Amazon Project Kuiper Broadband Constellation

[butters]

Amazon formally announces space business solutions business unit...

https://blog.aboutamazon.com/innovation/amazon-web-services-unveils-new-space-business-segment


Your one stop end-to-end shop for your platform needs...

So notice how it is under the AWS division, I think that is a strong indicator that Kuiper's biggest client will be Amazon/AWS itself, and will mostly make money off of new AWS subscriptions. This is a pretty big leg up on the other constellations, as it is business/enterprise oriented and not consumer facing primarily and therefore has a must stable initial revenue service. Plus it makes AWS much more competitive for future Government contracts.

Between this, backing Rivian and his rocket program, I'm starting to wonder about Bezos seeming to chase everything Musk industries is doing. I'm expecting a tunnel boring company next.

There's also Zoox, the self driving company Amazon just purchased for $1.3 billion.

I actually fully expect a Boring play from Bezos. If Boring tunnels work out, Amazon is going to want a similar way to move goods rapidly. I doubt Bezos asks Musk to build him routes.

Amazon tends to invest in technology that help its core business.ZooX is for self drive delivery vehicles for delivering Amazon goods. AWS was for internal use that found external market, Prime internet TV used AWS servers. Kuiper is to support AWS and its customers.

Amazon annual revenue is $280B why would Bezos try to copy Musk whose business are 1/10 size.

It's possible that Gen-2 Starlink with space-to-space laser links could be used to connect cloud datacenters (i.e. AWS or more likely GCP) for geographic data replication and other cloud services, but if we're to understand that Kuiper is a pure space-to-ground constellation, then I don't see how it would play a useful role in AWS infrastructure.

Amazon won't be able to build an AWS datacenter without terrestrial fiber links by mounting a Kuiper terminal on the roof. They wouldn't be able to offer the cross-region data replication performance that AWS customers expect. The closest thing to an AWS use-case that I can envision is a TDRS-like capability using Kuiper to enhance their AWS Ground Station service.

The larger rationale for Amazon's investment in Kuiper is expanding access to their e-commerce platform. This should be pretty clear. The business play is not about AWS, it's about introducing more of the Earth's inhabitants to the convenience of The Everything Store. They can't grow their customer base if billions don't have reliable Internet access.

[Mammutti]

https://twitter.com/thesheetztweetz/status/1288954064760377344

The FCC authorizes Amazon to build its Kuiper internet constellation of 3,236 satellites, which would likely compete with SpaceX's Starlink in the realm of low Earth orbit broadband service: BREAKING

https://twitter.com/CHenry_SN/status/1288951906409340929

Amazon's Kuiper constellation received FCC approval today. The commission is requiring Amazon to submit an updated debris mitigation plan once its satellite design is finalized. https://spacenews.com/amazons-kuiper-constellation-gets-fcc-approval/

[gongora]

So Kuiper is officially in the 2020 processing round, and is required to coordinate with/not interfere with constellations approved earlier.  We'll see what that means in practice.

a) Prior to commencing operations in the 17.8-18.6 GHz and 18.8-20.2 GHz and 27.5-30 GHz bands, Kuiper must certify that it has completed a coordination agreement with or make a showing that it will not cause harmful interference to any operational system licensed or granted U.S. market access in the NGSO FSS processing rounds referred to in Public Notices DA 16-804, 31 FCC Rcd 7666 (IB 2016) and DA 17-524, 32 FCC Rcd 4180 (IB 2017).
...
b. Kuiper must launch the space stations, place them in the assigned orbits, and operate them in accordance with this authorization and 47 CFR § 25.164(b). Section 25.164(b) requires Kuiper to launch and operate 50 percent of its satellites no later than July 30, 2026, and Kuiper must launch the remaining space stations necessary to complete its authorized service constellation, place them in their assigned orbits, and operate each of them in accordance with the authorization no later than July 30, 2029. 47 CFR § 25.164(b).

[gongora]

Amazon receives FCC approval for Project Kuiper satellite constellation
By Day One Staff
on July 30, 2020

Last spring, we announced Project Kuiper, an initiative to build a low earth orbit (LEO) satellite constellation capable of providing reliable, affordable broadband service to unserved and underserved communities around the world.

Today marked a key milestone for the project, with the Federal Communications Commission (FCC) granting Amazon approval by a 5-0 vote to deploy and operate our constellation of 3,236 satellites. The authorization allows Project Kuiper to deliver satellite-based broadband services in the United States, helping expand internet access to households and communities across the country.

