Adaptive Distributed Coverage Control for Learning Spatial Phenomena in Unknown Environments

This study focuses on efficiently deploying a multi-robot system (MRS) to achieve optimal area coverage in the presence of an unknown event that requires monitoring. We introduce a control algorithm based on coverage to enable a team of robots to learn and estimate the spatial process of this region while ensuring good coverage. The robots’ observations are influenced by environmental noise. We use Gaussian processes (GP) to model the spatial process, and the multi-robot team optimally covers the estimated process. We evaluated the algorithm through simulations and real platform experiments.