Humanoid Path Planning over Rough Terrain using Traversability Assessment

TL;DR

This paper introduces a two-stage planning framework for humanoid robots to navigate rough terrain by assessing traversability and smoothing paths, demonstrated on a Boston Dynamics Atlas robot.

Contribution

A novel terrain-aware path planning method that integrates traversability assessment with path smoothing, specifically designed for humanoid robots over challenging terrains.

Findings

Effective in simulation across various terrains

Robust paths that handle small gaps in traversability

Successful real-world demonstration on Atlas robot

Abstract

We present a planning framework designed for humanoid navigation over challenging terrain. This framework is designed to plan a traversable, smooth, and collision-free path using a 2.5D height map. The planner is comprised of two stages. The first stage consists of an A* planner which reasons about traversability using terrain features. A novel cost function is presented which encodes the bipedal gait directly into the graph structure, enabling natural paths that are robust to small gaps in traversability. The second stage is an optimization framework which smooths the path while further improving traversability. The planner is tested on a variety of terrains in simulation and is combined with a footstep planner and balance controller to create an integrated navigation framework, which is demonstrated on a DRC Boston Dynamics Atlas robot.