Experiments on Space Robot Arm Path Planning Using the Sensors Database, Part II

Abstract

We propose three sensor-based pathplanning methods to simultaneousl y control the orientation of the main body and the arm position of a space robot using only arm motion. In our study, the database of relationships among the current sensor output, the robot motion, the change of sensor output and/or the sensor output after the motion is required. We then use breadth-first search, modified breadth-first search or A-search to determine the arm path. It is theoretically possible to operate the robot and gather data simultaneously as well as to easily modify the database. We confirmed the validity of our methods in experiments using a drop shaft, which provides ten seconds of microgravity during free fall. This experiment with a space robot using a drop shaft was the first trial of its kind in the world.