A project of this scale requires significant effort and resources, and, due to the nature of LEO constellations, it is not the kind of initiative that can start small. You have to commit. Amazon will invest more than $10 billion in Project Kuiper. This investment will create jobs and infrastructure around the United States, build and scale our ground network, accelerate satellite testing and manufacturing, and let us deliver an affordable customer terminal that will make fast, reliable broadband accessible to communities around the world.

“We have heard so many stories lately about people who are unable to do their job or complete schoolwork because they don’t have reliable internet at home,” said Dave Limp, Senior Vice President, Amazon. “There are still too many places where broadband access is unreliable or where it doesn’t exist at all. Kuiper will change that. Our $10 billion investment will create jobs and infrastructure around the United States that will help us close this gap. We appreciate the FCC's unanimous, bipartisan support on this issue, and I want to thank Chairman Pai and the rest of the Commission for taking this important first step with us. We’re off to the races.”

Project Kuiper will deliver high-speed, low-latency broadband service to places beyond the reach of traditional fiber or wireless networks. It is inspired by customers in every corner of the world: by families working and learning together from home; by scientists and researchers operating in remote locations; by first responders providing disaster relief; and by companies of all sizes moving their business online. Project Kuiper will serve individual households, as well as schools, hospitals, businesses and other organizations operating in places without reliable broadband.

“We are doing an incredible amount of invention to deliver fast, reliable broadband at a price that makes sense for customers,” said Rajeev Badyal, Vice President of Technology for Project Kuiper. “LEO-based broadband systems like Project Kuiper present a huge number of challenges, and we have assembled a world-class team of engineers and scientists who are committed to delivering on our vision for Project Kuiper and keeping space a safe, sustainable environment for everyone. Combine that with Amazon’s deep expertise in networking and infrastructure and its ability to finance such a huge undertaking, and I am optimistic about the impact we can have for these unserved and underserved communities.”

We are also committed to working with public and private sector partners that share our vision for the project. In addition to providing ground station service directly to customers, Project Kuiper will also provide backhaul solutions for wireless carriers extending LTE and 5G service to new regions. Together, these projects will expand broadband access to more households in the United States and around the world.

Project Kuiper will be designed and tested in our all-new research and development facility opening in Redmond, Wash. The Kuiper team is growing quickly, and those inspired by our mission and interested in joining a diverse, world-class team can view open roles here.

[Robotbeat]

That's exciting. It's something with the same scale and ambition as Starlink while also (almost certainly) leveraging reusable rockets like New Glenn.

Now we just need Blue Origin to start really executing on New Glenn (etc) and we'll have like 2 SpaceXes.

Competition is good!

[ncb1397]

Is this the facility that will be home to Project Kuiper?  The fountain statue is interesting. Looks like satellites orbiting earth.

https://www.google.com/maps/@47.6700326,-122.1188277,3a,75.4y,44.17h,78.92t/data=!3m6!1e1!3m4!1svHJnnHHkdwHQjqo35CCJNg!2e0!7i16384!8i8192

[TrevorMonty]

This is one LEO constellation where financing isn't an issue. While launch by Blue isn't given, will be surprised if they aren't first if not sole launch provider.

Even if SS launch is cheaper may not be in Amazon's interest to give money to competing constellation company.

[ChrisWilson68]

Now we just need Blue Origin to start really executing on New Glenn (etc) and we'll have like 2 SpaceXes.

I really hope they do start executing.  I do think we'll finally see New Glenn within the next 2-4 years.

5 years ago I would have been excited to see New Glenn and I'd say there's another SpaceX.  But SpaceX has moved on.  I'd really love to see Blue Origin have a real Starship competitor.  Maybe they'll get there, but unfortunately I think it will take a while.  So I don't think we'll have 2 SpaceXes for at least the next decade or so.

But still, one SpaceX and one New Glenn is better than just one SpaceX and one ULA.

