Designing of web-based lattice path controller for indoor environment by using multiple quadrotors

Unmanned aerial vehicles (UAVs) become very popular in last years by the help of increasing computing power per area and per cost. While UAVs with global positioning system (GPS) can easily operate to flight autonomously, this and such sensors data cannot be always trusted. And most of the cases for small and nano scale UAVs, we cannot use these kinds of sensors because of cost, complexity and weight. Safely and reliably operating close to unknown indoor or GPS-denied environments requires improving UAVs' sensing, localization, and control algorithms. To solve and improving for a UAV is one problem, extending it to a multiple UAVs is another problem. We study and develop a framework for multiple UAVs to command and control from web-based front-end. In our experiments, a quadcopter platform called AR.Drone from Parrot Inc. is preferred and used because of its open-source API and kernel level arrangements. Test flights inside a warehouse validate the framework is capable of control one or more quadcopters in given different lattice-based path from a web browser. UAVs are capable of localizing its position with using IMU, front and bottom camera. All UAVs are interconnected to each other and controller computer through a self-healing network, so if one or more quadcopter fails because of lack of battery or any other circumstances, the rest of the group continues its mission.