GPF-BG: A Hierarchical Vision-Based Planning Framework for Safe Quadrupedal Navigation

Safe quadrupedal navigation through unknown environments is a challenging problem. This paper proposes a hierarchical vision-based planning framework (GPF-BG) integrating our previous Global Path Follower (GPF) navigation system and a gap-based local planner using Bézier curves, so called $B$ézier Gap (BG). This BG-based trajectory synthesis can generate smooth trajectories and guarantee safety for point-mass robots. With a gap analysis extension based on non-point, rectangular geometry, safety is guaranteed for an idealized quadrupedal motion model and significantly improved for an actual quadrupedal robot model. Stabilized perception space improves performance under oscillatory internal body motions that impact sensing. Simulation-based and real experiments under different benchmarking configurations test safe navigation performance. GPF-BG has the best safety outcomes across all experiments.

I. Introduction

Quadrupedal robots have demonstrated superior terrain traversability compared to traditional wheeled robots [1]–[3]. Significant progress has been made during the past few years to improve the robustness and agility of legged locomotion control [4]-[10], which enables the incorporation of exteroceptive sensors for autonomous navigation (e.g. Fig. 1). Taking into account legged robot morphology, prior navigation works have been mostly focused on rough terrain traversability [2], [11]–[15], multi-modal planning [2], [16]–[19], and multi-robot exploration [20], [21]. However, navigation safety and obstacle avoidance have not been formally analyzed, and the tested scenarios are limited to less dense environments. Safe navigation through unknown or partially unknown environments for quadrupeds is critical to the field deployment of legged robots yet remains under explored.