We're taking advantage of the increasing popularity of connected devices to solve a common problem—finding your stuff. Specifically, we designed navigation and control algorithms for a tiny flying robot to really quickly locate Bluetooth-enabled devices. This is especially useful because even if you can't see the thing you're trying to find (e.g., if it's buried under a pile of clothes), our drone can still sense and locate it. Our algorithm also isn't exclusive to Bluetooth devices—it relies only on signal strength, which is available with pretty much every wireless standard (Wi-Fi, RF, etc.).
In general, the problem of finding things based on signal strength is hard because the information you have is very noisy and only gives you "distance" rather than "direction." To solve this problem, we look to how fruit flies locate bananas. They exhibit a distinctive flight pattern consisting of long segments of forward flight interspersed with rapid right-angle turns. By turning each time odor concentration starts to decrease, flies are able to locate sources of food very efficiently. Our drone mimics these behaviors.
We also designed a web interface where you can add devices and command the drone to locate them, and monitor its progress in real-time.