Skid-Steering Telenavigation with Extended Predictive Polygon Method

Abstract

The future moon missions will support extensive exploration activities, including remote operation between astronaut missions. The recent discovery of a lunar cave opens exciting possibilities for future exploration, as these natural formations could provide crucial protection for astronauts from the Moon's harsh surface conditions and harmful solar radiation. Despite significant advances in autonomous driving systems, their application in lunar exploration is constrained by uncertainties in environmental reconstruction and incomplete scene analysis due to occlusions. As a result, teleoperation emerges as a vital fallback solution for such missions. However, one of the primary challenges in bilateral teleoperation is the communication delays, which can destabilize and impair its performance. To address this, the current work extends the position-force predictive polygon method to skid-steering wheeled mobile robots (WMRs). By incorporating an intuitive haptic interface and a passive position-scaling mechanism, the system allows the human operator to adjust their commanded curvature input, ensuring it better matches the turning radius of skid-steering WMR. Experimental validation demonstrating the passivity of the proposed framework, along with discussion on how position scaling affects the system's overall performance is presented.