Can These Small Satellites Solve the Riddle of Internet From Space?

Satellite internet is notoriously expensive, and the business of providing it is notoriously brutal. A startup called Astranis founded by rocket scientists from Stanford University and MIT wants to change that.

The company’s satellites are about the size of a mini-fridge, while traditional satellites are closer to the size of a bus. “Traditional satellites cost hundreds of millions to build,” says co-founder and CEO John Gedmark. “These cost tens of millions.”

Gedmark says Astranis focuses on the software that controls its satellites, which make them more flexible than traditional satellites. By using software-defined radio, for example, the company can more easily change what radio frequencies a satellite uses.

Astranis is joining a new space race led by Elon Musk’s SpaceX and Richard Branson-backed OneWeb to build a new breed of satellite internet network that can reach the entire globe and perhaps compete with more traditional forms of broadband, like cable internet and cellular data services. But Astranis is taking a different approach than other space companies by focusing on bringing down costs.

Traditional satellite communications systems float in what’s called geosynchronous orbit, around 22,000 feet above the Earth. These satellites can provide internet access to remote parts of the Earth, as well as airplanes. But the connections can lag, which isn’t good for real-time applications like online gaming or video conferencing. SpaceX and OneWeb both aim to overcome this problem by launching satellites into what’s called low Earth orbit, which ranges from roughly 100 to 1,250 miles above Earth.


The problem is that in order to reach the entire world from low Earth orbit, these companies need hundreds or thousands of satellites, raising the system’s cost. Previous attempts at building low Earth orbit networks ended in bankruptcy, including the Bill Gates-backed Teledesic and satellite-phone companies Globalstar and Iridium.

SpaceX and similar companies, like Jeff Bezos-backed Blue Origin, are trying to reduce the costs of launching rockets, which lower the cost of building such a network. But it’s not yet clear whether these companies could offer internet access at rates that subscribers can afford, and skeptics worry this will end up costing more than just trenching fiber and building cellular towers.

Astranis is sticking to geosynchronous orbit, but it’s building small, cheaper satellites that the company hopes will bring the costs for end-users into the same price range as cellular data.

That plan won’t address the latency issues of geosynchronous orbit satellites, but the company just might be able to bring affordable high-speed internet to places where laying fiber isn’t practical, such as the Pacific islands. Astranis plans to sell bandwidth to internet service providers, rather than directly to end-users. It already has one test satellite in space, and plans to launch its first commercially available satellite next year. Eventually the company could launch dozens or hundreds of satellites; other geostationary satellite communications providers operate anywhere from one to dozens of satellites. Gedmark estimates each Astranis satellite will have a bandwidth capacity of about 10 gigabits per second, which isn’t much compared with a wholesale fiber-optic link, but should be enough to help get remote areas online.

The company received a vote of confidence Thursday from venture-capital firm Andreessen Horowitz, which announced a $13.5 million investment. Andreessen Horowitz partner Martin Casado said the Astranis team was the main attraction. Gedmark has a degree in aerospace from Stanford and previously worked for the Xprize Foundation. Co-founder and CTO Ryan McLinko studied aeronautics and astronautics at MIT, designed hardware for satellite company Planet Labs, and oversaw the creation of the flight-control system for Sierra Nevada Corporation’s Dream Chaser spacecraft. Other team members previously worked for SpaceX, Orbital, and Google.

“Lots of companies understand the software or networking side, but these guys really understood the space side,” says Casado.

Sattelite companies often need billions in capital, but Gedmark wouldn’t discuss how much money Astranis will need to raise to meet the company’s goals. Gedmark says he expects each satellite the company launches to be profitable.

Astranis isn’t alone in trying to find a compromise between gigantic geostationary satellites and networks of thousands of low Earth orbit satellites. O3B, founded by Greg Wyler before he started OneWeb, places satellites between low Earth orbit and geostationary orbit. That reduces latency and cuts down on the total number of satellites needed to provide service. But Gedmark argues that by sticking to geostationary orbit, Astranis is able to reach more places with less gear. “We can send a small satellite out to geo and start making an immediate dent in this problem,” he says.

Space Race