That's the first question and a relatively easy way to ask it; the harder version is "can Amazon get their Kuiper internet satellites into orbits fast enough to be a player in the satellite-based internet access world?"
What prompted this thought was this morning's 6:54 AM launch of the ULA Atlas V carrying Amazon's next 27 satellites into Low Earth Orbit. It's worth pointing out that this was only the second launch of a full load of operational satellites for Amazon's Project Kuiper, a network envisioned to become a competitor to SpaceX's Starlink. The first launch of a batch like this was Monday, April 28th, eight weeks ago.
Amazon is moving toward a goal of 3232 satellites in low-Earth orbit, enabling coverage of most of the populated world. Back in April of '22, word broke that Amazon had chosen "everybody except SpaceX" to launch "the majority" of their 3,236-satellite Project Kuiper constellation. Amazon had contracted for 68 rocket flights from United Launch Alliance, Arianespace, and Blue Origin. Before those 68, they signed on for five Atlas V launches of the initial satellites in that eventual constellation. The first two (prototype) Kuiper prototype satellites launched Friday, October 6th on an Atlas V after being swapped to it from the continually-slipping out maiden flight of Vulcan centaur.
Amazon has procured more than 80 launches with four companies to put all of these satellites into orbit. ULA won the lion's share of the launch contracts to deploy more than half of the Kuiper constellation. Amazon purchased the last nine Atlas V rockets before ULA retires the vehicle in favor of the newer Vulcan rocket. Shortly thereafter, Amazon signed a contract for 38 Vulcan flights, each of which ULA says will deliver 45 Kuiper satellites into orbit.
Europe's Ariane 6 rocket is under contract for 18 missions with Kuiper satellites. The New Glenn from Blue Origin, owned by Amazon founder Jeff Bezos, will launch at least 12 times to add satellites to the Kuiper network. Amazon has a contract option with Blue Origin for up to 15 additional New Glenn launches.
SpaceX's Falcon 9 was contracted for a number of launches as well, also back in 2023. The Falcon 9 is the only rocket that has actually flown at the kind of cadence envisioned for the Kuiper network. The F9 currently flies at an average rate of once every two days, primarily on missions with SpaceX's own Starlink broadband satellites.
Amazon's competitor will launch the next batch of Kuiper satellites on a Falcon 9 as soon as next month. Those satellites are already at Cape Canaveral for final launch preparations, suggesting that Amazon is finally overcoming difficulties in activating its spacecraft factory in Kirkland, Washington. A source recently told Ars that satellites for the fourth Kuiper launch are undergoing final testing in Kirkland. For the first time, Amazon will soon be in a position of waiting on rockets before launching them.
One of the main criteria that Starlink was designed to meet is latency - largely set by the amount of time radio signals take to travel from Earth to the satellites. As with everything involved in the design of systems like radios, there are compromises. To get the lower latency, the satellites need to be lower to minimize the time for the radio signals to go to/from higher orbits, but being lower limits the circular area they can communicate with, so there need to be more satellites in orbit to get the desired coverage. As of May 30, 2025, there are currently 7,578 Starlink satellites in orbit, of which 7,556 are working. They have been reported to have filed paperwork with an international regulator to loft up to 30,000 additional spacecraft. They will be launching their own orbital internet for a long time to put up 30,000 satellites 20 (or so) at a time.
With the Kuiper constellation containing the 3232 number of satellites quoted above, they don't have to put as many satellites in LEO as SpaceX does, but ULA is working on expanding their launch capabilities.
ULA aims to double its launch capacity at Cape Canaveral to meet Amazon's demands. And Amazon is investing heavily to make it happen, committing $2 billion to pay for the expansion of ULA's manufacturing and launch capacity, with around $500 million going toward upgrades at the Florida launch base.
