Vision-directed path planning, navigation, and control for an autonomous mobile robot

This paper presents a model and sensor based path planning, navigation and control system for an autonomous mobile robot (AMR) operating in a known laboratory environment. The goal of this research is to enable the AMR to use its on-board computer vision to: (a) locate its own position in the laboratory environment; (b) plan its path; (c) and navigate itself along the planned path. To determine the position and the orientation of the AMR before and during navigation, the vision system has to first recognize and locate the known landmarks, such as doors and columns, in the laboratory. Once the AMR relates its own position and orientation with the world environment, it is able to plan a path to reach a certain prescribed destination. In order to achieve on-line visual feedback, an autonomous target (landmark) acquisition, recognition, and tracking scheme is used. The AMR system is designed and developed to support flexible manufacturing in general, and surveillance and transporting materials in a hazardous environment as well as an autonomous space robotics project funded by MRCO and the Canadian Space Program related to the Freedom Space Station.