Vision-Based Reactive Planning and Control of Quadruped Robots in Unstructured Dynamic Environments

TL;DR

This paper introduces a vision-based reactive planning and control framework for quadruped robots that enables real-time adaptation in unstructured, dynamic environments by combining offline pre-planning with online reactive adjustments.

Contribution

It presents a novel V-RPC system integrating pre-planned trajectories with real-time visual feedback for adaptive quadruped robot navigation.

Findings

Effective in dynamic, unstructured environments

Real-time trajectory adjustment based on visual perception

Improved adaptability over static planning methods

Abstract

Quadruped robots have received increasing attention for the past few years. However, existing works primarily focus on static environments or assume the robot has full observations of the environment. This limits their practical applications since real-world environments are often dynamic and partially observable. To tackle these issues, vision-based reactive planning and control (V-RPC) is developed in this work. The V-RPC comprises two modules: offline pre-planning and online reactive planning. The pre-planning phase generates a reference trajectory over continuous workspace via sampling-based methods using prior environmental knowledge, given an LTL specification. The online reactive module dynamically adjusts the reference trajectory and control based on the robot's real-time visual perception to adapt to environmental changes.