United Launch Alliance's Atlas V rocket soars through sky over Cape Canaveral,
Florida, shortly after sunrise Monday. Credit: United Launch Alliance
I started this post by asking "can Amazon get their Kuiper internet satellites into orbits fast enough to be a player in the satellite-based internet access world?" I don't know. It sounds like they're doing well but the only player mentioned here that seems to have demonstrated they can do it is not United Launch Alliance.
22 out of 7500 satellites are not working? I've been out in the real world, and that kind of performance for that complex a device as a satellite is jaw-dropping.
ReplyDeleteSpaceX seems pretty adamant about deorbiting sats that aren't working, so a better question is how many they dropped in the ocean for failing to work properly. That "7,578 Starlink satellites in orbit, of which 7,556 are working" is a snapshot at some time.
DeleteAh, ok, that adds context to the number. Thanks.
DeleteSpaceX has removed many of the earlier 1st gen Starlinks that all the astronomers were whining about because they supposedly clouded up the sky. Later 1st gen and later generations are much less shiny.
DeleteIn a word, horseshit. That would be Bezos. What businessman says, Anyone but (the most capable launcher)?
ReplyDeleteThere seems to be a lack of seriousness in the effort to get Kuiper fully operational. It all seems very petty.
Other than a billionaire personally at odds with another billionaire, how is Space X a competitor of Amazon?
I wonder what Amazon stockholders think about such small mindedness.
In other news, tracking orbiting sats and junk seems a growth industry. If Kuiper does get up, Amazon will either spin off a company just to track and maintain the constellation, or sub out to a contractor.
"I wonder what Amazon stockholders think about such small mindedness."
DeleteIn August of '23, shareholders of a pension fund that includes Amazon stock sued Amazon claiming the company, founder Jeff Bezos, and its board of directors breached "their fiduciary duty" and failed to consider SpaceX during the launch service procurement.
I stumbled across that in this old post while researching some points for this post.
I saw a report somewhere that in order to keep their license, Amazon must launch about 3,000 Kuipers by end of 2026. Loss of license invalidates any further launch licenses, supposedly.
ReplyDeleteI don't see them launching more than 500 by end of 2026. I'd be surprised if they get more than 100 up by end of this year.
in none of the old SciFi stories did I ever read about aliens who wanted to land on Earth having problems evading the satellite net surrounding our planet.
ReplyDeletehas anyone done a study to determine the negative effects these two (or more) nets will have on various parameters necessary for transmission (and reception) of signals beyond the sphere of their envelopment?
are there any studies showing the effects that they will have on "space travel"?
(the phrase is in quotes, but I'm not joking)
The best answer I know of is there are no such studies. The simplest explanation is that the surface area of the earth is 197 million square miles. The equivalent area is marginally bigger because of increasing the diameter marginally to include orbital altitudes. There's not enough satellites in higher and geosynchronous orbits to justify pushing the diameter more than "a few" hundred miles, which makes 197 million a lower limit that we know the reality is bigger than. Now divide that 197 million square miles by the number of active satellites, which Live Science says is 11,700. In fact, since the number gets easier, let's make that 19,700 active satellites. It's not, but it could be in a decade.
DeleteThat would mean one satellite every 10,000 square miles. And satellites are small, so almost all of that 10,000 square miles is empty.
It just seems like running into a satellite is a pretty low risk.
SiG, that's the 'big sky' theory. The flaw is that it is based on a static projection, whereas reality is dynamic.
DeleteWhile most of that 10,000 sq. mi. is empty, which of those sq. mi. are empty changes every moment.
A launch from earths surface may have a narrow window, at least narrower than that suggested by the theory.
Yes, collision, even near hits, is low risk. But not as low as suggested by the theory.
Put another way; congestion will occur sooner than the theory would allow. This because any one orbiting body occupies many locations over time in the dynamic realm of reality. Bread crumb tracks of multiple orbiting bodies would feature an expanding and contraction cycle as bodies come near then further.
At what time will your launch vehicle occupy which altitude for all points of that distance from Earth surface to final orbit?
What is closest point of contact to any of the thousands up there now?