Motion Planning for Relocatable Robots Performing On-Orbit Locomotion and Manipulation Tasks

Abstract

In-space assembly is a key technology for the future development of large infrastructures in space, from space stations and telescopes, to solar power plants or planetary bases. Such structures are much larger than cargo areas in current launchers, therefore they must be sent in separate pieces that are assembled in situ, typically using relocatable robotic manipulators. The efficient exploitation of the locomotion and manipulation (loco-manipulation) abilities for such robotic systems requires suitable planning tools. In this paper, we present a motion planning approach for exploiting loco-manipulation abilities of self-relocatable space robots, assuming that they move over specific interconnects that provide the required mechanical, power and data connectivity. The proposed approach consists of three planning layers: a high-level planning for obtaining the contact sequence, a low-level planning for the joint trajectories, and a validation layer. The motion planner provides plans for single locomotion and manipulation tasks, as well as combined loco-manipulation tasks. The approach is illustrated with examples for two robotic systems: MOSAR-WM, a relocatable walking manipulator, and a multi-arm robot (MAR) equipped with two arms attached to a central torso