SniffySquad: Patchiness-Aware Gas Source Localization with Multi-Robot Collaboration
Yuhan Cheng, Xuecheng Chen, Yixuan Yang, Haoyang Wang, Jingao Xu, Chaopeng Hong, Xiao-Ping Zhang, Yunhao Liu, Xinlei Chen
TL;DR
This work tackles real-world gas source localization by addressing patchy plume distributions that challenge traditional GSL methods. It introduces SniffySquad, which combines patchiness-aware active sensing based on Langevin diffusion with a potential-instructed, heterogeneous role-adaptation strategy for a team of mobile robots. The approach reframes GSL as an MCMC sampling problem, using surrogate optimization and Langevin dynamics to robustly converge to the source despite multimodal and noisy observations, while adaptively balancing exploration and exploitation across robots. Evaluations in field trials and simulations demonstrate substantial gains in success rate (20%+ ) and path efficiency (30%+), highlighting improved robustness to emission intensity, geometry, initial distance, and temperature settings, with practical impact for rapid, safe gas-leak mitigation.
Abstract
Gas source localization is pivotal for the rapid mitigation of gas leakage disasters, where mobile robots emerge as a promising solution. However, existing methods predominantly schedule robots' movements based on reactive stimuli or simplified gas plume models. These approaches typically excel in idealized, simulated environments but fall short in real-world gas environments characterized by their patchy distribution. In this work, we introduce SniffySquad, a multi-robot olfaction-based system designed to address the inherent patchiness in gas source localization. SniffySquad incorporates a patchiness-aware active sensing approach that enhances the quality of data collection and estimation. Moreover, it features an innovative collaborative role adaptation strategy to boost the efficiency of source-seeking endeavors. Extensive evaluations demonstrate that our system achieves an increase in the success rate by $20\%+$ and an improvement in path efficiency by $30\%+$, outperforming state-of-the-art gas source localization solutions.