As one deleted comment stated:
"Stacking images and object detection can pretty much nullify this problem. Worst case scenario you throw out a couple minutes of data. It might add an extra step to data processing which is unfortunate but in the vast majority of cases won't be a problem."
Pretty much sums it up. This is a problem that has been blown out of proportions by media. Sure, amateur astronomy will take a slight hit, but it is nowhere the issue being presented.
Kessler syndrome doesn't apply at altitude so low and all of the satellites are brought by gravity in few months, they need constant readjustments to stay in low orbit.
We are talking about (at most - including other future companies) roughly 50,000 satellites in LEO, each about the size of a car, and they burn up in the atmosphere in a few years.
Given LEO is at a higher altitude, the "surface area" at that orbit is larger, but lets pretend its only 510 million km^2 (i.e. Earth's surface area).
If I evenly distributed 50,000 cars across 510 million km^2, you would need to search 10,200 km^2 to find 1 car.
Even 1 satellite per 1 km^2 in LEO would not impede astronomy, and that future requires 500+ million such satellites.
Amazon and SpaceX operating roughly 0.01 million satellites each is inconsequential.
Cons: this + Starlink + OneWeb make wide-field Earth-based astronomy much harder, making it less likely that we'll notice new astronomic phenomena or inbound city-killer near earth asteroids.
SpaceX is testing a sunshield on one of their newest Starlink satellites. I hope we'll see Kuiper launch with those and other reflectivity mitigations from the beginning instead of being surprised by the whole problem like SpaceX (claims to be).
Here's some coverage that we did in Orbital Index about the issue: https://orbitalindex.com/archive/2020-02-20-Issue-52/#starli...
It's a resource allocation efficiency issue, through and through. The government is always going to pick "winners and losers" in high value industries like defense, resource extraction, and aerospace but all of those investments have very rough feedback loops that are disconnected from the decision making process. If government investments/grants came at least with profit/revenue sharing arrangements, there would be a positive feedback loop and an opportunity for the system to evolve and improve even if the net income change after administration overhead is zero or even slightly negative.
Obviously this isn't tenable for basic scientific research but our industries do very little of that anyway, even with government funding available.
The target market is not elderly people retired to cabins in the woods. The antenna costs for individual users would be prohibitive (10s of thousands). SpaceX is aiming for military/government first, then fixed-point industrial (oil rigs, mining camps, large ships) and then aviation (internet onboard airliners). That's the priority structure.
Another angle is to look at the orbital inclinations to see which customers Musk is targeting. At the moment he isn't sending sats further north than London (+/-51*). That suggests their first/primary/target market doesn't exist north of London. It looks like they are going after either the military drone or luxury aviation consumers. I don't see communities in Canada's north getting connected anytime soon.
That is such a ridiculous claim.
Of course, we'll have to see how good these efforts of relative to normal broadband but they do offer hope that you won't have to choose where to live based on the broadband options.
Which I suppose does help, but only for people who probably already have 4G. So, not very rural.
But you're right. A lot of places that are pretty patchy or non-existent today for cell service, especially 4G, will likely have the same problem even after 5G rolls out.
I just looked at a property this week that's 15 minutes from a major regional hospital and has no access to reasonable wired internet.
I think a lot of people here assume you're either in Manhattan/SF or in some remote section of Wyoming. But, as you say, you don't need to get far away from dense civilization for broadband to be problematic.
I live about 40 miles outside of a major city off a somewhat main road. While I do get Comcast, Verizon FIOS isn't available at my address (in site of the fact that they constantly send me mailings).
Laying fibre and putting up 5G towers is cheaper, they can be serviced and upgraded.
Satellites make sense for governments, as it is harder to screw with them for adversaries. But for commercial use, what's exactly the benefit?
Given that this is Elon we're talking about, we'll call it 300 sats at $20 million per launch.
Per their FCC application, SpaceX were allowed an upper eventual limit of 30,000 sats. Let's assume that upper limit is what's required for a global comms network.
(30,000 / 300) * 20,000,000) = $2 Billion in launch costs. Assuming a lifetime of 5 years, that's $4 Billion per decade.
30,000 sats at $250K per gives us $7.5 Billion in hardware costs. Twice per decade is $15 Billion.
That's 19 Billion per decade. Round and double it, just to be safe. 40 Billion per decade, so 4 Billion per year.
Could you build and maintain a global fiber network for $4 Billion per year?
