Author Topic: Amazon Project Kuiper Broadband Constellation  (Read 194994 times)

Offline su27k

  • Senior Member
  • *****
  • Posts: 6414
  • Liked: 9104
  • Likes Given: 885
Re: Amazon Project Kuiper Broadband Constellation
« Reply #280 on: 11/03/2022 02:48 am »
https://twitter.com/LionnetPierre/status/1586709356661379072

Quote
Amazon wants to target "tens of millions of customers", and will focus on lowering the user terminal price. The "bill of materials" alone is under 400$ he says. Admittedly the Kuiper Ka band satelllite & terminal are more complex than Starlink's Ku one. 2/

I don't think it was known before that the "under $500" figure for Kuiper user terminal is actually the BOM cost. If I'm not mistaken the price of a consumer electronics would be a lot higher than the BOM cost, usually 3x to 5x more, so Kuiper's user terminal doesn't look that cheap does it?

Offline edzieba

  • Virtual Realist
  • Senior Member
  • *****
  • Posts: 6494
  • United Kingdom
  • Liked: 9936
  • Likes Given: 43
Re: Amazon Project Kuiper Broadband Constellation
« Reply #281 on: 11/03/2022 08:18 am »
https://twitter.com/LionnetPierre/status/1586709356661379072

Quote
Amazon wants to target "tens of millions of customers", and will focus on lowering the user terminal price. The "bill of materials" alone is under 400$ he says. Admittedly the Kuiper Ka band satelllite & terminal are more complex than Starlink's Ku one. 2/

I don't think it was known before that the "under $500" figure for Kuiper user terminal is actually the BOM cost. If I'm not mistaken the price of a consumer electronics would be a lot higher than the BOM cost, usually 3x to 5x more, so Kuiper's user terminal doesn't look that cheap does it?
Depends on whether they sell terminals at a profit in additional to charging for services, the model many telecomms companies (and SpaceX) use of subsidising the terminal and making up that cost through service charges, or the model other telecomms companies use of leasing the terminal as part of the service charge (with no upfront cost).

Offline Tywin

Re: Amazon Project Kuiper Broadband Constellation
« Reply #282 on: 11/07/2022 08:26 pm »
speculation, maybe because of the delay of the Ariane 6, Amazon purchase a contract with H3 rocket?
The knowledge is power...Everything is connected...
The Turtle continues at a steady pace ...

Offline TrevorMonty

Re: Amazon Project Kuiper Broadband Constellation
« Reply #283 on: 11/30/2022 07:11 pm »
Interview with Clint Crosier from AWS and services they offer to space businesses eg satellite operators.



AWS now pushing their APIs to customers satellites, so satellite can run latest and greatest APIs applicable to operators  needs. Examples were processing of images by satellite to discard images with clouds, another example to monitoring ship movement. Satellite would detect when ship moves and download images once it starts leaving port, uptil then images are discarded.

I can see Kuiper satellites providing AWS services to satellites where images etc are transmitted to Kuiper satellite and it does processing. They never discussed this but suspect that is where things are going.

Sent from my SM-T733 using Tapatalk


Offline Asteroza

  • Senior Member
  • *****
  • Posts: 2910
  • Liked: 1126
  • Likes Given: 33
Re: Amazon Project Kuiper Broadband Constellation
« Reply #284 on: 12/01/2022 12:18 am »
Interview with Clint Crosier from AWS and services they offer to space businesses eg satellite operators.



AWS now pushing their APIs to customers satellites, so satellite can run latest and greatest APIs applicable to operators  needs. Examples were processing of images by satellite to discard images with clouds, another example to monitoring ship movement. Satellite would detect when ship moves and download images once it starts leaving port, uptil then images are discarded.

I can see Kuiper satellites providing AWS services to satellites where images etc are transmitted to Kuiper satellite and it does processing. They never discussed this but suspect that is where things are going.

