale it up to serve 100000000 subscribers let alone 5-6 billion or more. We would need some kind of physics breakthrough for that to scale properly, and I'm not even sure if physically it would even be possible to do that no matter what you threw at it. Starlink isn't magic as a lot of people seem to think it is.

[IncreasePosts]

Ground based infrastructure is much easier to justify in densely populated areas. So, dense areas get ground infra, and the dispersed rural population can get satellite infra

[fc417fc802]

> it cannot serve everyone in a densely populated area,

I also suspect that to be the case but in order to be more objective I wonder. What's the theoretical maximum bandwidth per square meter (or other unit area) that it can deliver?

[unusualmonkey]

Starlink is useful, but people often confuse coverage with capacity density.

The rough hierarchy is: fixed fiber/cable > Wi-Fi > cellular > LEO satellite > traditional GEO satellite. Starlink is a huge improvement over old satellite internet, but it is still fundamentally a satellite system, not dense terrestrial infrastructure.

As a rough illustration, suppose one Starlink beam covers about 63 square miles and has 6 Gbps of usable downlink. At New York City density, that footprint contains about 1.85 million people. That works out to roughly 3 kbps per person. For comparison, network planners typically budget on the order of Mbps per person to comfortably support peak demand in dense urban environments.

You can improve that with more satellites, more beams, more spectrum, and better hardware. But even a 100x improvement only gets you to about 300 kbps/person. Additionally, this all adds cost... that isn't needed for the vast majority of time or space that the sattelites will be over. If the sattelites orbiting over New York also orbit over the Sahara Desert - every $ spent improving NY capacity is also improving the Sahara... but with no return.

The reason fiber, cable, Wi-Fi, and cellular work so well in cities is spatial reuse. Capacity can be reused block by block, building by building, apartment by apartment, tower by tower, and access point by access point. A beam from hundreds of kilometers overhead covers a much larger area.

So Starlink is excellent for rural areas, ships, aircraft, remote sites, disaster recovery, and backup connectivity. But it is not a replacement for terrestrial broadband in dense cities. It solves coverage much better than it solves urban capacity.

[fc417fc802]

> suppose one Starlink beam covers about 63 square miles

Is that the actual figure though?

> even a 100x improvement only gets you to about 300 kbps/person.

TBF even that would be quite useful for lots of things seeing as it's faster than the DSL service I once had. However agreed that it isn't going to replace fiber service and overbuilding to such an extent for outliers would be absurd.

Still, that's quite decent considering it's an estimate of a near worst case scenario.

[ck2]

note that including dead things in orbit

(that we currently have no way to remove)

is actually 32,000 not just 14,000

what we need is the investment for "space roombas" that go around bumping things out of orbit that are dead or did not de-orbit properly

the problem is all that atmospheric burnup creates a lot of toxic pollution

* https://www.bloomberg.com/graphics/2025-space-orbit-satellit...

[JumpCrisscross]

> we currently have no way to remove

Nature takes care of this for us in LEO. I’ve seen no serious plans to put millions of anything anywhere else.