Concise description

SICK image processing systems integrated into welding robots guarantee the reliable recognition, tracking and checking of weld seams.

Task

WISE WELDING develops and constructs a whole range of special machines, including machines for the production of stainless steel plant components for the pharmaceutical, chemical and food-processing industry. In this particular case the business had been requested to optimize the welding process of a tungsten inert gas welding robot. Since the parts to be connected could have tolerances, the welding path in the robot control had to be adjusted by manual determination of the actual geometries. This procedure was monotonous, prone to errors and time-consuming. This impaired the availability of the robots and the level of process quality. WISE WELDING therefore searched for a possibility of automatically detecting the welding path even with tolerances of up to +/- 10 mm and hence of controlling the robot more precisely. Another goal was to be able to check the quality of the weld seam. The quality demanded included an accuracy of 0.1 mm and reliability even with reflection differences in the materials. The challenges in the context of sensors and tungsten inert gas welding are electromagnetic fields and strong extraneous light influences. The solution: Vision sensor components and solution support from SICK.

Implementation

Recognition of weld seams and robot visual guidance with a camera system increase the productivity and quality

Recognition and automatic adjustment of the welding path as well as quality inspection of the weld- seams is done with image processing systems (Ranger E), which WISE WELDING, supported by SICK, has integrated directly into the robots. The camera solution offers maximum measuring and transmission speeds as well as an extremely high accuracy for recognition of even the smallest details. Prior to welding parts, a master part is taught-in. Then all other parts can be processed automatically. To this end, the camera transmits an image of the parts to an analyzing software developed by WISE WELDING prior to welding. This software recognizes where the weld seam should be and transmits the control parameters to the robot control which positions and guides the welding head accordingly. In this, neither the high-frequency electromagnetic fields of the tungsten inert gas welding process nor the extremely bright light effects affect the precise robot guidance and sub-millimeter precision control of the weld seams.

Customer benefits

For this demanding task SICK was able to offer WISE WELDING a perfectly suited camera solution from its extensive image processing portfolio. Thanks to the comprehensive support provided by SICK, which ranged from application analysis over commissioning right up to training, WISE WELDING was able to integrate the camera solution into its welding robots in the best possible way. Due to the fact that a master part is taught-in only once and the optimum welding path is now determined much faster, the robots' availability and productivity is significantly increased and a faster return on investment is achieved. The high-precision weld seam inspection not only ensures consistent process quality, but also better surface quality of the parts around the welded joints.