Modeling Deformable Linear Objects for Autonomous Robotic Outfitting of Lunar Surface Systems

Abstract

The construction of in-space facilities is a significant capability for the establishment of a longterm human presence in space. Autonomous robotic technologies are a critical tool to enabling the construction and maintenance of such permanent facilities. This paper evaluates the outfitting portion of the construction process, focusing on modeling required for robotic manipulation of cable-like objects, referred to as deformable linear objects (DLOs). DLOs contain a high degree of variability, which makes accurate real-time autonomous operations a difficult task. Different modeling methods for DLOs such as discrete mass–spring systems and Cosserat elastic rod models range in problem complexity and accuracy, a trade-off that must be understood to fully realize autonomous cable routing technologies. This paper validates different DLO models through simulation and a hardware experiment, highlighting the size of the state space and accuracy of different approaches. The development of DLO manipulation models for outfitting enables an autonomous architecture for the construction of lunar surface systems.