Small drones first. The FAA must release a set of guidelines by Aug. 14, 2014 covering small, hobbyist-type unmanned aircraft that weigh less than 55 pounds and can be monitored by direct line of sight. At the moment, hobbyists simply enjoy an exception to a much larger regime that mostly forbids the use of drones in national airspace. (You can also apply today for an "experimental certificate" from the FAA, but it comes with strict restrictions.) The new system would likely require pilots to get an FAA permit to fly small drones.
Drone test sites. Two dozen states have applied to host six test ranges the government will use to gather information about how drones — presumably both small and large — will change the way we live. That information is going to inform future attempts at rulemaking. The sites are being chosen as we speak, and a decision is expected sometime within the next month and a half.
Sense and avoid. Human pilots rely on radar and their eyesight to steer clear of hazards. But drones have to be taught how to distinguish those threats from the rest of the environment — no easy task. So the government has been working on what it calls "sense and avoidance" technologies that would help drones avoid both other planes and ground-based dangers like mountains or communications towers. Beyond the technology itself, rules will have to be established governing how sense-and-avoid is to be integrated into a national system.
Certification. We'll need trained professionals for this brave, new world of unmanned aircraft. Given how different drones can be from manned planes, the old ways of doing things won't be sufficient. Pilots will need new skills as they adapt to flying by computer screen rather than by looking out the window. Air traffic controllers, too. The training programs themselves will need to be developed, as will the standards for testing and certificates.
Security. The FAA says it wants to be sure that the link between a pilot and a drone is never broken. To that end, it has proposed greater research into the vulnerability of the command and control links. It's also investigating how a drone should act if its systems have been compromised. Not a bad idea, considering the havoc that might be wreaked by a malfunctioning drone — hacked or otherwise.