An automated four-DOF robot screw fastening using visual servo

Abstract

Techniques for determining and correcting alignment of an automated robot screw fastening using visual servoing strategy are developed and proposed in this paper. The new scheme controls continuously the screw aliment stages to avoid process caused failures. To achieve a desired screw aliment, a four DOF robot combined with visual servo feedback is designed to monitor and control. Two camera were equipped at the industrial robot, which is used to measure position error of screw aliment in real time. Tracking of screw aliment was employing image based on Canny edge enhancement and Hough transform algorithms. Visual features were transformed to the robot coordinate frames to allow designing manipulated a robot pose trajectory directly in the Cartesian. Results from experiments are presented which the automated screw insertion system using image processing feedback can be improve the efficient assembly performance.