Adaptive linear robot control for tracking and grasping a dynamic object

Abstract

Versatile vision systems are being employed increasingly for automated visual feedback intelligent robotic control to perform complex manufacturing tasks especially for tracking and grasping dynamic objects a the conveyor by generating the optimum-tracking trajectory for the robot. Currently automated visual feedback robot systems use vision and robot system as separate tools. The best solution for these kind of problems still can not be implemented successfully with ease. Therefore, we tried to develop a new single adaptive prediction and execution algorithm for the picking and placing of dynamic objects in real time in a robot work-cell by integrating a CCD camera with frame grabber, artificial neural network, an optical flow technique and an industrial robot into a single application. The implementation of this proposed system is done only in one dimension because of the time constraint. After the robot learning process, it is shown that a KUKA robot is capable of adaptively tracking and intercepting dynamic objects at an optimal rendezvous point on the conveyor accurately in real time.