Serpentine Manipulator Planning and Control for NASA Space-Shuttle Payload Servicing

Abstract

The use of a highly-redundant manipulator, so-called 'serpentine' manipulator, is proposed as a solution for servicing space-payloads during the payload installation process before launch on the space-shuttle. The staging of the system would be inside a large cleanroom area, the Payload Changeout Room (PCR), which sits on the Rotating Service Structure (RSS), allowing it to be swung into mated contact with the space-shuttle as it sits on the launchpad at the John F. Kennedy Space Center in Cape Canaveral, Florida. This report is not so much concerned with the design and implementation issues associated with a serpentine manipulator, but rather with the planning, control and user-interface issues. We thus present a brief introduction to the actual application environment and its restrictions, followed by the theoretical background and implementation issues of planning and control algorithms developed specifically for a serpentine manipulator operating within the confined spaces of the PCR and the space-shuttle cargo-bay. The approach used to generate a continuous path is to develop and use a minimizing continuous energy curve for the manipulator, and then fit the discretized link-sections to that curve. The operator is then able to modify this curve directly, allowing for shaping of the serpentine, and for the placement of obstacles that can then be avoided. This algorithm can be implemented in real-time, and curve-parameters can also be adjusted such as flexibility and curvature of the curve. An A* method is used to search for the shortest distance between two goal points in order to construct the curve. In order to avoid obstacles (real and artificial), available from a database or which can be placed by the operator using the operator interface display, we use an obstacle potential field FIRAS function. This report outlines the NASA applications background and the technical research and implementation issues of the different planning and control algorithms. The entire system was simulated on an SGI IRIS, and a video-tape has been made of the different operational modes of the planning and operator control displays, showcasing the different capabilities of the serpentine manipulator motion planning and control system. (33 pages)