Amazon’s Project Kuiper aims to rival Elon Musk’s starlink internet business

After Amazon.com Inc. successfully launched its initial satellites, there was a brief period of concern when it seemed that one of them may have been lost. The two prototypes were sent into orbit over the Atlantic Ocean at 2:24 p.m. Eastern on Oct. 6. Initially, both satellites were in contact with an Amazon antenna located on the Indian Ocean island of Mauritius. However, during a handoff to another station, only one of the vehicles was able to establish communication. Amazon attempted to locate the second satellite by scanning the sky behind the first one, but there was no response.

The incident threatened to dampen the spirits of employees who had gathered to celebrate the launch at Postdoc Brewing, not far from Amazon’s Seattle-area space operation. The team had spent years building satellites from scratch and spent months launching them. Now that they were up, Amazon had to get in touch to make sure the solar panels started up. If not, the batteries would run out and the satellites would fail, a major setback for the retail and cloud services giant, which is already late in the race to build a profitable business selling Internet access from low orbit.

In Amazon’s Mission Operations Center, a conference room filled with large video screens, computers and cases of energy drinks, satellite operations manager Yonina DeKeyser and her deputy worked to piece together the pieces of data they had collected. Between the third and fourth contact, the guidance, navigation and control team called: the lost satellite was OK. The information flowing in could only have come from a couple of healthy spacecraft. Project manager Rajeev Badyal shouted in victory.

An Amazonian looking at his phone at the brewery broke through the noise and raised his fisted hands as he shouted “We have the power!” His coworkers cheered. The team later discovered that some of Amazon’s ground-based antennas had been looking in the wrong place, thinking that another satellite had passed the first.

Amazon executives tend to describe their satellite project, Project Kuiper, in philanthropic terms, emphasizing its potential to connect people in remote or impoverished areas with education and global trade. Less altruistically, Amazon also hopes the $10 billion-plus project can turn it into a global telecommunications giant. The company plans to sell roof antennas to individual internet users, cloud computing and data recovery services to companies, and connections to wireless companies to connect remote cell towers to their networks starting in 2025.

Project Kuiper is one of the Seattle-based company’s biggest bets, one of the few to survive two years of austerity measures that have eliminated many of the speculative projects started late in Jeff Bezos’ tenure as CEO. It’s a huge undertaking in an arena that has seen more bankruptcies than successful companies. Broadband is already widely available, and in many places that don’t have it, it’s not clear whether people can afford space-based internet. Some Amazon watchers see Project Kuiper as another front in the race between Bezos and fellow billionaire Elon Musk, whose SpaceX operates the Starlink constellation of internet satellites.

Amazon is betting that its system will advance the latest technology and can provide the capacity and Internet speeds to compete not only with Starlink, which has a big head start, but also with terrestrial telecommunications companies. At least Amazon is building an alternative to Musk’s service at a time when governments and companies are looking for ways to reduce their reliance on the unpredictable and controversial businessman.

Over the past two months, Amazon engineers have put their maiden satellite through several tests. They’ve made a video call, bought a set of toy rockets on Amazon.com and tested a laser system designed to extend the reach of each satellite. Now comes the really hard part. To meet the terms of the license granted by regulators, Amazon must build — and find a ride into space — every day until July 2026, the equivalent of two satellites a day.

“Building two satellites is very difficult,” Badyal said. “Building 3000+ is exponentially more difficult.”

Project Kuiper, named for the belt of dwarf planets, ice and rock behind Neptune, was born out of a thought experiment, says Dave Limp, Amazon’s longtime consumer electronics executive. Bezos had periodically asked executives to consider distant obstacles that could slow down the company’s operations. This exercise led Amazon to spend billions on warehouse robots and fleets of aircraft, large equipment and delivery trucks.

About six years ago, Amazon executives became fixated on broadband internet. Their various businesses, including retail websites, movie studios, and enterprise software, all depend on Internet access. “If you wanted to grow, you had to find these hundreds of millions of people who weren’t currently Amazon customers,” Limp said in an interview. “Well, what’s the limit to getting them there?”

Among other ideas, Amazon explored drones and balloons for Internet broadcasting, approaches tried and rejected by Facebook, now Meta Platforms Inc. and Alphabet Inc.’s Google. Amazon decided to deliver the Internet from satellites.

The idea was not new. In the 1990s, a company called Teledesic was located near the suburban Seattle garage where Bezos founded Amazon. At the time, most communications satellites rested in a geostationary orbit, which matched the Earth’s orbit and fixed each craft in place from the perspective of someone on the ground. Such satellites support Global Positioning System, weather monitoring and in-flight web browsing.

Teledesic’s envisioned satellites in a much lower orbit, in the space station region, could take advantage of the shorter distance to Earth to better compete with terrestrial telephone and Internet companies. Despite backing from Microsoft Corp.’s Bill Gates and wireless mogul Craig McCaw, the company folded after Dotcom’s collapse. Rockets were expensive, and the aerospace industry wanted to continue making custom satellites for governments.

