Monte Carlo SLAM method for a Small Mobile Robot with Short-Range Sensors

K. Yamada, Y. Nakada, and T. Matsumoto (Japan)


Autonomous Mobile Robot, Simultaneous Localization And Mapping (SLAM), Rao-Blackwellized Sequential Monte Carlo, Hidden Variable, Grid Map


We propose a novel extension of the grid-based Monte Carlo SLAM approach for a small mobile robot with short range distance sensors. The proposed approach consid ers probabilistic hidden variables in the model, instead of the noisy local maps generated in deterministic conversion processes. These hidden variables reduce the effect of un wanted sensor noise in the mapping processes. To evaluate the proposed approach, it is tested against a numerical ex periment based on a simulator of a small mobile robot, Khepella II, with various sensor noise levels.

Important Links:

Go Back