FCC lifts looming deadline for Amazon Leo satellite broadband constellation

The waiver "serves the public interest by promoting a second large satellite broadband constellation."

See full article...

 

[S_T_R]

LEO Satellite data is a natural non-monopoly

You get world wide coverage with the most basic system. And there are multiple vendors right now.

There is only one mega constellation currently in service, but we are discussing one of the next one. The Chinese are launching as well.

Yeah, it's not a monopoly, but it's not necessarily going to be super competitive either.

1. Spectrum is a limitied resource.
2. Feasible orbits are limited. This is compounded by #1. You only have so much spectrum, so each satellite has limited bandwidth available to it. Too low and the satellite falls out of orbit. Too high and each satellites service area is too large. Even with beamforming, you can run into interference issues if the birds are too high.
3. Massive barriers to entry. You have regulatory hurdles on both the spectrum and spaceflight side. Then there's the staggering capital cost.

On the plus side, terrestrial telecoms are a competitor, which places a soft cap on what space-based providers can reasonably charge. All in all, it's not a traditional utility that requires monopoly regulation. It's, like cellular telecoms, a natural oligopoly.

The big caution is consolidation. If SpaceX uses its bloated valuation to buy a terrestrial telecom, that would be bad for consumers.

 

[Dan Corcoran]

I predicted this when Carr blasted Amazon for complaining about SpaceX. Carr must have known at the time he was goign to give Amazon a pass.
https://meincmagazine.com/space/2026/...after-it-criticizes-spacex-megaconstellation/

 

[ArcaneTourist]

LEO Satellite data is a natural non-monopoly [...]

Space is big, but this space.com op-ed suggests that trying to determine the amount of sats that will fit in LEO by looking at volume is like trying to estimate highway capacity by looking at how many cars could be parked on the highway.

The article suggests that the limit might be on the order of hundred(s) of thousands of satellites.

Personally, it seems like it depends greatly on advances in traffic control.

Starlink has about 10K sats currently, I think - but SpaceX has talked about going to 100K sats.

EDIT: This is per orbital shell. I'm not sure how many shells there are that are suitable for Internet constellations.

 

[Malmesbury]

Yeah, it's not a monopoly, but it's not necessarily going to be super competitive either.

1. Spectrum is a limitied resource.
2. Feasible orbits are limited. This is compounded by #1. You only have so much spectrum, so each satellite has limited bandwidth available to it. Too low and the satellite falls out of orbit. Too high and each satellites service area is too large. Even with beamforming, you can run into interference issues if the birds are too high.
3. Massive barriers to entry. You have regulatory hurdles on both the spectrum and spaceflight side. Then there's the staggering capital cost.

On the plus side, terrestrial telecoms are a competitor, which places a soft cap on what space-based providers can reasonably charge. All in all, it's not a traditional utility that requires monopoly regulation. It's, like cellular telecoms, a natural oligopoly.

The big caution is consolidation. If SpaceX uses its bloated valuation to buy a terrestrial telecom, that would be bad for consumers.

The ITC (the international body dealing with spectrum) seems to think that there is enough for multiple constellations - more than are planned.

The regulatory barriers have allowed through 2 American mega constellations and several Chinese ones (planned).

The cost barrier is of the order of $10 billion - which is not exactly insane given the scale of the resulting ISP.

 

[Tomokatsu]

They entered into a market (being an isp) that gives them no benefits,other than to be a major customer of their former CEO and large shareholder. Being an internet service provider doesn't help them, they have no real expertise, and it is a massive capital outlay...for what?

I know people point to aws, but this has nothing to do with aws. This is more akin to them competing with comcast. Take bezos rocket company out of the picture, and Id bet great money this never even gets considered.

I get your point, but I do think you might be overly fixated on consumer services. It remains to be seen, but there is a material commercial connectivity business where AWS might have synergies. For example, think about aviation, or something like off-shore oil & gas industries.

It also omits the idea of utilizing their "Prime" brand - which has expanded into all sorts of other adjacencies.

All this to say - I don't personally love the idea of Amazon trying to bundle internet in with all of their other services, but I don't think anything about this screams grift.

 

[ArcaneTourist]

Apparently no one at the FCC has heard of the Kessler Effect. If you have seen the movie Gravity you know what I mean. With the current number of LEO satellites there is a significant chance of chain reaction collisions during the next severe solar storm event. Adding thousands more LEO satellites increases the odds to near certainty. Something the FCC should have considered and something to consider if you are thinking about jumping onto the SpaceX IPO

When people mention Kessler, they're usually talking about an event that would render space unavailable for generations.

