Computer-Vision-Based Autonomous Robotic Part Repairing∗

In this paper, a computer-vision(CV)-based robotic autonomous part repairing system is developed. Robotic autonomous part repair is needed to retrofit high-value parts as robotic-based machining system offers advantages in manufacturing flexibility, adaptability, precision, and low cost. Although CV-based robotic manipulation has been explored for applications, challenges emerge in CV-based robotic machining applications due to the more stringent precision and accuracy in part identification, the inevitable eccentric misalignment in data acquisition, the artifacts of the laser-scanned data, and the needs for simultaneous force and path tracking. The contribution of this work is the development of an experimental-based approach to quantify and compensate for the eccentric misalignment, and then, identify and quantify the defect on the part. We illustrate the function and performance of the CV-based robotic autonomous repairing through experiment.