[edzieba]

I'm surprised by the prevalence of the "Amazon want Kuiper for DC to DC links" argument:
Firstly, because they already have those links at much higher bandwidths and lower cost than via a satellite network: it's called fibre optics. Google, Amazon, etc not only own their own fibre links between fixed datacentres (not just hooking to an upper layer ISP, actually run and own their own point-to-point links), they own their own trans-oceanic fibre optic links between countries too! Even if you posit a surprise rollout of brand new datacentres located in the middles of several nowheres, it still remains cheaper to run fibre to them rather than launch a megaconstellation. At absolute best you could use satellite links to bridge the gap between building and comissioning your datacentre and running fibre to it in the event you build a datacentre and forget to run fibre until you're finished.
Secondly, security. That argument is also bunk. You encrypt before transmission over any link that leaves your physical building (and ideally any link that leaves any given box), and that encryption provides 100% of the protection against snooping or tampering, regardless of the physical medium those packets are carried over (inside an optical fibre, or written on a postcard strapped to a carrier pidgeon). Physical ownership of the PHY layer means naff-all in terms of security, as does the geographic route that data passes through. Legislative snooping affects satellite data just as much as data over microwave links, fibre, or POTS: if the local government tells you "let us snoop your traffic or you don't get to operate", they don't care if your backhaul is satellite or fibre. Even the active malice argument isn't a great one: If someone can locate and cut multiple fibre links, they can locate and whack-with-a-hammer your dish, or just squirt RF noise at the sufficient-but-low power needed to jam it.

Amazon are gunning for the same market as Starlink and Oneweb: Fixed terminals in rural areas (below the density where running fibre/copper wins economically, as it does in suburban and urban areas), and high-bandwidth mobile terminals outside of cell-site coverage.
You need to tick the "we physically can't run fibre there / we can't afford to run fibre there" box before 'satellite network' becomes a reasonable choice. Datacentres do not tick that box.

[Robotbeat]

I think Amazon is pretty ambitious about their constellation, though. I suspect similarly ambitious to Starlink.

Fiber is really cheap for moving data around between data centers; no argument there. But ultimately, I think people underestimate the overhead in the last-mile for consumer data. Megaconstellations, with reusable rockets and continual improvement, can be a bigger competitor (even outside of rural markets) than you might think.

Average monthly usage for Comcast (Xfinity) is 308GB per customer. That's less than 1 Megabit/s average speed. Each Starlink launch is roughly 1 terabit/s of capacity. Each Falcon 9 launch can service about a million customers on average... although there's overhead for multiple reasons that probably reduces that by a factor of 10. With $80/month per customer (Comcast's average monthly revenue per user is $150-160, BTW) and a satellite lifetime of about 5-6 years, that's about $500 million in revenue per launch at Comcast-like data usages (obviously user terminals and overhead has to come out of that, too, but I'm illustrating a point, here...).

New Glenn is at least twice the capacity to LEO, and eventually it (or the follow-on launcher) will be reusable. I think there will be a lot more competition with the Comcasts and Verizons of the world than one might think. Particularly as a lot of city users who don't use a lot of bandwidth might migrate to Kuiper or Starlink in the near term, forcing Comcast and Verizon to have a higher average per-customer bandwidth usage and lower overall bargaining power...

[Asteroza]

I'm surprised by the prevalence of the "Amazon want Kuiper for DC to DC links" argument:
Firstly, because they already have those links at much higher bandwidths and lower cost than via a satellite network: it's called fibre optics. Google, Amazon, etc not only own their own fibre links between fixed datacentres (not just hooking to an upper layer ISP, actually run and own their own point-to-point links), they own their own trans-oceanic fibre optic links between countries too! Even if you posit a surprise rollout of brand new datacentres located in the middles of several nowheres, it still remains cheaper to run fibre to them rather than launch a megaconstellation. At absolute best you could use satellite links to bridge the gap between building and comissioning your datacentre and running fibre to it in the event you build a datacentre and forget to run fibre until you're finished.
Secondly, security. That argument is also bunk. You encrypt before transmission over any link that leaves your physical building (and ideally any link that leaves any given box), and that encryption provides 100% of the protection against snooping or tampering, regardless of the physical medium those packets are carried over (inside an optical fibre, or written on a postcard strapped to a carrier pidgeon). Physical ownership of the PHY layer means naff-all in terms of security, as does the geographic route that data passes through. Legislative snooping affects satellite data just as much as data over microwave links, fibre, or POTS: if the local government tells you "let us snoop your traffic or you don't get to operate", they don't care if your backhaul is satellite or fibre. Even the active malice argument isn't a great one: If someone can locate and cut multiple fibre links, they can locate and whack-with-a-hammer your dish, or just squirt RF noise at the sufficient-but-low power needed to jam it.