Sent from my SM-T733 using Tapatalk

So called edge computing advances are laying the groundwork for this, along with some pioneering work on software defined satellites where customer software containers are pushed to isolated application CPU's onboard for payload data processing. AWS has extensive operational experience with running customer software containers, as well as custom chips (partially optimized for containers), so this is well within their wheelhouse. With the application processor isolated from housekeeping systems, customers are free to upload new containers at will without extensive review by the satellite operator.

The trouble though is the model. Running a CPU poor but data rich sat, and offloading data to Kuiper, will mean more extensive communication capabilities for ISL's, which gets into questions regarding RF vs laser. We might see the reverse though, in the satellite form of AWS Outpost (a service where AWS deploys managed AWS racks within your own on-premises server rooms, where you get priority use of racks, but AWS gets to use idle capacity for themselves). An example being a relatively self-contained 3U compute module you slap onto your sat zenith side that connects to onboard systems via Spacewire, running preprocessing containers to shave down data before pushing to kuiper sats running more intensive container workloads relative to the data stream. Though that presupposes the RaspberryPi supercomputer hypothesis, where a swarm of low-to-mid capability compute can constitute a useful federated resource.

That recent AMD announcement for a space rated high performance SoC with FPGA would be a good fit for such an application processor in an addon module that can be "bought" off the shelf from AWS.
« Last Edit: 12/01/2022 12:20 am by Asteroza »

Offline JayWee

  • Full Member
  • ****
  • Posts: 1021
  • Liked: 1033
  • Likes Given: 2044
Re: Amazon Project Kuiper Broadband Constellation
« Reply #285 on: 12/01/2022 03:30 am »
Examples were processing of images by satellite to discard images with clouds
If I use Kuiper as a relay, I already built an expensive satellite -> why would I let the feeble Kuiper handle it?
OR
Quote
another example to monitoring ship movement. Satellite would detect when ship moves and download images once it starts leaving port, uptil then images are discarded.
Is Kuiper planning to put optical sensors on the sats? Then it'd make sense to have pre-filtering on the sats themselves with customers uploading various filters.

But - letting customers run an instance of their SW on all the sats in the megaconstellation (otherwise it doesn't really make sense for global coverage) - feels just weird.

Offline su27k

  • Senior Member
  • *****
  • Posts: 6414
  • Liked: 9104
  • Likes Given: 885
Re: Amazon Project Kuiper Broadband Constellation
« Reply #286 on: 01/04/2023 10:11 am »
Amazon financials doesn't look great these days, I wonder if this will affect Kuiper:

Amazon secures $8 billion term loan

Quote from: reuters.com
"Given the uncertain macroeconomic environment, over the last few months we have used different financing options to support capital expenditures, debt repayments, acquisitions, and working capital needs," an Amazon spokesperson told Reuters in a statement.

<snip>

Amazon had about $35 billion in cash and cash equivalents and long-term debt of about $59 billion at the end of the third quarter ended Sept. 30.

Offline Tywin

The knowledge is power...Everything is connected...
The Turtle continues at a steady pace ...

Offline seb21051

  • Full Member
  • **
  • Posts: 229
  • Michigan, USA
  • Liked: 113
  • Likes Given: 371
Re: Amazon Project Kuiper Broadband Constellation
« Reply #288 on: 01/23/2023 07:22 pm »
Could someone please explain what is meant by Amazon using Kuiper as a "Backhaul" data path? I was a trucker and backhaul meant trying to get a load out of a black hole like Miami just cover the cost of fuel to return to base.

Offline edzieba

  • Virtual Realist
  • Senior Member
  • *****
  • Posts: 6494
  • United Kingdom
  • Liked: 9936
  • Likes Given: 43
Re: Amazon Project Kuiper Broadband Constellation
« Reply #289 on: 01/23/2023 08:54 pm »
Could someone please explain what is meant by Amazon using Kuiper as a "Backhaul" data path? I was a trucker and backhaul meant trying to get a load out of a black hole like Miami just cover the cost of fuel to return to base.
'Backhaul' in a networking context are the high bandwidth links between centralised sites (hundreds of gigabits to multiple terabits of aggregate capacity). Kuiper (or Starlink, or GEO-based telecomm) is a catastrophically bad fit for these links, and fibre optics are far more economically effective even before any bandwidth concerns. At absolute best it may end up as a 3rd or 4th line failover link after multiple geographically redundant fibre lines. Since data is encrypted in transit, transmitting it via an antenna yelling bits into space (and listening for bits yelled back down) is no more or less secure than transmitting those same bits over a buried fibre.
Kuiper, like Starlink, is an 'edge' system.

