Online map building for autonomous mobile robots by fusing laser and sonar data

This paper presents a practical method of online building maps automatically by fusing laser and sonar data. Sensor modeling is used to translate range data into grid status information of the local map, and Bayes´ rule is employed to update the global map with this local map. For online mapping and pose estimation, the problem of simultaneous localization and mapping is solved by applying scan matching using laser data. A selective method is proposed to fuse laser and sonar data for better obstacle detection and mapping. Finally, navigation using a fuzzy controller is implemented based on fusion of sensor data, which enables the algorithm to build maps autonomously.