Amazon are gunning for the same market as Starlink and Oneweb: Fixed terminals in rural areas (below the density where running fibre/copper wins economically, as it does in suburban and urban areas), and high-bandwidth mobile terminals outside of cell-site coverage.
You need to tick the "we physically can't run fibre there / we can't afford to run fibre there" box before 'satellite network' becomes a reasonable choice. Datacentres do not tick that box.

For bulk data, terrestrial fiber is hard to beat (except with a station wagon full of tapes), but for low latency links there may be a market. I could imagine an AWS inter-region/interAZ low latency tier for their internal network access, optimized for certain syncing stuff. Depending on size, DB replication, as well as some special case BCP-DR syncing might be willing to pay for that.

AWS is rich enough to not need to specifically go after the FCC rural broadband initiative subsidies, but may be interested as a potential anchor "customer" perhaps.

There's always the specter of Amazon trying to expand their Kindle whispernet MVNO cellular service however as part of their information hegemony endgame...

[FutureSpaceTourist]

https://twitter.com/thesheetztweetz/status/1295359326169726978

Amazon is looking for engineers to develop the customer terminals for its coming Project Kuiper satellite internet service:
https://www.amazon.jobs/en/jobs/1237208/senior-customer-terminal-architect
https://www.amazon.jobs/en/jobs/1237380/senior-customer-terminal-network-architect

[TrevorMonty]

This article view is same as mine, in that Kuiper is extension of AWS. The odd rural customer is nice extra revenue stream but not essential to business plan.

https://spectrum.ieee.org/tech-talk/aerospace/satellites/amazons-project-kuiper-is-more-than-the-companys-response-to-spacex

“With Amazon, it’s a whole different ballgame,” says Zac Manchester, an assistant professor of aeronautics and astronautics at Stanford University. “The thing that makes Amazon different from SpaceX and OneWeb is they have so much other stuff going for them.” If Kuiper succeeds, Amazon can not only offer global satellite broadband access—it can include that access as part of its Amazon Web Services (AWS), which already offers resources for cloud computing, machine learning, data analytics, and more.

[meekGee]

This article view is same as mine, in that Kuiper is extension of AWS. The odd rural customer is nice extra revenue stream but not essential to business plan.

https://spectrum.ieee.org/tech-talk/aerospace/satellites/amazons-project-kuiper-is-more-than-the-companys-response-to-spacex

“With Amazon, it’s a whole different ballgame,” says Zac Manchester, an assistant professor of aeronautics and astronautics at Stanford University. “The thing that makes Amazon different from SpaceX and OneWeb is they have so much other stuff going for them.” If Kuiper succeeds, Amazon can not only offer global satellite broadband access—it can include that access as part of its Amazon Web Services (AWS), which already offers resources for cloud computing, machine learning, data analytics, and more.

I'd go as far as saying that without Kuiper, AWS stands at risk in comparison to a StarLink/Microsoft or StarLink/Google partnership.

They simply cannot afford to have a constellation gap!

[Asteroza]

They simply cannot afford to have a constellation gap!

If someone could photoshop in Bezos onto the character General Buck Turgidson from Dr. Stranglove, I think we'll be satisfied...

[FutureSpaceTourist]

https://twitter.com/thesheetztweetz/status/1299032138440994818

A visualization of Amazon's Project Kuiper satellite internet constellation, by @NSR_SatCom:

$AMZN

[gongora]

https://twitter.com/thesheetztweetz/status/1299032138440994818

A visualization of Amazon's Project Kuiper satellite internet constellation, by @NSR_SatCom:

$AMZN

That's only one inclination.  There are other shells at 51.9 and 33 degrees.

[FutureSpaceTourist]

https://twitter.com/thesheetztweetz/status/1303822634560491521

Amazon $AMZN has brought on Peter Marquez as the company's new Head of Space Policy. He's previously worked on the White House National Security Council as the Director of Space Policy, as well as a VP for both Orbital Sciences and Planetary Resources.

https://www.linkedin.com/posts/peter-marquez-55756740_im-happy-to-have-joined-the-amazon-family-activity-6709558665323577344-DEpP

[gongora]

[CNBC] Amazon streamed 4K video from a satellite with a prototype antenna for its Kuiper internet network

Amazon testing Ka-band phased array antenna that will be used for Kuiper.

[gongora]

[Amazon] Custom-built antenna architecture will allow Amazon to deliver a small, affordable customer terminal to connect unserved and underserved communities around the world.

[Amazon] Nima Mahanfar discusses the science behind Project Kuiper customer terminal antenna