Single Robot Exploration: Simultaneous Localization and Uncertainty Reduction on Maps (SLURM)

Abstract

During exploration of an unknown environment by a single robot, the robot is driven by two conflicting goals: to explore as fast as possible, and to produce the most accurate map. While fast exploration necessitates minimizing traversal of already mapped territory, accurate mapping requires that the robot passes over previously explored areas to reduce the localization and map uncertainty. This problem has been labelled as exploration versus exploitation. In this paper the problem of mapping a camera sensor network by a mobile robot has been used to demonstrate the effect that different exploration strategies have on uncertainty and speed of exploration. Simulation results using a realistic noise model are presented for different environments and for different strategies.