Communication and path planning strategies of a robotic coverage operation

In this paper, we consider the scenario where a mobile robot is tasked with visiting a number of Points Of Interest (POIs) in a workspace, gathering their generated information bits, and finding a good spot for communication, in order to successfully transmit all the collected bits to a remote station. The goal of the robot is to minimize its total energy consumption, which includes both communication and motion energy costs, while operating in a realistic communication environment and under a given time budget. We show how to find the optimum communication point and plan the optimum path to cover all the POIs by posing the problem as a Mixed Integer Linear Program (MILP) and by using probabilistic metrics to assess the link quality over the workspace. More specifically, we prove a number of properties for the optimum communication location and transmission strategy at this location, as a function of motion and communication parameters and channel prediction quality. Furthermore, we show that if the communication demand is low, the optimum path becomes the minimum-length path on the POIs asymptotically. On the other hand, if the communication demand is high, given enough time, the optimum path becomes the minimum-length path on the set of POIs and the point with the best predicted channel quality in the workspace. Finally, the performance of our framework is verified in a simulation environment.