Robots to Navigate Hiking Trails

Summary

The episode explores the challenges of autonomous robot navigation on unstructured hiking trails, emphasizing the need to perceive, plan, and adapt to dynamic obstacles like fallen trees, mud, and erosion. Researchers combine LiDAR-based geometric terrain analysis with camera-driven semantic segmentation to create a fused traversability map that prioritizes trail staying while allowing safe off‑trail detours, using a hierarchical planner with a custom RRT* and local arc trajectories. Field tests on a Clearpath Husky at West Virginia University's arboretum demonstrate the system’s ability to reliably follow trails, handle hazards, and support applications such as trail monitoring, environmental data collection, and search‑and‑rescue. Future work aims to broaden the dataset across seasons, improve robustness to lighting and weather, and refine the balance between trail adherence and efficiency.