Offline JayWee

  • Full Member
  • ****
  • Posts: 1021
  • Liked: 1033
  • Likes Given: 2044
Re: Amazon Project Kuiper Broadband Constellation
« Reply #290 on: 01/23/2023 09:26 pm »
Could someone please explain what is meant by Amazon using Kuiper as a "Backhaul" data path? I was a trucker and backhaul meant trying to get a load out of a black hole like Miami just cover the cost of fuel to return to base.
...
Kuiper (or Starlink, or GEO-based telecomm) is a catastrophically bad fit for these (backhaul) links
...
While generally true, there is a use case for NGSO backhaul for cell towers in sparsely populated areas. Think remote islands where running fiber wouldn't be profitable.

Starlink example: https://www.datacenterdynamics.com/en/news/kddi-launches-first-starlink-powered-mobile-tower-in-japan/

Quote
The launch of its first commercial operation is in Hatsushima, an island located off the Atami coast in Sagami Bay. ...
The company will eventually expand its network of remote towers to 1,200.

Speaking of which - I wonder how much of this business will Kuiper attract while being so late compared to SL.
« Last Edit: 01/23/2023 09:52 pm by JayWee »

Offline TrevorMonty

Re: Amazon Project Kuiper Broadband Constellation
« Reply #291 on: 01/23/2023 10:52 pm »
Could someone please explain what is meant by Amazon using Kuiper as a "Backhaul" data path? I was a trucker and backhaul meant trying to get a load out of a black hole like Miami just cover the cost of fuel to return to base.
Their primary means of moving data between AWS servers scattered around world and also large customers.
Currently terrestrial fibre which isn't necessarily secure given there are 10-100,000kms of fibre that somecone could potentially tap into and easedrop.


Online mn

  • Full Member
  • ****
  • Posts: 1116
  • United States
  • Liked: 1006
  • Likes Given: 367
Re: Amazon Project Kuiper Broadband Constellation
« Reply #292 on: 01/24/2023 12:12 am »
Could someone please explain what is meant by Amazon using Kuiper as a "Backhaul" data path? I was a trucker and backhaul meant trying to get a load out of a black hole like Miami just cover the cost of fuel to return to base.
Their primary means of moving data between AWS servers scattered around world and also large customers.
Currently terrestrial fibre which isn't necessarily secure given there are 10-100,000kms of fibre that somecone could potentially tap into and easedrop.

See above
https://forum.nasaspaceflight.com/index.php?topic=47811.msg2451623#msg2451623

Offline su27k

  • Senior Member
  • *****
  • Posts: 6414
  • Liked: 9104
  • Likes Given: 885
Re: Amazon Project Kuiper Broadband Constellation
« Reply #293 on: 02/01/2023 04:03 am »
https://twitter.com/stanshull/status/1620469046398566400

Quote from: Stan Shull
Open #space industry jobs in the #Seattle area & #WashingtonState are above 1,000 for the 18th month in a row.

A rapid decline in open roles at #Amazon's #ProjectKuiper propels #AerojetRocketdyne to the number 3 spot for space hiring in WA State after #BlueOrigin and #SpaceX.

Quote from: Ben Brockert
What number of Kuiper listings were there in October? It is unusual for a company like that to wind down hiring to that degree.

Quote from: Stan Shull
Went from 189 at end of October to 14 yesterday (in WA state, not total).  It's unusual, at least in the nearly 3 years I've been monitoring. Could be due largely due to Amazon hiring freeze.

Offline ThatOldJanxSpirit

  • Full Member
  • ****
  • Posts: 958
  • Liked: 1488
  • Likes Given: 3648
Re: Amazon Project Kuiper Broadband Constellation
« Reply #294 on: 02/01/2023 12:57 pm »
Only 43 months until half the constellation needs to be up and the signs aren’t looking good.

