OneWeb's Plan to Beat SpaceX to Provide Satellite Internet to Everyone on Earth

​An employee of OneWeb walks on stage, looks at a mockup thatched roof just sitting there, and at the small metal box sitting upon it. Greg Wyler, the man who wants to connect the entire world to the internet, asks if she can install the device, which will create large wifi and 4G hotspots around the world. She pulls one solar panel open, then another. She walks offstage. The process takes five seconds.

"That's it," Wyler says. "It's done."

Wyler is the man who could challenge Elon Musk in the race to connect the world with fast, space-based internet that can hit any spot in the world. He's already built one successful space internet company, and is now in the process of building another.

Before Musk announced his plan to put 4,000 internet-providing satellites in space, Wyler had launched O3b, a company that has 12 internet satellites in orbit and is already the largest internet provider in the Pacific. OneWeb is his next venture, a project that, he says, will start with 648 low Earth orbit satellites that can beam internet to anywhere in the world.

But rather than beaming the internet to your cell phone, it'll instead beam it to the devices that Wyler showed off for the first time Tuesday at the Satellite 2015 conference in Washington, DC.

"These are terminals that will cost $250 that you can put on top of your barn. It'll give you 16 megabits per second and create an LTE, 3G, and 2G operator to empower you to extend your own network," he said. "These terminals are small and they're light, they're easy to install, and you can put them where you want."

"The dream is, how do we connect all the schools of the world? We have a vision and a plan to allow 2 million schools to get connected," he added. "You put the terminal on top of the school, there are no buttons, no wires, nothing to point. You just place the terminal on the school."

The goal, Wyler says, is not to become an internet service provider or to beam internet directly to your cell phone or computer, but instead to sign contracts with local telecom providers to extend their own networks. In theory, most people who use OneWeb will never know that they're actually using it.

For simplicity's sake, imagine you have Verizon service (though much of the early work will probably be done in rural Africa, Latin America, and Asia). Your phone will use Verizon's towers all the time, but if you drive outside of Verizon's network, you'd instead use a OneWeb terminal, which Verizon would connect to and pay OneWeb to access. Verizon's incentive in buying into the system would be to sign up more customers in more remote areas.

OneWeb hopes the plan will bring improved internet access to rural areas at a faster clip than traditional infrastructure. It's really expensive and not very lucrative to wire the developing world with fiber—Wyler has tried with a past company.

"I've walked through mud, spliced fiber in a rainstorm, put the fiber back together," Wyler said. "Fiber in Africa is really, really hard … mobile operators are good at high GDP and high population densities. At lower densities and lower GDPs, they're bad. They have no economic reason and no economic means to extend their networks there."

Wyler began combatting this problem with O3b, but the latency and speeds just aren't good enough—the satellites orbit too high for fast internet speeds. The more satellites that are launched, and the lower they fly, the faster the internet speed will be—Wyler and Musk have both suggested low Earth orbit satellite connections may one day rival fiber optic speeds.

"It's like if you're living on the fourth floor, and your hot water heater is in the basement, you wait. That's latency. Well, [geosynchronous orbiting internet-providing satellites] are on the 36th floor," he said. "We want to bring you to the first floor so you're on par with everyone else."

A second version of the ground terminal can be popped on the roof of a car or ambulance to provide it constant cell phone and internet access.

"If you put it on your car, it just works. It doesn't take anything from the operator, it doesn't turn on until the operator's signal is not around," Wyler said.

From the ground infrastructure standpoint, the project sounds both doable and promising.

It also sounds audacious, and it is. For OneWeb to have any hope of succeeding, it's going to have to figure out how to launch all those satellites, for cheap. SpaceX probably isn't going to help him, which is why Wyler has partnered with Virgin Galactic instead.

Virgin Galactic's Rocket Launcher One is a still-under-development launch system that will, like its space tourism missions, start on a normal plane. The plane will fly high into the atmosphere, a rocket will drop and then launch itself into orbit, saving lots of money on fuel costs.

"The advantages are purely technical—by starting at a higher altitude, we avoid launching through the densest part of the atmosphere, we're skipping the steepest part of the curve [to orbit]," George Whitesides, CEO of Virgin Galactic, said Monday. "If you're launching a system carried out over the ocean and at a high altitude, it's a lot simpler and cheaper and faster."

It works in theory. The problem is, Virgin Galactic hasn't done any live tests yet (it says a dry, on Earth test of the rocket is forthcoming this year). If OneWeb is going to be successful, it's got to get to space somehow.

Maybe that's why Wyler spent very little time Tuesday talking about the satellite infrastructure and instead focused on how it'll work on the ground.

"We're not a satellite company," Wyler said. "We just happen to have satellites in the system."