A Framework for Human-Aware Navigation in Quadruped Robots using 3D LiDAR Sensing

Molnár, Kamilla and Rózsa, Zoltán and Szirányi, Tamás (2026) A Framework for Human-Aware Navigation in Quadruped Robots using 3D LiDAR Sensing. In: Intelligent Robotics FAIR 2026 (IntRob '26). (In Press)

Text ACM.pdf - Accepted Version Restricted to Repository staff only until 31 December 2027. Available under License Creative Commons Attribution. Download (1MB)

Abstract

The paper presents a framework for real-time, socially aware human overtaking using a Boston Dynamics Spot quadruped robot equipped with a Livox MID-360 3D LiDAR. The core of the system is an adaptation of a high-performance object tracking and trajectory prediction pipeline originally developed for autonomous vehicles. We propose a predictive goal generation method that calculates an optimal overtaking pose based on the tracked human and a dynamic lookahead distance. To ensure smooth and stable robot motion, an Exponential Moving Average (EMA) filter is applied to the generated goal coordinates, mitigating the jitter inherent in real-time point cloud processing. Experimental results in real-world scenarios demonstrate over 90\% success rate during uniform walking, while also revealing the challenges of high-dynamic movements. The results underscore the viability of porting automotive-grade perception stacks to legged robotic platforms for enhanced human-robot interaction.

Actions (login required)

View Item