KuiperSat 1 and 2 has slipped from Q4 22 to probably Q2 23. Launch vehicle availability doesn’t look so good. I’d suggest it’s unlikely that there will be significant non-governmental availability on Vulcan and Ariane 6 for at least a couple of years. New Glenn, who knows - first flight 2025 according to Eric. Amazon’s cash pile doesn’t look so big anymore. And now we see a hiring freeze when we’d expect to be seeing a big ramp. Meanwhile the competition is ramping up to 100 flights per year.

I suspect that meeting the FCC timescale has slipped well into the realms of fantasy. I also suspect the business case doesn’t close.

Online DanClemmensen

  • Senior Member
  • *****
  • Posts: 6013
  • Earth (currently)
  • Liked: 4725
  • Likes Given: 2006
Re: Amazon Project Kuiper Broadband Constellation
« Reply #295 on: 02/01/2023 01:45 pm »
Only 43 months until half the constellation needs to be up and the signs aren’t looking good.

KuiperSat 1 and 2 has slipped from Q4 22 to probably Q2 23. Launch vehicle availability doesn’t look so good. I’d suggest it’s unlikely that there will be significant non-governmental availability on Vulcan and Ariane 6 for at least a couple of years. New Glenn, who knows - first flight 2025 according to Eric. Amazon’s cash pile doesn’t look so big anymore. And now we see a hiring freeze when we’d expect to be seeing a big ramp. Meanwhile the competition is ramping up to 100 flights per year.

I suspect that meeting the FCC timescale has slipped well into the realms of fantasy. I also suspect the business case doesn’t close.
Half the constellation is 1638 satellites. 1638/43 is 38 satellites per month. They have nine Atlas Vs on contract and ULA can launch one Atlas V per month per launch pad (one at VSFB and one at CCSFS).

(Note: The following are made-up example numbers) However, if they can launch 50 satellites/month they only need 33 months, and that would need to start ten months from now. I have no idea how big a Kuiper satellite is. If Atlas V can carry 25 Kuipers, they get another 4.5 months (total 14.5, or April 2024) before they need Vulcan in full operation and launching 50 Kuiper/month.

This moves it from the realm of "fantasy" into the realm of "not provably impossible", but there are still so many things that must go perfectly that I am highly skeptical. I don't even see how they can move from test launch to volume satellite production in less than a year.

Offline ThatOldJanxSpirit

  • Full Member
  • ****
  • Posts: 958
  • Liked: 1488
  • Likes Given: 3648
Re: Amazon Project Kuiper Broadband Constellation
« Reply #296 on: 02/01/2023 03:28 pm »
Only 43 months until half the constellation needs to be up and the signs aren’t looking good.

KuiperSat 1 and 2 has slipped from Q4 22 to probably Q2 23. Launch vehicle availability doesn’t look so good. I’d suggest it’s unlikely that there will be significant non-governmental availability on Vulcan and Ariane 6 for at least a couple of years. New Glenn, who knows - first flight 2025 according to Eric. Amazon’s cash pile doesn’t look so big anymore. And now we see a hiring freeze when we’d expect to be seeing a big ramp. Meanwhile the competition is ramping up to 100 flights per year.

I suspect that meeting the FCC timescale has slipped well into the realms of fantasy. I also suspect the business case doesn’t close.
Half the constellation is 1638 satellites. 1638/43 is 38 satellites per month. They have nine Atlas Vs on contract and ULA can launch one Atlas V per month per launch pad (one at VSFB and one at CCSFS).

(Note: The following are made-up example numbers) However, if they can launch 50 satellites/month they only need 33 months, and that would need to start ten months from now. I have no idea how big a Kuiper satellite is. If Atlas V can carry 25 Kuipers, they get another 4.5 months (total 14.5, or April 2024) before they need Vulcan in full operation and launching 50 Kuiper/month.

