Weather and wind are also a little calmer at that layer of the atmosphere, reducing the burden associated with designing the crafts. It'll still take a careful balance of priorities, though. The drone has to be large enough to fit the right amount of solar panels, so that it can gather enough power to operate on batteries at night. It has to be small enough so it doesn't need a ton of energy to remain in the air for months or years. It has to be cheap, so you can build a lot of them. And it has to be controllable — "unlike balloons," wrote Zuckerberg, in a jab at Google.
But the drones are just a platform. What's more interesting are the communications technologies Zuckerberg is considering loading onto them. Among them? Lasers.
Transmitting data by light is hardly a new idea; in fact, it's how terrestrial fiber optic networks carry information over incredibly long distances very quickly. NASA is experimenting with its own laser that's been mounted on a probe bound for the moon; last year, scientists successfully tested the data link by downloading information from the probe at a rate of over 600 Mbps — 60 times faster than the current national average. Today, even newer lasers are being developed that could quadruple that figure.
These lasers provide an advantage over more traditional communications methods that rely on propagating radio waves. The farther a radio signal gets from its source, the weaker it becomes — and the more power it takes to boost the range. So long as they have a clear shot at their target, lasers are limited only by the speed of light. And they require much less power.
There are downsides to lasers. Because of how narrow their beams are, they always have to hit the receiver spot-on — or poof, no Internet.
"The level of accuracy required is the equivalent of needing to hit a dime from 10 miles away," wrote Zuckerberg, "or hit the Statue of Liberty from California."
It'd be much easier just to aim the laser once and then forget about moving it. And that implies a satellite floating in geostationary orbit high above the Earth. Because the satellite would be in a fixed position, obtaining a laser lock would be a lot less complicated than maintaining a cheaper fleet of satellites at a lower altitude — which would amount to a bunch of constantly moving targets.
At the same time, beaming a laser from 20,000 miles up makes you vulnerable to obstructions — clouds, weather — in a way that beaming a laser from 300 miles up doesn't. If and when the laser gets blocked, you'll need a backup way of transmitting the Internet. And from geostationary orbit, wrote Zuckerberg, Internet by radio waves is going to introduce a bit of lag into your call with Grandma.
That's why Facebook's drones are going to be so important. If the lasers don't work out, the drones could blanket a moderately dense area with microwave Internet, allowing many devices to pick up the signal — like a kind of drone-based Wi-Fi.
The final product will likely involve a combination of these technologies, depending on the density of the area being served. And it'll involve a lot of tradeoffs. But as pie-in-the-sky as it may be, it's clear Zuckerberg has done a lot of thinking about what is and isn't possible. If you were doubting for a second whether he was serious — stop.