Nowadays, Unmanned Aerial Vehicles (UAVs), such as drones, are becoming very popular because they can access difficult locations for humans. Traditionally, UAVs are remotely controlled by a human operator, which implies establishing a communication with the UAV in order to send the human operator's orders in real time and also transmit the environment visualized by the UAV. This problem becomes even worse when more than one UAV is considered. Many solutions have been considered lately to tackle the automation of the UAVs control, most of them related to deep learning and reinforcement learning techniques. Specially, those facing the problem of managing the routes of a fleet of UAVs to visit different locations.
In this project, an online mission planning system based on deep reinforcement learning and graph neural networks has been developed, which is capable of managing the routes of a fleet of UAVs in real time and in an optimal way to achieve visiting as many locations as possible with a region-sharing strategy. The cooperation strategy is specially designed for non-communications environments among the UAVs (or radio silence) during the mission execution.