Scheduling drones to cover outdoor events

Task allocation is an important aspect of many multi-robot systems. In this paper, we consider a new task allocation problem that appears in multi-robot aerial cinematography. The main goal is to distribute a set of tasks (shooting actions) among the team members optimizing a certain objective function. The tasks are given as sequences of waypoints with associated time intervals (scenes). We prove that the task allocation problem maximizing the total filmed time by $k$ aerial robots (drones) can be solved in polynomial time when the drones do not require battery recharge. We also consider the problem in which the drones have a limited battery endurance and must periodically go to a static base station. For this version, we show how to solve the problem in polynomial time when only one drone is available.