Detection of assembly variations for automatic program adaptation in robotic welding systems

Abstract

This paper proposes a novel approach for detection of assembly variations to adapt the programming of a robotic welding system. A matching process is performed between the different parts of an assembly CAD model and a measured point cloud using an Iterative Closest Point algorithm to determine assembly variations. Experimental validation is performed with an industrial robot equipped with stereo camera and welding gun. A method for accurate calibration of sensor and welding gun is presented as it influences the quality of the welding process. Experimental results show an increased weld seam quality and the robustness of the proposed approach in industrial applications.