An image-guided mobile robotic welding system for SMAW repair processes

Conventionally, shield metal arc welding (SMAW) has been performed by manual operator, and hence it is not easy to apply in hazardous areas such as in nuclear power plants. An image-guided mobile robotic welding system is developed in this study for the automation of the SMAW process. The position of cracks can also be determined by this image-guided mobile robotic welding system. We installed two CCD cameras, a five-axis SCORBOT-ER VII robot and SMAW equipment on a mobile vehicle. The cameras are installed in parallel and can be used to detect the position of any cracks beforehand; the robot is then driven at a constant welding speed along the welding path. Based on the binocular images captured by the CCD cameras, an on-line algorithm is used to compare the disparities between the two images and to obtain the welding paths for the cracks. Since the electrode is consumed during the SMAW process, the robotic joint trajectories need to be modified to keep the end of the electrode on tracking on the desired welding path. The control system determines the feeding velocity of the electrode from the measured welding current and joint positions and modifies the joint trajectories correspondingly. The experimental results show that our system can effectively perform SMAW operations resulting in good quality welding. Also, the variation of the welding current in our system is much smaller than that operated by even senior technicians.