Pose uncertainty in occupancy grids through Monte Carlo integration

Abstract

We consider the dense mapping problem where a mobile robot must combine range measurements into a consistent world-centric map. If the range sensor is mounted on the robot, as is usually the case, some form of SLAM must be implemented in order to estimate the robot's pose (position and orientation) at every time step. Such estimates are typically characterized by uncertainty, and for safe navigation it can be important for the map to reflect the extent of those uncertainties. We present a simple and computationally tractable means of incorporating the pose distribution returned by SLAM directly into an occupancy grid map. We also indicate how our mechanism for handling pose uncertainty fits naturally into an existing adaptive grid mapping algorithm, which is more memory efficient, and offer some improvements to that algorithm. We demonstrate the effectiveness and benefits of our approach using simulated as well as real-world data.