About a decade later, Musk took the idea and cut out the middlemen. His rocket company, Space Exploration Technologies Corp., cut the cost of getting into orbit and decided to build satellites in-house. Musk hired future Kuiper chief Badyal to bring it to life.

Badyal was born in India and spent much of his childhood in Kuwait, where his architect father had been posted. He came to the United States to study and earned a master’s degree in electrical and computer engineering from Oregon State University. He found work at the nearby Hewlett Packard campus, helping to design the print head that transfers ink to the page in inkjet printers. He later worked on the first optical mice, sparing future generations the task of cleaning a dirty trackball, before moving to Microsoft, where he helped create the company’s ill-fated Zune music player.

The rapid technological development of consumer electronics made it possible for companies outside the aviation industry to build satellites. People like Badyal, adept at on-the-fly design changes and mass manufacturing, had the right tools for a new generation of satellites that could be built quickly and cheaply. After joining the Starlink project in 2014, Badyal set up shop in Redmond, Washington, D.C. The first two satellites were launched on a SpaceX rocket four years later.

In June 2018, Musk flew to Seattle. Soon after, Badyal and much of his team were gone. Co-workers were told they had been fired. Badyal says she and Musk just decided to part ways. Musk put a second lieutenant in charge and told him to strip down the design to the bare minimum to make the barebones system work as quickly as possible. Today, SpaceX says it builds six satellites a day. There are more than 5,000 of them in orbit, and they serve more than 2 million customers.

Bezos’ principle is that Amazon only enters new fields when it has the expertise – or when it can acquire it quickly. Amazon’s satellite initiative was a two-page draft when Limp heard the Starlink founding team was looking for work. He called Badyal in August 2018.

Two months later, Badyal and five fellow Starlink alums were at Amazon sketching out a new constellation in two conference rooms covered by a black curtain that invited curious workers to climb through. “It was very safe,” said Naveen Kachroo, one of the first workers. When their plan — 3,236 satellites criss-crossing the Earth at an altitude of 590 to 630 kilometers — became public months later, Musk called Bezos a copycat on Twitter.

Amazon engineers designed the terminal, a gadget that customers would one day use to receive data from satellites, which they believed they could make for about $750. Bezos sent them back to the drawing board. It had to be even cheaper. Amazon’s antenna manager Nima Mahanfar and his team combined some of the antenna functions, and the company says it can now build its main terminal, an 11-inch-square terminal, for less than $400. It offers Internet speeds of up to 400 megabits per second, which is about twice the median broadband speed for US homes.

Project Kuiper employs more than 1,600 people, both consumer electronics veterans and career aerospace experts. DeKeyser, a satellite operations manager, has a master’s degree in aerospace engineering and says quitting Amazon would have been unthinkable earlier in his career. The team is a rare organization within Amazon, led mostly by people from elsewhere. Chief Satellite Engineer Paul O’Brien, Kachroo and Mahanfar all worked on Microsoft’s Zune.

“You have to innovate much faster” than traditional space manufacturing, said Badyal, a mustachioed, gray-haired engineer with a gravelly voice and a penchant for classic cocktails. His office at Project Kuiper’s headquarters in Redmond, in a building that once manufactured forklifts, overlooks a research and development lab where engineers make custom aluminum parts, assemble circuit boards and test antennas in a cavernous anechoic chamber.

Amazon’s satellites combine technology that pushes the cutting edge—including optical satellite links, more commonly known as space lasers—with simple, proven components that limit cost or weight. “Kuiper is designing spacecraft that are fewer, larger and more powerful” than SpaceX’s first-generation vehicles, said Caleb Henry, who tracks private-sector space companies at Quilty Space. “There is a real difference between these two design philosophies.”

The satellites arrive in space packed inside the nose of the rocket and begin their first orbit steeply until an automated system reorients them toward Earth. By then, the solar panels folded in at launch should deploy automatically, relying on nearly a century-old technology: actuators that heat a plug of wax that expands and depresses a bolt that releases the array.

When a customer loads a website, the home terminal sends a signal to a circular array of up to thousands of antenna modules that look like small, green, two-pointed Legos. The bowl-shaped gateway antennas route the request to Amazon’s ground stations that lead to the Internet. Responsive data starts back up and then down to the terminal from one group of Lego blocks.

All this happens in milliseconds when the boat is traveling at 17,000 miles per hour (27,359 km/h). When a satellite zooms out of view, another should be visible. Each has its own propulsion system. The Amazonians compare the power of the thruster to the flapping of a dragonfly’s wings, which can overcome the pull of gravity when shot for hours in the vacuum of space.

