Multi-UAV agent-based coordination for persistent surveillance with dynamic priorities

A system to coordinate a team of unmanned aerial vehicles (UAV) applied for surveillance tasks is presented in this paper. Such scenario demands continuous monitoring of the region and take account of its priority areas which will require control interactions, both in mission planning and execution. The methodology approach is based on priority gradients combined with region partitioning, where the points of greatest interest are updated more frequently. The proposed system is based on the coordination of embedded agents using spatial coordination on strategical positioning and role exchange. The system architecture is composed by three modules, namely, the ground control station, for mission planning and control through user interaction; a quadrotors team, capable of navigation by inertial and satellite systems; and the control agents embedded in aircrafts, responsible for communications, task allocation and payload mission control. The preliminary results are validated using both computer simulation and a real flight with quadrotors. Telemetry data from the results were compared and it demonstrate that the proposed methodology converges to an effective solution.