Amazon Begins Project Kuiper Deployment

Amazon is a company which is used to making big investments.  Over the past 30 years, it has revolutionized the e-commerce industry by developing products such as its online store, its Amazon Prime streaming service, and its Kindle tablet.  However, even by the standards of its founder, Jeff Bezos, the Project Kuiper satellite constellation is a significant gamble.  Amazon reportedly plans to spend up to 20 billion dollars on the network of spacecraft [1], which aim to provide Internet access to people in remote areas.  Over the past three months, Amazon has taken a major step towards a reliable broadband service by completing its first three launches.

Kuiper’s development has proven to be more challenging than expected.  The program was officially announced in 2019.  Its FCC communications license, issued one year later, stipulated that it would deploy half of the planned 3,236 satellites by July 30^th, 2026.  Even with the deployment of the constellation now underway, matching that target seems highly unlikely.  Kuiper’s largest hurdle was building a production line capable of producing satellites en masse.  Unlike traditional geostationary communications satellites, a profitable megaconstellation in low Earth orbit requires multiple satellites to be built per week at a cost of no more than a few million dollars per vehicle. 

Yet, at the same time, these satellites must be advanced.  Their low altitude is naturally conducive to low communications latency, as radio waves travelling at the speed of light need to travel a short distance between a user and a satellite overhead.  Yet, because the satellites are only visible across a relatively small area, information must typically pass through a chain of spacecraft before reaching its destination.  Kuiper and its largest competitor, SpaceX’s Starlink, both use lasers to beam up to 100 Gb per second of data between individual satellites.

Securing a stable of launch vehicles was also a challenge.  Unlike SpaceX, Amazon does not produce its own rockets.  Instead, it had to buy launches from United Launch Alliance (ULA), Arianespace, and Blue Origin.  All of these companies were in the process of developing new heavy-lift launch vehicles, which only debuted in the past year and still have low flight rates.  To bridge the gap, Amazon also purchased nine of ULA’s heritage Atlas V boosters and three SpaceX Falcon 9s.  In October of 2023, Kuiper’s program managers decided to use the first Atlas V to place two prototype satellites into orbit.  While this may sound like overkill for such a small payload, it demonstrates Amazon’s determination to bring Kuiper to market as rapidly as possible.

The “Protoflight” satellites’ six months in orbit inspired several improvements to their mass-produced descendants.  By April, the first batch of operational Kuiper spacecraft were ready for deployment.  For the KA-01 mission, Amazon opted to use another Atlas V.  As twilight spread across Cape Canaveral on April 28^th, the venerable booster’s twin RD-180 engines and five solid rocket boosters roared to life.  The rocket, flying in its most powerful configuration, had to use its full performance to place all 27 satellites into orbit.  In aggregate, they were the heaviest payload ever carried by the Atlas V in its 20 years of service.  Over 18 minutes after liftoff, the Centaur upper stage’s engines cut off as planned, completing the first of ULA’s many deliveries on behalf of Amazon.  Amazon later confirmed that all 27 satellites were functioning well.

The second Kuiper launch campaign, KA-02, was not nearly so smooth.  On June 16^th, the next Atlas V, with another 27 satellites aboard, entered its countdown sequence as scheduled.  However, the launch was scrubbed when a purge line within one of the main engines turned out to be warmer than anticipated [2].  After a short trip back to its vertical integration building for maintenance, the Atlas rolled back to the launch pad.  The second countdown was routine, and the rocket lifted off early in the morning of June 23^rd.  From that point onward, the mission was nearly identical to its predecessor.

Yesterday, Amazon’s first SpaceX Falcon 9 joined the action.  While the Falcon 9 team is more accustomed to deploying their own company’s satellites, SpaceX was more than happy to also sell launches to Amazon to increase the latter company’s launch cadence.  The flight, designated KF-01, was somewhat unusual, as it featured a factory-fresh Falcon booster.  Most of SpaceX’s launches feature reused first stages, one of which has completed an impressive 29 journeys to the edge of space and back.  The new vehicle, B1096, performed flawlessly, placing another 24 satellites into orbit.

Over the coming months, the Kuiper team and their four launch providers will strive to increase their satellite deployment rate.  Amazon has already booked two more Falcon 9s; six more Atlas Vs; 38 ULA Vulcans; 18 French Ariane 6s; and 12 Blue Origin New Glenns (with an option for an additional 15 of the latter).  These contracts should be sufficient to cover the constellation’s needs for some time to come. 

Kuiper’s financial prospects are also uncertain.  Its largest competitor, Starlink, has a six-year head start with its satellite launches.  Nearly 8,000 SpaceX satellites are already in service.  However, Amazon has some advantages of its own, including the ability to package Kuiper access with other services, such as Amazon Prime and AWS analytics.  SpaceX CEO Elon Musk’s disputes with other nations, such as Canada and Ukraine, has also created business opportunities overseas.  Both companies must also manage a limited customer base, which constrains their revenue [3].  The coming years will be an interesting period as these massive corporations attempt to prove the viability of megaconstellations – driving up the U.S. launch rate in the process.