Research Analyzer
← Back ICRA 2026

APREBot: Active Perception System for Reflexive Evasion Robot

ZIHAO XU, Kuankuan Sima, Junhao Deng, Zixuan Zhuang, Chunzheng Wang, Ce Hao, Jin Song Dong

PDF

AI summary

Key figure (auto-extracted from paper)
Combining LiDAR’s omnidirectional scanning with camera-based active focusing significantly improves dynamic obstacle avoidance safety and agility for quadruped robots under strict reaction time constraints.
Quadruped robots dynamic obstacle avoidance active perception LiDAR-camera fusion reflexive evasion Sim2Real

Problem

Single-sensor systems fail to provide both global coverage and high-resolution detail needed for legged robots to reliably detect and evade dynamic obstacles within strict reaction time windows.

Approach

APREBot uses a three-stage pipeline where LiDAR continuously monitors threats, the robot actively turns to focus an RGB-D camera on the most dangerous obstacle, and a unified threat metric drives adaptive avoidance behaviors.

Key results

  • Active hierarchical perception framework for quadruped dynamic obstacle avoidance
  • Threat-aware perception mechanisms combining LiDAR scanning, camera tracking, and short-horizon prediction
  • Extensive Sim2Real validation showing consistent safety, efficiency, and robustness gains over baselines
  • Adaptive threat-aware avoidance blending navigation retreat with reflexive evasion gaits

Why it matters

Enables safer, more agile autonomous navigation for legged robots in dynamic, safety-critical environments where rapid and reliable obstacle avoidance is essential.

Abstract

Reliable onboard perception is critical for quadruped robots navigating dynamic environments, where obstacles can emerge from any direction under strict reaction time constraints. Single-sensor systems face inherent limitations: LiDAR provides omnidirectional coverage but lacks rich texture information, while cameras capture high-resolution detail but suffer from restricted field of view. We introduce APREBot (Active Perception System for Reflexive Evasion Robot), a novel framework that integrates reflexive evasion with active hierarchical perception. APREBot strategically combines LiDAR- based omnidirectional scanning with camera-based active focusing, achieving comprehensive environmental awareness essential for agile obstacle avoidance in quadruped robots. We validate APREBot through extensive Sim2Real experiments on a quadruped platform, evaluating diverse obstacle types, trajectories, and approach directions. Our results demonstrate substantial improvements over strong baselines in both safety metrics and operational efficiency, highlighting APREBot’s potential for dependable autonomy in safety-critical scenarios. Paper homepage: https://aprebot-2026.github.io/.

Index terms

Legged Robots Collision Avoidance

Related papers