Amazon hitches a ride with SpaceX for Project Kuiper launches Amazon is signing a contract with SpaceX for three Falcon 9 launches to help "support deployment plans" for the Project Kuiper satellite broadband initiative. This is all about adding capacity to support the schedule, Amazon insists. Yet of the heavy-lift rocket launches Amazon has so far procured from different providers, a …
I predict they will lose some spacecraft in a launch.
Integration of a satellite launch bus and the rocket that launches it is not a trivial task. I'm sure all their providers will be very professional and helpful but Project Kuiper now needs four separate integration teams to deal with them.
One of the big lessons of big systems development is courtesy of IBM: "Adding people to a late project makes it later." One corollary of this is "not adding people will result in overwork for the people you have."
Meh - I suspect that the integration will not be the biggest challenge they face.
And each of the launch suppliers will be keen to ensure that integration is done properly.
The chances of the Ariane 6, Vulcan and Glenn rockets all being completely reliable from the get go however... I suspect there is a decent chance they'll lose some to launch issues.
Atlas and Falcon both have good demonstrated reliability.
"Atlas and Falcon both have good demonstrated reliability."
The last time Falcon blew up on the launch pad and destroyed the payload, it was a satellite for Facebook's foray into providing services to unserved areas. Coincidence?
It's difficult to see the business case for a mega-constellation in LEO much less several.
Yes - coincidence.
If it happens again... then I'll revisit that opinion.
Starlink provides services that others can't - it works well, and is clearly generating significant revenue...
There is also a wierd market which they haven't tapped yet - and that's only going to be available once the laser comms come online, the transatlantic latency can be better than oceanic fibre which is important for high frequency traders - and they will pay for it, and potentially pay alot.
While Bezos probably drank a whole bottle of pepto-bismol after signing this deal, it fits. Amazon won't want their main competitor profiting off of their launches, so they are willing to fund even experimental launch platforms. Unlike a 30-year com sat out in geo-syncronous orbit, their Kuiper sats have a limited life span and are undergoing rapid development, and they are deploying a large constellation, so they can afford to lose some to launch failures.
What they can't afford is to let SpaceX get so far ahead they are permanently closed out of the market. So they need to reliably get enough of their block one sats up in orbit and generating data for them. SpaceX has a good record and the best launch cadence.
SpaceX will play along, as they still have first mover advantage, and they are probably gunning for a duopoly to keep the regulators at bay. If the new launch platforms deliver, they will be pushing more and bigger sats into their own Starlink constellation, and if Bezo's is on the sidelines when that happens, he may not get a chance to catch up. He can't afford to wait, but will still try to either pay other launch partners or play catch up with New Glen.
"What they can't afford is to let SpaceX get so far ahead they are permanently closed out of the market. So they need to reliably get enough of their block one sats up in orbit and generating data for them. SpaceX has a good record and the best launch cadence."
I suspect the primary motivation is the licence terms. IIRC, Kuiper has to have a certain number od sat in orbit by a certain date or their licence is revoked. It was already looking a bit close to the bone before the latest round of launcher delays.
Jeff is not afraid to throw huge amounts of money at a market he wants to control.
I have confidence in the payload integration teams at SpaceX, ULA and Ariane Space (even though some of them have had time for staff turn over without active contracts). I was worried about Amazon having to work with four launchers at the same time but it probably won't work out like that. ULA/Atlas goes first - waiting on satellites from Amazon. Ariane is probably second because ULA/Vulcan will be waiting for engines. SpaceX/Falcon does not start until mid 2025 so they are likely third. The difficult race to predict is Blue Origin vs SpaceX/Starship. Starship has more launches booked than licenses (will likely get more licenses per year once environmental effects of Starship can be measured from experience). New Glenn has a bigger engine problem (7/launch) than ULA (2/launch).
Perhaps some Kuipers will go up on the next generation Antares.
Even if all three of the other rockets (Ariane 6, Vulcan and Glenn) make their initial launches "on time", there is a big difference between that and having the sort of launch cadence needed for Kuiper. Heck, can BO put out enough BE-4 engines for both New Glenn and Vulcan to meet their schedules (which also include non-Kuiper launches)? While NG's booster is to be reusable it's going to take a number of launches before they will be reflying boosters - more BE-4 engines.
I would be pleasantly surprised if all three launch providers could meet their schedules. But I expect SpaceX will get at least a dozen extra Kuiper launches to cover it.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2025
