Exploiting Inaccurate A Priori Knowledge in Robot Exploration

Exploration is a task in which autonomous mobile robots incrementally discover features of interest in initially unknown environments. Usually, robots follow exploration strategies to select the next best locations to reach in partially explored environments. Most of the current exploration strategies ignore prior knowledge about the enviroments to explore that, in some practical cases, could be available. In this paper, we present a method that includes a priori knowledge in an exploration strategy for a mobile robot. In particular, our exploration strategy selects the next best locations the robot should reach by exploiting the knowledge of the floor plan of the indoor environment that is being explored. Although the floor plan can be inaccurate (e.g., it typically does not include furniture and could represent a topology that does not fully match with that of the actual environment), we experimentally show, both in simulation and with real robots, that knowing the floor plan improves the exploration performance under a wide range of conditions.