So even if LEO is more expensive, which it will not be, the expense will be covered by large financial firms.
Also companies like AWS pay a lot of money for bandwidth between their datacenters. Once Amazon's satellites are up and running, AWS will continue to need government permissions to build new datacenters (including ground stations) but will not need to interact with any other company for bandwidth. Improving AWS' profit margin for existing datacenters.
Finally, today AWS cannot build a datacenter in low cost areas because they need great internet connectivity. However, a LEO constellation enables them to build in significantly cheaper locations (think about the land/construction costs in Moldova vs Germany i.e. 6x reduction).
These alone are huge benefits that outweigh the costs, but adding general purpose internet traffic will only improve cost effectiveness.
 https://www.numbeo.com/property-investment/country_result.js... vs https://www.numbeo.com/property-investment/country_result.js...
The ability for people to take pictures being diminished is unfortunate, but if I had to choose between the two options I'll choose the one that significantly increases a lot on people's standard of living
Maybe in the future when we aren't ruled by capitalist overlords that make expanding traditional internet infeasible we can stop launching and replacing these satellites and can take pretty pictures again
Huge portions of the world are simply so sparsely populated that running wires is cost prohibitive.
And it's not a theoretical problem - there's plenty of people who have already had their work impacted
Software to remove the satellites would be great, given that the satellites are already interfering with astrophotography, can someone point me to the software to remove the starlink satellites from my images?
There are a few use cases/reasons to put telescope observatories on earth.
One reason is because it's cheap enough for hobbyists. An ordinary enthusiast can go out in their backyard and point a couple hundred dollars worth of gear at the sky and get shockingly gorgeous images with little more than a webcam and a cheap scope. Software like what I linked above can help remove satellites from those images. Post-processing of images is a (nearly) cost-free but complexity-expensive way to remove noise, distortion, and other artifacts from the images produced by cheap gear, might as well remove satellites while you're at it!
The second reason is because it's cheap enough for scientists and students. A few universities have their own observatories. They're abundantly aware that not only satellites but also air and light pollution get in the way of good seeing. There are all kinds of effort devoted to fixing the problems that air and light pollution cause, but the easiest fix is just to move the telescope. A university can contract time for its students with an observatory in a favorable location (low light pollution, clear skies, right latitude/hemisphere, high altitude, good seeing) and do science through that remote interface. These are, I think, the most important users to take care of with space-based telescopes. Maybe their Chilean observatory doesn't have the server hardware or network connection to do stacking and post-processing to remove satellites, and images taken by the old method of long exposure astrophotography are getting ruined by satellites. Time on a space-based telescope would, for many use-cases, be as good or better than time on a terrestrial observatory. And there are zero air or weather issues!
The third is because it's the only place where we can put telescopes of the required size for a particular imaging task. Maybe they're in an interferometric array, maybe they're 3+ meters in diameter, who knows. There are only a couple dozen telescopes in this category in the world, the few astronomers using these telescopes don't need HN commenters to point them to software to remove satellites!
I do acknowledge that processing for a hobbyist does increase the learning curve and reduce the raw awesomeness of the hobby a bit - "why not just go to nasa.gov and download someone else's preprocessed images" isn't the same as seeing something from your own backyard through nothing but a mirror and a piece of glass. I also acknowledge that there's a ton of money invested in terrestrial observatories, but everyone needs to keep up with the times.
Many of you will be too young to remember this but there was a glut of these companies/constellations in the 1990s. For example, Teledesic . For awhile 20 years ago we had Internet access of international flights. We have it again but there was 10-15 years in the middle where we didn't.
The reason was that there just wasn't that much demand for LEO satellite Internet and the capacity way exceeded demand.
We're at the beginning of another cycle like this. Not all these companies will survive. My money is on Starlink being one of the survivors because SpaceX, of any of these companies, is vertically integrated with the biggest cost: launches. Even better, their launch costs are the lowest.
Maybe Amazon can do this since Bezos also is throwing big money at Blue Origin but, at best, I think Blue Origin is years behind SpaceX.
I think you have to factor is the access cost though. There wasn't much demand at that cost.
I like the simplicity of the Starlink design, in that it's simply ground-to-satellite-to-ground. There are no satellite relays. I wonder if this will limit access in more remote places (eg Alaska, on the ocean, Antarctica). I guess SpaceX has decided this limitation is worth the cost-savings.
