Robot Path Planning for Dimensional Measurement in Automotive Manufacturing

This paper addresses the robot path planning problem in our effort to develop a fully automated dimensional measurement system using an eye-in-hand robotic manipulator. First, the CAD-based vision sensor planning system developed in our lab is briefly introduced; it uses both the CAD model and the camera model to plan camera viewpoints. The planning system employs a decomposition-based approach to generate camera viewpoints that satisfy given task constraints. Second, to improve the efficiency of the eye-in-hand robot inspection system, robot path planning is studied, which is the focus of this paper. This problem is rendered as a Traveling Salesman Problem (TSP). A new hierarchical approach is developed to solve the TSP into its suboptimality. Instead of solving a large size TSP, this approach utilizes the clustering nature of the viewpoints and converts the TSP into a clustered Traveling Salesman Problem (CTSP). A new algorithm, which favors the intergroup paths, is proposed to solve the CTSP quickly. Performance of the new algorithm is analyzed. It is shown that instead of a fixed performance ratio as reported in some existing work, a constant bound can be achieved which is related to the diameter of the clusters. Experimental results demonstrate the effectiveness of the robot motion planning system. The proposed path planning approach can obtain sub-optimal solutions quickly for many large scale TSPs, which are common problems in many robotic applications.