Amazon at one point sought to manufacture the satellites for $500,000 apiece and keep them under 500 kilograms (1,100 pounds), according to two people familiar with the matter. The size and weight of Amazon’s upcoming production models could not be learned. Based on Amazon’s launch vehicles, Quilty Space estimates the weight of the Kuiper satellites at 600-800 kilograms. A photo released by Amazon of the prototypes on their way to launch showed each one enclosed in a cubic steel box about the height of a person.

Kachroo, now Kuiper’s director of business development, says Amazon sells connections directly to individual Internet users and through wireless and broadband providers, depending on the country. Amazon has announced cooperation with Verizon Communications Inc. in the United States, Vodafone Group Plc in Europe and Africa, and NTT in Japan. Service tests will begin in the second half of next year, and Amazon expects to eventually sell to tens of millions of customers.

“We want to serve businesses, governments, schools, hospitals and mobile operators, so we don’t have one channel or segment where we make money,” Kachroo said. Amazon, which has not disclosed prices, has licenses in more than 15 countries so far, including Brazil, Canada, France, Mexico and the United States.

The company offers businesses and governments private connectivity through its Amazon Web Services unit, guaranteeing a quality of service that SpaceX has yet to offer. AWS, the largest seller of leased computing power and data storage, will in the next few years be able to offer product bundles that include Internet access, an advantage that Amazon’s cloud computing rivals cannot match on their own.

Kuiper’s staff doesn’t usually bring up Starlink (another Bezosism: don’t bully the competition), but analysts say Amazon has a chance to differentiate itself by simply running a satellite business free of Musk’s personal drama or business entanglements. Other companies build what the industry calls mega-constellations, but Starlink’s is by far the largest and most capable.

Taiwanese officials, who are looking for backup Internet connections in case of war with China, are wary of relying on Musk, who has business ties to Beijing, Bloomberg has reported. In Ukraine, Starlink has been a lifeline since the Russian invasion, but earlier this year it emerged that Musk refused Kiev’s request to extend coverage to allow Ukraine to advance. The world’s richest man has called for an end to the conflict on terms favorable to Russia, and his biographer published text messages between Musk and Ukraine’s deputy prime minister.

SpaceX, which did not respond to requests for comment, also avoids the long-term contracts and exclusivity agreements that corporate customers typically seek, said Lluc Palerm, an analyst at Research NSR. “They are not considered the best partner in the industry”

In an interview, Kuiper’s head of regulatory affairs, Julie Zoller, did not specify how Amazon would deal with the political mess, saying the company would defer to the State Department. Zoller, who started his career installing satellite equipment at US military bases, acknowledged that Amazon doesn’t imagine CEO Andy Jassy negotiating terms of service via text message. “Customers are literally saying, ‘Why can’t you go faster,'” Kachroo said. “They love that there’s competition.”

Executives claim Kuiper is on schedule, but the company hoped to get its prototype out almost a year earlier. The proof is in orbit: each boat’s aluminum hull component will be engraved with the names of the people working on the project in August 2022.

Amazon’s first trip into space—on a new rocket built by the startup—exploded on the launch pad. Its second, new Vulcan Centaur, built by United Launch Alliance, the US space agency, was due to launch this summer before an explosion during testing. Desperate to get its satellites flying, Amazon leased the Atlas V, a 21-year-old ULA rocket capable of carrying much heavier payloads. The launch was hiring a rocket-equivalent city bus to transport two people to the movies.

Now Amazon has to get the rest of the satellites up. Project Kuiper’s is the largest commercial launch order in history, including 47 ULA launches as well as ArianeGroup and Bezos’ Blue Origin rockets. But only one of those rockets — the Atlas, which Amazon has booked for eight launches — has flown. Blue Origin has never sent a spacecraft into orbit, and the rocket it hopes to get is years behind schedule. (Limp, Badyal’s old boss, left Amazon this month to run Blue Origin.)

Earlier this month, Amazon booked three launches with SpaceX, an awkward deal due to Kuiper’s tight launch schedule and lack of options. Amazon says it has been in discussions with every major launch provider for years. It also denies that the decision was influenced by a lawsuit filed by a pension fund that said Amazon did not consider using SpaceX in part because of competition between Bezos and Musk, which drove up costs. Amazon says the claims are unfounded.

ULA is expanding its plant in Alabama and retrofitting a facility in Cape Canaveral, Florida, to quickly stack Kuiper satellites on rockets and transport them to the launch pad. Suppliers of rocket engines and avionics equipment are increasing their production. “Everything is on track to be done on time,” said ULA CEO Tory Bruno. “As long as we don’t have to completely change the design, we’ll be fine.”

Even if Amazon’s satellites do make it to space, Project Kuiper’s Federal Communications Commission license requires 1,618 of them to be there by July 2026 and the rest three years later. Amazon plans to build them at its own manufacturing facility in Kirkland, Washington, where crews will continue to assemble machines and make utilities. So Amazon’s first satellites are being assembled at Kuiper headquarters, which is being rebuilt from a research and development facility into a collision production line.