High-speed control and navigation for quadrupedal robots on complex and discrete terrain.
🤖 AI Summary
To address the challenge of high-speed autonomous navigation and dynamic control for quadrupedal robots on complex, discontinuous terrains, this work proposes a hierarchical planning–tracking architecture. At the planning level, we introduce a novel three-stage framework—sampling, heuristic filtering, and physics-based rolling evaluation—that synergistically integrates model-based optimization with neural-network-accelerated computation. At the tracking level, we propose an adversarial generative objective distribution training strategy to enhance foot placement accuracy and robustness. The method is validated on our custom-built robot Raibo, achieving autonomous capabilities including vertical wall running, 1.3 m obstacle jumping, stepping-stone traversal at 4 m/s, and full autonomy on 30° slopes, staircases, and multi-scale unstructured obstacles. These results significantly extend the locomotion capability frontier of quadrupedal robots in geometrically complex, discrete, and non-structured environments.
Application Category
📝 Abstract
High-speed legged navigation in discrete and geometrically complex environments is a challenging task because of the high-degree-of-freedom dynamics and long-horizon, nonconvex nature of the optimization problem. In this work, we propose a hierarchical navigation pipeline for legged robots that can traverse such environments at high speed. The proposed pipeline consists of a planner and tracker module. The planner module finds physically feasible foothold plans by sampling-based optimization with fast sequential filtering using heuristics and a neural network. Subsequently, rollouts are performed in a physics simulation to identify the best foothold plan regarding the engineered cost function and to confirm its physical consistency. This hierarchical planning module is computationally efficient and physically accurate at the same time. The tracker aims to accurately step on the target footholds from the planning module. During the training stage, the foothold target distribution is given by a generative model that is trained competitively with the tracker. This process ensures that the tracker is trained in an environment with the desired difficulty. The resulting tracker can overcome terrains that are more difficult than what the previous methods could manage. We demonstrated our approach using Raibo, our in-house dynamic quadruped robot. The results were dynamic and agile motions: Raibo is capable of running on vertical walls, jumping a 1.3-meter gap, running over stepping stones at 4 meters per second, and autonomously navigating on terrains full of 30° ramps, stairs, and boxes of various sizes.
Problem
Research questions and friction points this paper is trying to address.
High-speed navigation for quadruped robots on complex terrain
Optimizing foothold plans in nonconvex, long-horizon environments
Achieving dynamic motions like wall-running and large gap jumps
Innovation
Methods, ideas, or system contributions that make the work stand out.
Hierarchical navigation pipeline for legged robots
Sampling-based optimization with neural network
Generative model for foothold target distribution
🔎 Similar Papers
No similar papers found.
