Proximity Detection and Collision Avoidance for Space Station Freedom Robot Manipulators

Abstract

The Teleoperated Robotics Section at the NASAIJohnson Space Center i s currently evaluating a world-model-based Proximity Detection and Collision Avoidance (PDCA) system for use on Space Station Freedom (SSF). An analysis of test data from this study will determine PDCA'S impact on SSF teleoperated robotic task performance under various work site viewing conditions. This PDCA system is an emulation of a concept proposed by the Canadian Space Agency (CSA) . The system's hierarchical world model is generated automatically from a graphical database that is also used to simulate the SSF environment during testing. This world model includes all the fixed obstacles, modelled as convex polytopes, in the robot's workspace. Line segments with radial shells, referred to as clyspheres, approximate all moving robot and payload parts. A University of Michigan algorithm that calculates distances between arbitrary convex polytopes provides proximity data for use in collision avoidance. This multipart evaluation i s in its simulation stage, and testing is under way using detailed SSF graphical models and robot kinematic models. Future work is expected to include implementation i nto a previously developed hardware test bed.