A Micro-Computer Based Part/Tool Monitoring System for a Robotic Cell

Abstract

Abstract This paper describes the elements of a unique robotic cell integrated with a vision system. The cell consists of a PUMA 560 robot, a vision system, and a conveyor belt. The robot end effector is a drill used in a drilling operation. The vision system has a two fold function: monitoring tool wear, and recognizing parts passing over the conveyor belt. Both tool monitoring and part recognition are performed using the vision system. The vision system is PC-based and uses two CCD video cameras. One camera is mounted vertically overlooking the conveyor belt, and the other camera is mounted horizontally for tool wear monitoring. Algorithms have been developed that interface the vision system with the robot. Whenever a part reaches the view point of the camera, an image of the part is captured. A recognition algorithm has been developed that recognizes the part and signals the robot to perform a specific sequence of operation on that part. The part is then moved to the drilling station where the PUMA robot perform the operation sequence according to the stored process plan. After each part is manufactured, the robot moves the drill to the view point of the tool monitoring camera, where an image of the drill head is captured. This image is then processed with the developed tool monitoring algorithm and tool wear is identified.