Vision-based adjusting of a digital model to real-world conditions for wire insertion tasks

Abstract In small and medium-sized enterprises, the wiring of control cabinets is a very time-consuming process as it is mostly performed manually. Thus, a novel automated approach for wiring control cabinets with batch size 1 is provided. The task is executed by a lightweight robot and based on an ECAD model of a wired cabinet. Therefore, a method is described to export digital data from ECAD and adjust the digital information to real-world conditions. To detect the approximate real-world poses of the components, a 2D image from a stationary visual sensor is used for CAD-based template matching by normalized cross-correlation. A high accurate visual 3D sensor that is mounted on the robot end-effector is moved to the approximate poses of the component ports to measure the exact poses. The results prove the functionality of the developed method, which allows to position the robotic end-effector based on ECAD data, which is the prerequisite to measure the component port with high accuracy using the 3D sensor.