Seam tracking and welding bead geometry analysis for autonomous welding robot

Welding is a process recognized by the laborious work and hazardous work environment it takes place, but it is an important process in different industrial scenarios, like the shipbuilding industry. The use of robots has been increasing in recent years, reducing the human interference necessary for the process. This paper proposes a system for automated seam tracking and a geometric welding bead analysis systems based on passive monocular vision. Using an vision system attached to a linear welding robot, images of the gap between two plates are acquired, during the welding process in order to identify the groove's characteristics and perform a autonomous welding. These characteristics are used during the geometric weld bead analysis to detect geometric failures such as misalignment, lack or excess of fusion, among others.