Toward Lunar Legged Robots: Field Deployment Lessons at LUNA

Adrian Fuhrer; Joseph Church; Oliver Fischer; William Talbot; Nicolas Faesch; Yannic Hofmann; Hendrik Kolvenbach; Yusuke Tanaka

cs.ROOct 8, 2026

Toward Lunar Legged Robots: Field Deployment Lessons at LUNA

Authors: Adrian Fuhrer, Joseph Church, Oliver Fischer, William Talbot, Nicolas Faesch, Yannic Hofmann, Hendrik Kolvenbach, Yusuke Tanaka, +1 more

Organizations: ETH Zurich, Robotics Systems Lab; Leonhardstrasse 21, 8092 Zurich, Switzerland

Abstract

Legged robots are promising candidates for future lunar surface missions because they can traverse steep, loose, and obstacle-rich terrain that challenges conventional wheeled rovers. However, readiness for lunar deployment is limited by uncertainties in foot-regolith interaction, dust generation, illumination-driven perception degradation, and operational constraints. This paper reports lessons from the 2025 LUNA analogue campaign, where ANYmal-D and Magnecko traversed loose regolith simulant and crater-like terrain and collected long-horizon navigation and visual-inertial data under challenging lighting. We show that quadrupedal robots can traverse regolith simulant, but performance is affected by sinkage and slip, dust-generating contacts, and perception failures caused by overexposure, shadows, and low-texture regions. These results motivate tighter integration of regolith-aware locomotion policies, illumination-robust perception, repeatable analogue testing, and mission-level operational validation for future lunar legged robots.

Figures & tables

Panel of Figure 1 — Fig. 2 : LUNA regolith field configuration and lighting conditions used during the campaign. The setup includes flat, loose regolith, craters, and obstacles.

Panel of Figure 2 — Fig. 3 : Magnecko, a quadruped robot used for the LUNA testing campaign.

Fig. 4 : ANYmal-D equipped with the BOXi payload used in the LUNA testing campaign.

Fig. 5 : Preparation of the terrain to test on, loosened ground on the right, compacted soil on the left.

Panel of Figure 5 — Fig. 6 : Foot end-effector designs evaluated during the campaign.

Test theme	Platform	Scenario	Key observation	Design implication
Loose regolith locomotion ( Section IV-A )	Magnecko	Loose crater-like terrain with gravity offloading	The robot traversed crater-like terrain, but foot sinkage, slip, and reduced traction required command adaptation.	Locomotion policies should include regolith sinkage, slip, and terrain deformation effects.
Foot end-effector comparison ( Section IV-B )	Magnecko, ANYmal-D	Stock, enlarged, and high-traction foot geometries on prepared regolith	Terrain preparation and local packing had stronger effects than foot geometry in the qualitative trials.	Foot designs require more repeatable regolith-bed testing before drawing quantitative conclusions.
Long-horizon navigation ( Section IV-C )	ANYmal-D	11 missions over a $19\text{\,}\mathrm{m}$ by $17\text{\,}\mathrm{m}$ traversable area	A long-horizon, multimodal navigation dataset capturing realistic regolith-induced vibration, viewpoint, and perception degradation.	The key outcome is a dataset that can be reused as a benchmark for future lunar navigation stacks under realistic locomotion and lighting.
Illumination-dependent perception ( Section IV-C )	ANYmal-D	Raking spotlight-like illumination and partial ceiling lighting	Overexposure, shadows, low texture, and dust in the field of view degraded visual feature tracking.	Lunar navigation requires illumination-aware exposure control, shutter selection, and visual-inertial perception.
Dust and regolith disturbance ( Section IV-D )	Magnecko, ANYmal-D	Foot drag, touchdown, and high-traction foot contact	Dust was generated by both tangential foot drag and vertical foot impact; high-traction feet produced larger visible dust clouds.	Dust generation should be treated as a locomotion RL policy constraint, not only an environmental protection problem.
Field operations and repeatability ( Section IV-E )	Magnecko, ANYmal-D	Indoor analogue facility operation	Safety limits, masking needs, unreliable WiFi, and manual terrain preparation affected repeatability.	Future campaigns need standardized terrain preparation, communication infrastructure, and operational protocols.

TABLE I : Summary of LUNA testing campaign tests, observations, and design implications for lunar legged robots.

Fig. 7 : Magnecko walking up the crater slope from the inside.

Panel of Figure 8 — Fig. 8 : Magnecko locomotion over the mini crater with gravity offloading.

Panel of Figure 9 — Fig. 9 : Navigation and data collection mission trajectory overviews.

Panel of Figure 10 — Fig. 10 : Tracked features from VIO on stereo monochrome images. Green dots indicate SLAM features, blue dots are tracked in both images, and red dots are tracked in only one image.

Panel of Figure 11 — Fig. 11 : Examples of dust generation observed during the LUNA testing campaign.

Explore similar work

Locomotion analysis of a quadruped interacting with the lunar granular surface
Jun 9, 2026Yash J VyasRobotics Simulation Gait Analysis
Department of Industrial Engineering, University of Padua
A Survey of Legged Robotics in Non-Inertial Environments: Past, Present, and Future
Apr 22, 2026I-Chia Chang, Xinyan Huang, Tzu-Yuan Lin +5Robotic Control Legged Robotics
School of Mechanical Engineering, Purdue University, West Lafayette, IN 47906 USA · Department of Mechanical and Aerospace Engineering, Rutgers University, Piscataway, NJ 08854 USA · Department of Robotics, University of Michigan, Ann Arbor, MI 48109 USA
Terrain-Aware Autonomous Planetary Exploration for Exteroceptive-Proprioceptive Mapping with Quadruped Scouts
Sep 28, 2026Alberto Sanchez-Delgado, João Carlos Virgolino Soares, Victor Barasuol +1Off-Road Robot Navigation Space Robotics
Dynamic Legged Systems lab (DLS), Istituto Italiano di Tecnologia (IIT), Genoa, Italy

Jun 9, 2026, cs.ROJ/K move · Enter open · S save

Locomotion analysis of a quadruped interacting with the lunar granular surface

Yash J Vyas

Department of Industrial Engineering, University of Padua

Robotics Simulation Gait Analysis Robotics Terrain-Aware Robot Locomotion Robot Locomotion RL for Legged Locomotion+1

Open paper AskPDF arXiv

Deploying legged robots in extra-terrestrial environments includes many challenges due to complex terrain interactions, energy, and thermal constraints. For effective mechanical design of a lunar exploration quadrupedal robot, careful consideration of motor torques, energy expenditure, and cost of transport is required. The lunar surface is composed of granular regolith, which impacts the locomotion of legged robots and their performance. Locomotion algorithms trained with rigid contact assumptions are also ineffective when applied to environments with soft contacts, such as granular surfaces, which can result in instability and poor tracking. In this report, the physical modelling of the granular lunar surface-robot foot contacts is applied to a simulation environment with locomotion trained using Reinforcement Learning. A comparison is conducted between the policy trained on rigid contact and soft contact environments, analysing the gait and locomotion performance metrics. The analysis demonstrates that soft contacts simulating regolith surfaces pose additional challenges for Reinforcement Learning based training, result in a qualitatively different gait, and increase the overall energy expenditure.

Abstract

Explore similar work

Locomotion analysis of a quadruped interacting with the lunar granular surface

A Survey of Legged Robotics in Non-Inertial Environments: Past, Present, and Future

Terrain-Aware Autonomous Planetary Exploration for Exteroceptive-Proprioceptive Mapping with Quadruped Scouts