ArcNav: Real-Time Curvature Based Map-Free Path Planning in Unstructured Environments for Unmanned Ground Vehicles

Autonomous navigation in unstructured environments remains a fundamental challenge in robotics, particularly for skid-steer unmanned ground vehicles (UGVs), which must operate in unpredictable and dynamically changing terrains. This paper introduces ArcNav, a novel realtime, map-free path planning algorithm that enables UGVs to maneuver efficiently in unstructured environments without relying on precomputed maps. Inspired by Dubins and ReedsShepp curves, ArcNav dynamically adjusts turning radii based on real-time environmental conditions, optimizing both trajectory smoothness and energy efficiency.