Autonomous Robotic Control of the Deburring Process for Die-Casting Parts Using Vision-Based Burr Detection Information

This paper proposes an autonomous robotic deburring method for the deburring process of die-casting parts using vision-based burr detection information. A 3D laser scanner is used to acquire point cloud data of a die-cast workpiece, and 2D shape information is generated through image processing. The generated shape information is compared with CAD design data to identify burr regions and distributions. A speed-controlled robotic deburring path is automatically generated based on the detected burr information and a lookup table-based speed control scheme is applied. The effectiveness of the proposed method is validated through experiments using a robotic deburring system.