Navigation-Guidance-Based Robotic Interception of Moving Objects in Industrial Settings

Abstract

An important task for autonomous industrial robotic systems is the interception of moving objects. In order to achieve this objective, an on-line robot-motion planning technique that utilizes real-time sensory feedback about the object's motion is needed. In this paper, an Ideal Proportional Navigation Guidance (IPNG) based technique is utilized for on-line robot-motion planning. One must note, however, that navigation-guidance techniques were originally developed for bringing the interceptor into a collision course with (hostile) airborne targets. Therefore, in our case, a conventional tracking technique must be utilized as a subsequent phase to an initial IPNG-based robot-motion planning phase in order to ensure smooth interception. The implementation of the hybrid scheme in industrial settings, where one may not have access to the robot's dynamic model nor to the joints' controllers, is discussed. Real-time experimental results using an industrial robot and a computer-vision system are presented, confirming the (interception-time) superiority of our proposed scheme over conventional tracking techniques.