Amazon CEO Jeff BezosIn spite of being a latecomer to the race to deploy a constellation of low-Earth orbit (LEO) broadband Internet satellites, Amazon’s Project Kuiper will be a formidable competitor. SpaceX, OneWeb and Telesat already have test satellites in orbit, but Amazon has several strategic advantages.

For a start, each of the LEO broadband competitors plans to end the digital divide by providing global connectivity to end-users and small organizations in underserved areas, but they are also counting on high-margin customers—governments, enterprises, financial institutions, telephone companies, airlines, maritime companies and luxury yacht owners for early revenue. (A fifth company, LEOSAT, will focus exclusively on these commercial markets). Amazon’s complementary infrastructure will give them a strategic advantage with these early customers. They will be able to leverage Amazon’s established global Web and database services as well as their newly launched satellite ground-station service all of which will be integrated with the Project Kuiper constellation. Furthermore, when new end-users come online, they will be potential Amazon retail customers regardless of their satellite ISP.

The high-margin applications require inter-satellite laser links (ISLLs) for fast, secure long-distance communication and that technology is still under development. OneWeb has decided to forego ISLLs for their first constellation, and SpaceX launched their first 60 satellites without them and, as far as I know, has not said when they will be deploying satellites with ISLLs. Amazon may be working on their own ISLL technology or planning to partner with (or buy) Mynaric or one of the partners in the European project ORIONAS (Lasercom-on-chip for next-generation, high-speed satellite constellation interconnectivity). Note that there are political as well as technical barriers to ISSL deployment.

SpaceX and OneWeb have talked of consumer ground stations costing as little as $200, but that will require another critical technology that is still under development—cheap, mass-produced, electronically-steerable antennas the size of a “pizza box.” Telesat says they will concentrate on the maritime, aviation and cellular-backhaul markets until the cost of end-user antennas comes down. SpaceX is developing their own antenna and has filed for permission to deploy a million end-user ground stations, but an engineer working on the project told me they do not yet have an antenna that is cheap enough for the consumer market. OneWeb CEO Greg Wyler claims to have a self-funded side project that has developed a suitable fifteen dollar antenna and they may be ready to deploy. I don’t know whether Amazon has been working on small electronically-steerable antennas internally, but even if they have not, as with ISSLs, they have the funds to either partner with or purchase a company that is working on them.

Debris mitigation is another technology for which no one has a proven lead over Amazon at this time.

Amazon also gained ground on the others when Elon Musk reportedly became frustrated with the pace of development at Starlink and fired the vice president in charge of the satellite program, Rajeev Badyal, a veteran of Microsoft and Hewlett Packard and satellite designer Mark Krebs, who led Google’s aircraft and spacecraft teams before coming to SpaceX and playing a key role in developing their first two test satellites. Amazon subsequently hired Baydal, Krebs and other ex-SpaceX engineers. I wonder if they influenced Bezos’ decision to proceed with Project Kuiper.

Amazon has its own launch capability, but SpaceX has a clear lead in launch technology and capacity. Still, OneWeb has contracted with Amazon for five launches of perhaps 400 satellites starting in 2021, and one could imagine SpaceX serving their competitors as well. (I wonder if anti-trust law would require some sort of arm’s length pricing).

Amazon CEO Jeff Bezos has deep pockets so will not have to worry about raising money and, perhaps more important, he will have complete control over the project. SpaceX has had to go to the capital markets several times, OneWeb is working with a group of investors and collaborator/investors and Telesat has income from its established geostationary satellite business, but is owned by a somewhat contentious combination of Loral Space and Communications and a Canadian pension fund.

Finally, Bezos has had the skill and vision to build an array of highly successful, complementary companies from online retail to fulfillment infrastructure to Internet services to space. That is not to take anything away from the others—I suspect they were less surprised than I by the announcement of Project Kuiper. Whatever led to Amazon’s decision, it is good to see them involved in a competitive battle among would-be global Internet service providers.

Update Jul 22, 2019:

Megaconstellations points out that as a smart follower Amazon will also benefit from a matured ecosystem of suppliers and service providers facilitating mass production created and paid for by the first movers, OneWeb, Telesat and SpaceX.