DroneShield is an indieGOGO project from a DC aerospace engineer that aims to build a tiny, net-connected drone-detector/identifier. Based on a Raspberry Pi gumstick computer, it uses a mic to detect the audio signature of nearby drones, and then communicates about its findings over the Internet. The project promises free/open hardware and software specs on its main site. Ars Technica's Cyrus Farivar spoke to Chris Kyriakakis, a USC electrical engineering prof, who suggests the project is feasible, but believes it will need an array of mics for accurate identification. But John Franklin, who's running the effort, says the device will produce useful — if imperfect — output even with one mic.
The fully assembled drone detector costs at least $69 as a pre-order (as with all crowdfunded project, it's important to remember that you may never get your device). The project goal is to get them down to $20. For my part, I wonder how this would perform against active countermeasures: it's one thing to detect drones that aren't making any effort to remain hidden or fool detectors about which drone they are, but what about a drone that uses some technology (from playing a recording of a different drone to full-on modifications of its engines and blades) to sound different?
In any event, I expect that this is an intermediate step on the way to this thing disappearing into our phones and becoming an app that would make use of its open database of drone acoustic signatures. I can easily imagine a Drone Foursquare made by volunteers who upload drone "sightings" to realtime maps as they move around the world.