This moves it from the realm of "fantasy" into the realm of "not provably impossible", but there are still so many things that must go perfectly that I am highly skeptical. I don't even see how they can move from test launch to volume satellite production in less than a year.
I don’t disagree with your assessment Dan. SpaceX went from the first v0.9 Starlink launch to  1638 in around 30 months. So 10 months spare is about on track.

But the differences are stark. SpaceX did this with a mature F9 already operating at a cadence of about 2 flights a month. Apart from the nine Atlas Vs Kuiper has a ragbag of three launchers that haven’t even flown yet.

More fundamentally none of these companies is SpaceX. Achieving this will require focus and drive. A hiring freeze at Redmond right now doesn’t bode well for that.

I’ll concede that it is theoretically possible, but I’ll have a side of ketchup with my hat if it happens.

Online DanClemmensen

  • Senior Member
  • *****
  • Posts: 6013
  • Earth (currently)
  • Liked: 4725
  • Likes Given: 2006
Re: Amazon Project Kuiper Broadband Constellation
« Reply #297 on: 02/01/2023 04:03 pm »

More fundamentally none of these companies is SpaceX. Achieving this will require focus and drive. A hiring freeze at Redmond right now doesn’t bode well for that.

I’ll concede that it is theoretically possible, but I’ll have a side of ketchup with my hat if it happens.
I think ULA is competent. If it were not for the BE-4 engines, I think they might have a shot at launching Vulcan monthly starting in April 2024. This would be unprecedented in modern LV history, but it's not inconceivable. No new orbital LV since about 1970 has ramped up that fast. The quickest was F9, and it took four years to reach its tenth launch. I do not think Ariane 6 or New Glenn have any chance at all.

Kuiper will need some sort of extension on that deadline. I have no idea how likely that is. Alternatively, they could launch on F9, but absolutely everybody seems to think this will never happen.

All of this is in addition to the need to ramp up satellite production to 50 satellites/month by April 2024.

Offline RedLineTrain

  • Senior Member
  • *****
  • Posts: 2596
  • Liked: 2506
  • Likes Given: 10522
Re: Amazon Project Kuiper Broadband Constellation
« Reply #298 on: 02/01/2023 04:10 pm »
Only 43 months until half the constellation needs to be up and the signs aren’t looking good.

KuiperSat 1 and 2 has slipped from Q4 22 to probably Q2 23. Launch vehicle availability doesn’t look so good. I’d suggest it’s unlikely that there will be significant non-governmental availability on Vulcan and Ariane 6 for at least a couple of years. New Glenn, who knows - first flight 2025 according to Eric. Amazon’s cash pile doesn’t look so big anymore. And now we see a hiring freeze when we’d expect to be seeing a big ramp. Meanwhile the competition is ramping up to 100 flights per year.

I suspect that meeting the FCC timescale has slipped well into the realms of fantasy. I also suspect the business case doesn’t close.

You could be right that the business case doesn't close, but the FCC deadline isn't so hard.  They will just ask for more time and the FCC will almost certainly give them that time because they would like to see another strong player utilizing the frequencies.

Offline su27k

  • Senior Member
  • *****
  • Posts: 6414
  • Liked: 9104
  • Likes Given: 885
Re: Amazon Project Kuiper Broadband Constellation
« Reply #299 on: 02/09/2023 03:09 am »
Amazon gets key FCC approval for more than 3,000 LEO broadband satellites

Quote from: SpaceNews
The Federal Communications Commission approved Amazon’s plan Feb. 8 to deploy and operate 3,236 broadband satellites, subject to conditions that include measures for avoiding collisions in low Earth orbit (LEO).

Amazon got initial FCC clearance for its Ka-band Project Kuiper constellation in 2020 on the condition that it secured regulatory approval for an updated orbital debris mitigation plan.

The FCC said its conditional approval of this mitigation plan allows “Kuiper to begin deployment of its constellation in order to bring high-speed broadband connectivity to customers around the world.”

Tags: kuiper 
 

Advertisement NovaTech
Advertisement Northrop Grumman
Advertisement
Advertisement Margaritaville Beach Resort South Padre Island
Advertisement Brady Kenniston
Advertisement NextSpaceflight
Advertisement Nathan Barker Photography
0