Bayesian Map Learning in Dynamic Environments

We consider the problem of learning a grid-based map using a robot with noisy sensors and actuators. We compare two approaches: online EM, where the map is treated as a xed parameter, and Bayesian inference, where the map is a (matrix-valued) random variable. We show that even on a very simple example, online EM can get stuck in local minima, which causes the robot to get \\lost" and the resulting map to be useless. By contrast, the Bayesian approach, by maintaining multiple hypotheses, is much more robust. We then introduce a method for approximating the Bayesian solution, called Rao-Blackwellised particle ltering. We show that this approximation, when coupled with an active learning strategy, is fast but accurate. 1 Introduction The problem of getting mobile robots to autonomously learn maps of their environment has been widely studied (see e.g., [9] for a collection of recent papers). The basic diculty is that the robot must know exactly where it is (a problem called l...