Abstract
Volcanic gas emissions are key precursors of eruptive activity. Yet, obtaining accurate near-surface measurements remains hazardous and logistically challenging, motivating the need for autonomous solutions. Limited mobility in rough volcanic terrain has prevented wheeled systems from performing reliable in situ gas measurements, reducing their usefulness as sensing platforms. We present a legged robotic system for autonomous volcanic gas analysis, utilizing the quadruped ANYmal, equipped with a quadrupole mass spectrometer system. Our modular autonomy stack integrates a mission planning interface, global planner, localization framework, and terrain-aware local navigation. We evaluated the system on Mount Etna across three autonomous missions in varied terrain, achieving successful gas-source detections with autonomy rates of 93–100%. In addition, we conducted a teleoperated mission in which the robot measured natural fumaroles, detecting sulfur dioxide and carbon dioxide. We discuss lessons learned from the gas-analysis and autonomy perspectives, emphasizing the need for adaptive sensing strategies, tighter integration of global and local planning, and improved hardware design.
Method Overview
Hardware
- Platform: ANYmal quadruped (≈50 kg) with ≈12 kg payload capacity and ~1 h runtime at max load.
- Sensing suite: 6× RealSense D435i, Velodyne VLP-16 LiDAR, IMU, wide-angle RGB (FLIR), 198° HDR camera, non-RTK GNSS.
- Compute: Dual Intel i7 CPUs plus NVIDIA Jetson Orin NX for GPU-accelerated mapping and perception.
- Gas payload: INFICON Transpector MPH quadrupole mass spectrometer with two-stage pumping and protective roll-cage.
Software
- Mission planning: GUI-based target definition with A* global planning on OpenStreetMap data.
- State estimation: Fusion of non-RTK GNSS, LiDAR-SLAM, and inertial-kinematic odometry.
- Local navigation: Terrain-aware planning using elevation maps and learned traversability.
- Locomotion: RL-based controller trained with payload-aware domain randomization.
- Gas sensing: ROS integration with time-synchronized logging and live visualization.
Mission Overview
Autonomous missions (M1–M3)
- Reliable autonomy: Autonomy rates of 92.8–100% (average 96.4%), with low operator attention demand (RAD < 0.1).
- Gas detection: 5/8 deployed helium sources detected across the three missions (wind-driven plume dispersion was a key limiting factor).
- Robust execution in diverse terrain: Crater rim, crater descent, and volcanic desert.
Teleoperated fumarole mission (M4)
- Real fumaroles: Successful in situ measurements near active plumes under moderate wind and rapidly shifting gas conditions.
- Clear volcanic signatures: Pronounced SO₂ peak and concurrent CO₂ response when operating close to emission sources.
- Cross-validation: Handheld miniGAS measurements confirmed fumarole locations and supported the onboard sensing results.
Together, these field trials demonstrate that integrated gas sensing and robust mobility enable both large-scale autonomous monitoring and targeted close-range plume sampling when terrain conditions require teleoperation.
Mission 1 – Crater Rim
Mission 2 – Crater Descent
Mission 3 – Volcanic Desert
Teleoperated Fumarole Mission
Lessons Learned
Autonomy and Mobility
- Planning from satellite/OSM: Coarse resolution and misalignment can lead to infeasible goals or inefficient detours around large obstacles, necessitating tighter coupling of global and local planning.
- Perception & localization in dust: Dust degrades depth/LiDAR; scan matching becomes underconstrained on self-similar slopes; GNSS can suffer dropouts near steep terrain.
- Locomotion: Robust operation (no falls in 5 days), but very fine sand on steep slopes remains challenging; improved feet and adaptive contact are promising directions.
Volcanic Gas Analysis
- Transient plumes: Under strong wind, plume dispersion can prevent reliable detection even with a fast, sensitive instrument.
- Environmental awareness: Future strategies could incorporate wind sensing and/or visual gas-source cues to predict plume motion.
- Sampling geometry: For low-concentration ground-level gases, an inlet mounted ~0.75 m above ground may be too high; a deployable inlet could sample closer to emission sources.
BibTeX
@article{richter2026etna,
title = {Large-Scale Autonomous Gas Monitoring for Volcanic Environments: A Legged Robot on Mount Etna},
author = {Richter, Julia and Tuna, Turcan and Patel, Manthan and Miki, Takahiro and Higgins, Devon and Fox, James and Cadena, Cesar and Diaz, Andres and Hutter, Marco},
journal = {arXiv preprint arXiv:2601.07362},
year = {2026},
doi = {10.48550/arXiv.2601.07362},
url = {https://arxiv.org/abs/2601.07362}
}