In this thesis, we discuss the architecture, H/W and S/W of the sensor-based Intelligent Mobile Robot, CAIR-I and propose noise-insensitive ultrasound sensing technique for reliable obstacle avoidance and vision technique based on 2-phase algorithm for real-time object recognition and tracking. We also propose a method based on motion stereo by which a mobile robot can determine its position and orientation in the space.
Experimental results are provided using the ultrasound sensors, video camera, and processors established in the CAIR-I.