I'm curious to know what the effective range is from subscriber to access point (via satellite). And also what if you can talk to multiple Starlink satellites but only some of them are within range of an access point? I'm sure they've thought about this. I'm just curious on the technical details.
All in all, this time seems very different.
And be free of cox's monopoly price gouging as the sole provider of speeds over 5mb/s for my neighborhood, which is part of the second largest metropolitan area in the nation.
If cox and the rest had provided decent service and decent prices instead of gouging and underinvesting, they wouldn't be faced with competition from satellite clouds anywhere near as soon as they are going to be.
It's easy to make snark about this or that, but inexpensive, global, high bandwidth internet access seems like one of the most consequential technologies of the 21st century so far.
(Also, I wonder who will get the launch contract.)
I can't help but worry about where this ends though. 100,000 satellites orbiting the sky? Will our sky just be a bunch of moving dots, like you see in some sci-fi movies?
Most communication satellites (currently) have much higher altitudes, where such an event would be much worse as the debris would stay in orbit for much longer.
Boy, it's a great thing those satellites will only ever orbit above the US!
Boy, those astronomers are going to be so grateful for the extra bandwidth in their remote observatories!
I would very much like to live off of the Earth.
There is no more a theoretical situation that US can have a country locked our of software or hardware so larger countries would not want to have this dependency or have NSA/CIA intercept your satellites before launch and backdoor them(that is their job will be the excuse).
LEO constellations require lots of launches.
This is not to say it’s impossible, just that for it to be within an order of magnitude of the cost on SpaceX, someone else needs to come up with reusability as well.
I have no doubt that someone will, I just wonder about when, and if it will be in time to benefit anyone else’s presently announced LEO internet access constellations. It’s also possible that Amazon will just eat the cost of disposable rocket launches; it’s not like they couldn’t afford it if they wished.
Do you honestly see China,Russia or India governments giving money to a foreign company when they have their own space programs? The governments to not min-max for short term financials and you also have on top of that the security issue to trust US.
Satcom isn’t an issue for governments right now. It is, however, for billions of normal people who live far from population/commerce centers.
I'd guess we'd have a few people trying to shoot them down from their back yard, actually.
They will when economical access to LEO pays off in the form of cheap orbital telescopes.
this is a whole 'nother level of NIMBY.
Besides, at some point there will be a valid reason to have a visible device in orbit. I don't want any precedent set that some small number of narcissists can repeat e.g. the opposition to offshore wind power we've seen on Martha's Vineyard.
Does anyone here know if between this, Starlink, and OneWeb, it will ever become realistic to work remotely from a boat in, say, the middle of the South Pacific?
Thats really really expensive for not a lot of data. Click the wrong link from HN front page and you could burn through half that in one click!
I suppose it could be doable. You'd want to be setup with all your software and development environment before hand and just use the satellite connection for text-only email/chat. Maybe SSH to an outside server and use w3m for web browsing?
It seems like that would solve the light pollution problem. I have no relevant background, though.
James Webb: 6.5m (still on the ground)
More seriously though I think the race against time will be a real challenge.
There are few caveats to Amazon’s FCC approval. The company must launch half of the constellation by 2026 to retain its FCC license, and then the remaining satellites by 2029.
That is a lot of satellites to launch in 5 and half years which have to thread in with all the other scheduled launches and limited amount of launch sites to go from.
Is the consumer market for the larger antennas big enough to support the satellite costs?
I can also expect space photographers complaining even more going forward.
> Amazon claims that Kuiper will “provide broadband services to unserved and underserved consumers, businesses in the United States, and global customers by employing advanced satellite and earth station technologies,” according to the FCC’s approval document.
This is about getting internet to places where there is basically none at all now.
Listening to him, you can tell he's like a lot of other crony type Trump appointees; he seems to genuinely care about enhancing communication technologies in the US. I'm not saying I see eye to eye with him on all issues, but I think it's important for federal regulators to see the costs and benefits of policies, even if we fervently disagree with the final conclusion.
Link to the interview:
> you can tell he's not like a lot of other crony type Trump appointees
Ultimately it's IP; it's up to you what to do with it.
Also, protocols alone don't grant users addresses that they own and can use publicly and directly P2P, which is how we end up with all these sign-ins to owners of data centers.
Is this really a future we want? Or Will there be room for establishing dark satellite terrains?
Vertical, not horizontal monopoly.