From IEEE Spectrum:
Airspace above 500 feet is already well regulated by the FAA, but there's a potentially dangerous void that's about to get really crowded between 500 feet and either your skull or the ground, whichever comes first. At NASA's Ames Research Center in Moffett Field, Calif., in the heart of Silicon Valley, NASA engineers and researchers are working on a way to manage that void, and the system they're trying to put in place would help out small autonomous aircraft in a number of ways:
Airspace Restrictions: For manned airplanes, a lot of what air traffic control is all about is telling you where you can't go, as opposed to where you can. Especially around airports, airspace is heavily restricted over a succession of increasing altitudes, and whenever there are special air operations in place (like around air shows or forest fires), special no-fly zones pop up. Autonomous drones (and piloted ones, for that matter) will have to dynamically adapt to, and respect, airspace rules that may change rapidly. And for drones operating close to the ground, restricted airspace would, when possible, also include obvious stuff that you wouldn't want to run into, like buildings.
Flight Corridors: I don't think that we're going to see the sort of drone delivery that Amazon is promising within the next few years, and it sounds like Google doesn't either, with the company suggesting that it might be "a few years but less than a decade" before consumers see any tangible uses from the technology. My guess is that the first time we'll see delivery drones doing anything useful will be dedicated point-to-point service, where the landing areas can be carefully defined and controlled. That solves a huge amount of the problems that we identified with delivery drones, but it still leaves the issues with actually flying around: namely, running into stuff. Having established flight corridors would allow delivery drone operators to carefully define and control flight paths as well, ensuring that there are no obstacles or other aircraft for the drones to smash into.
Operating Areas: NASA suggests that we might start seeing drones for agricultural monitoring and inspections in about a year or so. It would be handy, and safe, to be able to schedule a time and an area where you want a drone to fly around doing work, with some amount of confidence that you wouldn't be bothered.
Again, we want to stress that the eventual adoption of a system like this would be fantastic for small autonomous drones, and it will likely also be necessary for their safe long-term commercial use, including delivery drones as proposed by Google and Amazon. But it's not going to solve every problem that they have, and companies like Amazon and Google have a lot ofmaaaaaybe impossible work to do before they'll be dropping packages off on your doorstep.
Via [ New York Times ]
much work we must work. First legalize security system of drones to protect the citizen (parachute etc). We need to think many drones on the air. Second attribute a identifier for each drone. Thrid create a virtual road with the air with controlled by a special semaphore... etc etc
I'd imagine any sort of ATC system for drones would be a fairly automated one. Having done ATC myself for the FAA, human occupied aircraft can keep you busy enough, I couldn't image adding a slew of drones to the mix. Pre-approved corridors certainly do make sense, even in less populated areas as they can be surveyed to ensure obstacle clearance. They use corridors already in more congested airspace's. I wonder if they'd still categorize a flight as VFR or IFR with respect to FPV or autonomous? :-)