The Space Station Robot Manipulator System (SSRMS) Path Planning Program is a computer program that, in comparison with software developed previously for the same purpose, supports operations of faster and more complex robots. Two especially notable features of the program are that (1) it makes for ease of description of the work space of a remote manipulator or other robot and (2) it takes advantage of redundant degrees of freedom of the manipulator by finding manipulator-link paths that avoid both mathematical singularities and physical obstacles. The program can be applied not only to space-station manipulators and other robots but also to manipulators and other robots used in remediation of waste sites and dismantling nuclear facilities. With moderate modifications, the program could even be used in reconfigurable manufacturing operations.

In the original International Space Station application, there is a need to assure the safety of the crew and equipment by calculating a safe path and an optimal trajectory for the SSRMS, which has seven degrees of freedom (DOFs). In the absence of the present or a similar program, determining a safe path through a field of obstacles is difficult because the end-effector trajectory and the necessary manipulator-joint-angle trajectories must be taken into account manually; as a consequence, the chosen path may not be the optimal path. The SSRMS Path Planning Program automates the path-planning process. It searches for, and finds, the optimal trajectory in a matter of seconds.

The idea for the program came from R. V. Mayorga, who proposed a path-planning method for a manipulator that is redundant in the sense that the number of its DOFs exceeds the number of task elements. An analysis of the method was performed for a 3-DOF manipulator operating in a plane; then the analysis was extended to a 7-DOF manipulator operating in three-dimensional space. Finally, all the equations of the method specific to the SSRMS were formulated, together with some clever ways of incorporating them into a computer code. The SSRMS Path Planning program, written in ANSI-C, evolved from this final analysis.

The SSRMS Path Planning Program follows the artificial-potential-field approach to planning the path of the end effector of the remote manipulator. Real-time joint-angle trajectory planning for enabling manipulator links to avoid obstacles, involving the use of a null-space vector, is implemented in this program. All of the equations of kinematics and inverse kinematics specific to the SSRMS are also included.

In Mayorga's method, the goal configuration is represented by an attractive potential and obstacles by repulsive potentials. The determination of the manipulator joint-angle trajectory involves the inversion of a potential-field matrix, but this is subject to difficulties when the manipulator approaches singular configurations. With respect to singular configurations, the SSRMS Path Planning Program incorporates two improvements: (1) the manipulator trajectory is determined directly from the gradient of the potential field, so that there is no need to invert nearly singular matrices; (2) a global perspective is added by providing for the predetermination of optimal (shortest-path) goal trajectory for the end-effector position. The attractive potential is then based on an optimal trajectory that guides the end effector around obstacles in the shortest possible distance.

The program reads two input files. The first file describes the work space; the second file contains constants, integration parameters, the initial configuration, and the goal configuration. From these inputs, the program calculates a path through the work space, going from the initial configuration to the goal configuration along a path that avoids obstacles for the end effector and the manipulator links. By varying the values of some or all of the constants, one can change a path to incorporate a wider or narrower clearance of obstacles, change the time to complete the task, and/or otherwise alter the nature of the task. The joint-angle trajectories are computed by another program, "PLUMECHECK," on a Silicon Graphics workstation. PLUMECHECK can display animation of the manipulator arm.

The SSRMS Path Planning Program is an improvement over programs developed previously for the same purpose in the following respects:

  • Path planning has been automated. The joint-angle trajectories for any path can be found in seconds. Therefore, many different scenarios can be tried to find an optimal path.
  • It is easy to describe the work space, including obstacles, forbidden regions, and other important features of the work space.
  • The redundancy of the SSRMS is not a disadvantage for this program; on the contrary, it is an advantage when there is a need to find singularity-free paths and avoid obstacles.

The SSRMS Path Planning Program has much to recommend it. Its potential field-based approach to avoidance of collisions by kinematically redundant manipulators has already been demonstrated in research laboratories. This program is being evaluated with respect to commercial applicability.

This work was done by Glenn V. Webb of Rockwell Space Operations for Johnson Space Center. MSC-22751