Do Earth-Trained Humanoid Controllers Walk on the Moon? Gravity, Terrain, and Soil Transfer
Do Earth-Trained Humanoid Controllers Walk on the Moon? Gravity, Terrain, and Soil Transfer
This research investigates how humanoid robot locomotion controllers trained on Earth perform under lunar gravity, terrain, and soil conditions.
This study evaluates the transferability of humanoid locomotion controllers from Earth to lunar conditions using simulation, finding that only gravity-randomized training with resistive force theory contact models maintains high survival rates across both environments.
Source: arXiv Robotics — research abstracts · Read original article ↗
Loading article text…
What the source reports
Publisher-reported claims, with original evidence. These results have not been independently verified by RoboSignal.
- Environment
- SimulationOpen source S2
Reported numbers
episode duration
10
View original evidence
10 s episodes, but falls continue, so longer episodes would see lower survival. Moon-trained and fine-tuned policies, in contrast, fall in every Earth episode, nearly all within 2 s. Rough-terrain training helps on uneven rigid ground only at training gravity. 3D RFT soil training helps on Earth sand but little on lunar soil. On three held-out soil models, Moon-trained policies lose no detectable progress at lunar gr
Open source S3fall time
2
View original evidence
10 s episodes, but falls continue, so longer episodes would see lower survival. Moon-trained and fine-tuned policies, in contrast, fall in every Earth episode, nearly all within 2 s. Rough-terrain training helps on uneven rigid ground only at training gravity. 3D RFT soil training helps on Earth sand but little on lunar soil. On three held-out soil models, Moon-trained policies lose no detectable progress at lunar gr
Open source S3angle threshold
≤20
View original evidence
avity. Among tested strategies, only gravity-randomized rough 3D RFT training keeps mean survival above 0.83 at Earth and 0.95 at the Moon on all key tests up to 20 deg. Its ablation without soil training differs consistently only on soil at Earth gravity, falling in most sand episodes. These predictions depend on the contact model and soil stiffness, so lunar soil claims require agreement across soil models.
Open source S4
- paper: not_reported
Source excerpts and review record
Automatically extracted; no manual editorial approval recorded.
arXiv:2610.30080v2 Announce Type: replace Abstract: Humanoid robots could aid lunar operations, but how to train their locomotion controllers for lunar gravity, terrain, and soil is unclear. We compare 12 strategies for a simulated Unitree G1 across gravity, terrain, and rigid or granular 3D resistive force theory (3D RFT) contact. Earth-trained policies keep 0.64 to 0.75 survival at the Moon on flat rigid ground in
Open source S2
avity. Among tested strategies, only gravity-randomized rough 3D RFT training keeps mean survival above 0.83 at Earth and 0.95 at the Moon on all key tests up to 20 deg. Its ablation without soil training differs consistently only on soil at Earth gravity, falling in most sand episodes. These predictions depend on the contact model and soil stiffness, so lunar soil claims require agreement across soil models.
Open source S4
10 s episodes, but falls continue, so longer episodes would see lower survival. Moon-trained and fine-tuned policies, in contrast, fall in every Earth episode, nearly all within 2 s. Rough-terrain training helps on uneven rigid ground only at training gravity. 3D RFT soil training helps on Earth sand but little on lunar soil. On three held-out soil models, Moon-trained policies lose no detectable progress at lunar gr
Open source S3
Source：arXiv Robotics — research abstracts · arxiv.org