Overview

Off-road robot navigation helps vehicles and legged robots move safely where roads, maps, and reliable infrastructure are absent, from forests and mines to planetary terrain. Systems combine camera, LiDAR, radar, satellite, and proprioceptive data to estimate terrain costs or mobility risk, then use methods such as A*, graph search, model predictive control, and learned trajectory policies to choose routes. Research evaluates these approaches in simulation and field settings, including forest, desert, orchard, and planetary-analog tests, with growing attention to long-horizon exploration, changing terrain, and low-power onboard operation. A central challenge is balancing reliable safety under unfamiliar or partially observed conditions with the generalization and computational efficiency needed for real-time autonomy.

Latest papers

CardsList

62 papers12 papers in the last four weeks, up 50% on the four weeks before. 0.1% of all new papers.

  1. Path Planning on Multi-level Point Cloud with a Weighted Traversability Graph

    Apr 30, 2025Yujie Tang, Quan Li, Hao Geng +3Motion PlanningOff-Road Robot Navigation

  2. AnyNav: Visual Neuro-Symbolic Friction Learning for Off-road Navigation

    Jan 22, 2025Taimeng Fu, Zitong Zhan, Zhipeng Zhao +7Off-Road Robot NavigationMotion Planning