Fuzzy logic based mobile robot area filling with vision system for indoor environments

We present an area filling mobile robot system for indoor environment based on fuzzy logic and behavioral control using up-to-date environmental heterogeneous information perceived from range sensors and vision system. For gathering environmental information in the proximity of the robot and avoiding collisions with obstacles when navigating, infrared and ultrasonic sensors are utilized on the mobile robot system. Based on human expert knowledge, several basic behaviors for area filling are implemented using fuzzy inference. By exploiting straight lines features found on standard suspended ceilings, orientation error is corrected effectively while following a path during area filling. A finite state machine mechanism is used to control the behavior transitions and accomplish the area filling in the desired region. Experiments indicate the effectiveness of the system.