The orbit used, by example, for Starlink satellites is so low that anything in that orbit will enter the atmosphere in a few years. Hence, a Kessler event at that altitude is not possible.

EDIT: Solar storms tend to make low flying sats enter the atmosphere and burn up sooner.
EDIT: The movie "Gravity" was fun, but the orbital mechanics were fantasy; several video games do a better job.

 

[Cloudgazer]

The article doesn't say much about the cost to Amazon for the different rockets.



A quick search says the F9 lists at ~ $74 million and NG is estimated at $68 to $110 million. But, Amazon's satellite project has a contract for a bulk discount on NG.

So, if it were flying, New Glenn would be the cheaper way for Amazon to get its sats to orbit. If it were flying...

The reason they get so few birds on F9 isn't the weight - it's the volume, had the designed with F9 in mind they'd get significantly more.

It's a little bit suspicious that they seemingly designed their birds to be dependent on extra big fairings and thus on the launcher created by their founder, especially when SpaceX had by this point demonstrated the incredible power of their flatpack deployments.

 

[r0twhylr]

Amazon did it to themselves by not including SpaceX in the initial big order. It took an investor lawsuit to uncover the detail that Amazon's board had spent less than a hour debating the launch contracts and had never even considered using Falcon 9. Thanks to the lawsuit, as well as delays with the other rockets, the company was forced to use Falcon, grudgingly.

True, but they also screwed themselves by not having satellites ready in bulk early enough. By the time their first prototypes flew, they were already running way behind the 8 ball for this FCC deployment milestone. And then after that, it took a long time to ramp up production to the point they have satellites to deploy. They had Atlas rockets sitting in storage for years waiting for satellites. Having Falcons standing by at that point wouldn't have made much difference if any.

The problems with Vulcan and NG are certainly affecting their ability to deploy their constellation, but even if they were flying, Amazon would need this FCC ruling.

 

[sigan7]

The reason they get so few birds on F9 isn't the weight - it's the volume, had the designed with F9 in mind they'd get significantly more.

At approximately ~571 kg per satellite (including dispenser¹), and a Starlink-esque total payload of 17,500 kg, that's still only 30 satellites instead of 24. But then you need to take into account that the Starlink deployment orbit is much lower than the Amazon deployment orbit of 465 km, which would reduce that number. Ultimately, no, it's not a significant difference and wouldn't change the equation.

¹ You might say that the dispenser could be optimized in a flatpack configuration. If that was the case, then it could also be optimized for New Glenn, which still has tons of room in its fairing and is likely mass-constrained at present.

 

[Baenwort]

I work in the comms sector - personally, I'd love to break the monopoly and drag the US into entertaining more municipal fiber. That said, it simply doesn't cover all use cases (not saying we shouldn't do it - just saying you still need both).

Fiber economics simply won't make sense everywhere, even terrestrial FWA is limited. Not to mention the defense applications, maritime, air, and ultra-long-haul that LEO satellites can offer.

Anywhere power lines can go, so can fiber.

Anything else is admitting we can't match the work of a hundred years ago and the rural electrification efforts.

Fiber is also cheaper than roads and we have those a lot of places fiber doesn't go.

 

[Cloudgazer]

At approximately ~571 kg per satellite (including dispenser¹), and a Starlink-esque total payload of 17,500 kg, that's still only 30 satellites instead of 24. But then you need to take into account that the Starlink deployment orbit is much lower than the Amazon deployment orbit of 465 km, which would reduce that number. Ultimately, no, it's not a significant difference and wouldn't change the equation.

Exactly - including dispenser. My point precisely. Starlink avoids the need for a heavy dispenser. Why not go the same route? Well because then NG wouldn't look nearly so appealing.

¹ You might say that the dispenser could be optimized in a flatpack configuration. If that was the case, then it could also be optimized for New Glenn, which still has tons of room in its fairing and is likely mass-constrained at present.

NG was supposed to have 45 tonnes to LEO, if it is mass limited on Leo launches then that would not be something that Amazon employees seeking to please Bezos would have anticipated or designed for. It is also not something BO have publicly admitted.

 

[NetMage]

Being an internet service provider doesn't help them, they have no real expertise, and it is a massive capital outlay...for what?

