Robots navigate gas leaks using diffusion and olfactory scene graphs
SmellDiffusion: Diffusion-Based Quadruped Navigation with Olfactory Scene Graphs
Robotics
Summary
Robots sent to find gas leaks need to identify the type of gas, find the leak's source, and get to it safely. The authors created SmellDiffusion, a system that helps robots estimate gas sources and plan paths using a new way of representing gas areas called olfactory scene graphs. Their method combines classical navigation and diffusion models to improve accuracy and speed compared to older approaches. Tests in simulation and robot runs showed it reduces errors in source location and navigates efficiently.
What this means in practice
- •For robotics teams: Guide quadruped robots to gas leak sources quickly and accurately using combined olfactory and diffusion-based planning.
- •For industrial safety engineers: Deploy autonomous robots to navigate hazardous gas leak environments for inspection and monitoring with improved source localization.
Authors
Faith Ogunwoye, Iana Zhura, Hajira Amjad, Timofei Kozlov, Didar Seyidov, Dmitrii Plotnikov, Fedor Fedorov, Dzmitry Tsetserukou
Abstract
A robot sent to a named gas leak must preserve gas identity, estimate the source, and navigate to the resulting goal. We present SmellDiffusion, a simulation pipeline that represents species-specific gas zones in an open-vocabulary olfactory scene graph and shares the selected goal between classical and diffusion planners. Its key components are a peak-local geometric gate for selective source correction and diffusion-based, gas-guided trajectory generation. Among 424 unique source-wind configurations in solved flow, 28 have a concentration peak displaced more than 0.5m from the source. A source-independent geometric gate, calibrated only on the training split and evaluated at the observed peak, detects 9 of 10 held-out displacements at 0.64 precision. Gating a precomputed forward-matching correction reduces mean error on the displaced cases from 1.468m to 0.592m (60%), using matching for only 14/204 cases. All-case mean error falls from 0.205m to 0.180m. All planners receive the same scene-graph source estimate as their goal. In a controlled comparison, best-of-ten diffusion achieves mean gas exposure comparable to gas-guided A* (0.0476 versus 0.0455). A single diffusion proposal takes 41.7ms, compared with 72.3ms for gas-guided A*, although best-of-ten sequential sampling increases total runtime. Plain A* also reaches the same goal and remains the fastest and shortest-path method. Six matched Gazebo runs give mean robot-to-source errors of 0.39m for A* and 0.31m for diffusion.