The development of a low cost autonomous robotic arc welding system

A significant challenge for robotic welding to be widely adopted is the time taken to program robot paths for new parts. While computer vision can be used to detect and locate weld joints, due to the large number of possible joint configurations, work piece materials and environmental impacts such as lighting conditions, autonomous weld joint detection and localisation remains a significant challenge. This paper introduces an autonomous robotic arc welding system that is capable of detecting realistic weld joints and calculating their position in the robot workspace with minimal human interaction. The proposed method is capable of detecting and localising butt and fillet weld joints regardless of base material, surface finish or imperfections. The welding results show that the proposed method is capable of producing high quality weld paths suitable for industrial applications.