Actually it is a considerable help to them, and like Google’s various rural Internet attempts, its point is to expand rural Internet presence so their customer base expands, and rural customers would be an ideal target base for them, as they will typically not have local or convenient alternatives to purchasing from Amazon for delivery.

 

[NetMage]

Apparently no one at the FCC has heard of the Kessler Effect.

Stop trying to make the Kessler Effect a thing. It was not a thing, it will never be a thing. Have you read how long it is supposed to take to occur? Think decades.

If you have seen the movie Gravity you know what I mean.

That’s like saying if you’ve seen the movie Herbie: Fully Loaded you know how Nascar racing works.

 

[Yui]

They had Atlas rockets sitting in storage for years waiting for satellites. Having Falcons standing by at that point wouldn't have made much difference if any.

That's a great point. I'd forgotten about the Atlas rockets being made well in advance because of ULA closing the production line to switch over to Vulcan. Amazon had rockets available to them but no satellites to launch.

 

[SpikeTheHobbitMage]

Anywhere power lines can go, so can fiber.

Anything else is admitting we can't match the work of a hundred years ago and the rural electrification efforts.

Fiber is also cheaper than roads and we have those a lot of places fiber doesn't go.

Rural electrification included setting up local generation for communities that were too remote to connect to the main grids. It was also heavily subsidized. There is a point where it is cheaper and more efficient to set up a WISP than to run fiber. Beyond that is a point where satellite becomes the optimal solution. Starlink moves the bar but the base equation doesn't change.

 

[Secondfloor]

Apparently no one at the FCC has heard of the Kessler Effect. If you have seen the movie Gravity you know what I mean.

Also, apparently, neither have you.

You know Gravity is fiction, right?

But congratulations on being the first in this thread (there’s always multiple) to comment without actually reading and understanding his work, nor comprehending why it cannot occur at the altitudes these satellites operate at.

And to top it all off, that doesn’t fall within the FCC’s authority.

Whats particularly amazing is this is the fifth time you have brought it up in the last six months and apparently have chosen to remain intentionally ignorant on the subject, even after being corrected.

 

[A Man A Plan A Canal Panama]

Lost or injured hikers will never have access to municipal fiber.

That's what Iridium is for

 

[Hamlet_Jr]

It really seems like people are picking sides based on who owns what company.

What is better for America, Amazon getting fucked over and all the launch providers losing 100 launches, or SpaceX losing its monopoly and having to engage in competition? All because of an arbitrary date on the calendar?

It isn’t a zero sum situation. In this particular case, a rising tide lifted all boats.

SpaceX are the guys who open source their tech. They have done a lot to foster the entire space economy.

 

[Cloudgazer]

It really seems like people are picking sides based on who owns what company.

What is better for America, Amazon getting fucked over and all the launch providers losing 100 launches, or SpaceX losing its monopoly and having to engage in competition? All because of an arbitrary date on the calendar?

It isn’t a zero sum situation. In this particular case, a rising tide lifted all boats.

SpaceX are the guys who open source their tech. They have done a lot to foster the entire space economy.

Blue Origin and Amazon have played the ref ... a lot.

There's a certain amount of schadenfreude about seeing them having to scrabble to avoid a red card.

 

[r0twhylr]

Anywhere power lines can go, so can fiber.

Anything else is admitting we can't match the work of a hundred years ago and the rural electrification efforts.

Fiber is also cheaper than roads and we have those a lot of places fiber doesn't go.

Just because fiber can go somewhere doesn't make it the best choice.

I work with companies who rely on different WAN connections and SD-WAN for reliabililty. If they have 2 x fiber connections taken out by a backhoe, they're screwed. Some kind of wireless - satellite, microwave, or 5G - is really best practice.

 

[r0twhylr]

That's what Iridium is for

It was always rare and expensive. Direct to cell coverage is orders of magnitude better.

 

[SportivoA]

Stop trying to make the Kessler Effect a thing. It was not a thing, it will never be a thing. Have you read how long it is supposed to take to occur? Think decades.

Eh, if the Chinese build two constellations in high-LEO to MEO and keep leaving the rocket bodies up there with the comms birds, then they might end up giving it a try. 100-1000+ year debris lifetimes (as they descend) is at least the opening salvo for such a disaster. If Amazon and SpaceX keep theirs in the active-stationkeeping-required portion of the upper atmosphere, they probably won't start the fire, as it were.