An embedded controller for quad-rotor flying robots running distributed algorithms

Abstract

Multiple collaborating quad-rotor flying robots are useful in a broad range of applications, from surveillance with onboard cameras to reconfiguration of wireless networks. For these applications, it is often advantageous to have the robot team be a distributed system. In this thesis, an embedded controller capable of running distributed algorithms is presented for the quad-rotor flying robot. The robot platform is first characterized to help guide the design of the embedded control module. These modules are fabricated and tested on the quad-rotor flying robots in both indoor and outdoor environments. To propagate state estimates throughout the robot team, a location-based multi-hop algorithm is proposed. Network limitations, such as sub-optimal bandwidth and finite communication range, are implemented in hardware-in-the-loop simulations to determine system performance. A novel coverage algorithm for multiple hovering robots with downward facing cameras is then demonstrated on the embedded controller. The results from numerous indoor and outdoor experiments are discussed.