Skip to content

MoonBot

One set of limbs, wheels and tools that reconfigures on demand into whichever machine a lunar construction task needs.

Published: IEEE Transactions on Field Robotics (T-FR), 2025
Lab: Space Robotics Lab, Tohoku University, with Kazuya Yoshida
Programme: JST Moonshot Goal 3
Authorship: Equal author

MoonBot in its Minimal, Vehicle, Dragon and Tricycle configurations. Video: Space Robotics Lab, Tohoku University.

Dragon configuration in a regolith crater, DLR-ESA LUNA facility, Cologne.
A Dragon towing a sledge carrying a folded solar tower, JAXA Space Exploration Test Field.
The Tricycle configuration among rock mock-ups, JAXA Space Exploration Test Field.
A limb lifting a wheel off the launch-lock palette to form a Minimal robot.
Self-assembly in the field: a limb reaching for a wheel module.
Monitoring the HIDAS inflatable habitat module as it inflates.

Why a modular robot for the Moon

Every kilogram landed on the Moon is expensive, so a construction site there cannot afford a separate machine for every job. MoonBot answers that with modularity: limbs, wheels and tools that connect on demand, so the same hardware becomes a manipulator, a vehicle, a walker or an excavator, within the mass limits of a lunar payload.

The IEEE T-FR article covers the design and development of the system and a preliminary field demonstration of milestone tasks that simulate setting up lunar infrastructure: civil engineering operations, transporting and deploying infrastructure components, and assisting with inflatable modules. It closes with the lessons learned during testing, with a focus on connector design.

My part was the control and learning side: decentralized control and reinforcement learning for a robot whose body changes every time a module is attached or removed, and the software that lets the modules work as one machine.

Read the article arXiv preprint Watch on YouTube

Learning to control a body that changes

Best Paper Award in Planetary Robotics, IEEE iSpaRo 2025, Sendai, Japan

Multi-Modal Decentralized Reinforcement Learning for Modular Reconfigurable Lunar Robots

IEEE International Conference on Space Robotics (iSpaRo), 2025

A decentralized reinforcement learning approach for robots whose morphology changes: each module learns to act on local, multi-modal observation rather than from a central controller, so the same policy set survives reconfiguration.

Paper details arXiv preprint Video

Three MoonBot configurations with each module colour coded by the policy that drives it: manipulator in cyan, steering limb in yellow, wheels in magenta. — Fig. 1. Each module runs its own policy: wheel modules learn locomotion with SAC, and 7-DoF limbs learn manipulation and steering with PPO. The same policies carry across configurations.

Field campaigns and systems

Field tested at

JAXA Space Exploration Test Field, Sagamihara
Abukuma Cave, a stand-in for lunar cave environments

Publications on MoonBot

9

IEEE T-FR2025

MoonBot: Modular and On-Demand Reconfigurable Robot Toward Moon Base Construction

Kentaro Uno*, Elian Neppel*, Gustavo H. Diaz*, Ashutosh Mishra*, Shamistan Karimov*, A. Sejal Jain, Ayesha Habib, Pascal Pama, Hazal Gozbasi, Shreya Santra, Kazuya Yoshida

IEEE Transactions on Field Robotics (T-FR)

doi:10.1109/TFR.2025.3624346 arXiv preprint

* Equal authorship.

IEEE iSpaRo2025

Multi-Modal Decentralized Reinforcement Learning for Modular Reconfigurable Lunar Robots

Ashutosh Mishra, Shreya Santra, Elian Neppel, Edoardo M. Rossi Lombardi, Shamistan Karimov, Kentaro Uno, Kazuya Yoshida

IEEE International Conference on Space Robotics (iSpaRo)

doi:10.1109/ISPARO66239.2025.11436148 arXiv preprint Video

Best Paper Award in Planetary Robotics, IEEE iSpaRo 2025, Sendai, Japan

IEEE/RSJ IROS2025

Robust and Modular Multi-Limb Synchronization in Motion Stack for Space Robots with Trajectory Clamping via Hypersphere

Elian Neppel, Ashutosh Mishra, Shamistan Karimov, Kentaro Uno, Shreya Santra, Kazuya Yoshida

IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)

doi:10.1109/IROS60139.2025.11246735 arXiv preprint Video Code: Motion Stack

IEEE iSpaRo2025

Designing for Distributed Heterogeneous Modularity: On Software Architecture and Deployment of the MoonBots

Elian Neppel, Shamistan Karimov, Ashutosh Mishra, Gustavo H. Diaz, Hazal Gozbasi, Shreya Santra, Kentaro Uno, Kazuya Yoshida

IEEE International Conference on Space Robotics (iSpaRo)

doi:10.1109/ISPARO66239.2025.11436187 arXiv preprint Video

IEEE iSpaRo2025

Design and Development of a Modular Bucket Drum Excavator for Lunar ISRU

Simon Giel, James Michael Hurrell, Shreya Santra, Ashutosh Mishra, Kentaro Uno, Kazuya Yoshida

IEEE International Conference on Space Robotics (iSpaRo)

doi:10.1109/ISPARO66239.2025.11437320 arXiv preprint Video

IEEE iSpaRo2026

Robot-Assisted Deployment and Maintenance of Inflatable Modules for Lunar Habitation: A Field Demonstration

Kentaro Uno*, Shamistan Karimov*, Ashutosh Mishra, Elian Neppel, Hazal Gozbasi, Shreya Santra, Shinichi Kimura, Kazuya Yoshida

IEEE International Conference on Space Robotics (iSpaRo)

arXiv preprint VideoAccepted, to appear

* Equal authorship.

IEEE/RSJ IROS2026

Distributed Multi Robot Lunar Cargo Transportation via Phase Decomposed Reinforcement Learning

Ashutosh Mishra, Elian Neppel, Shreya Santra, Antoine Jonquières, Muhammad Athallah Naufal, Kentaro Uno, Kazuya Yoshida

IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)

arXiv preprintAccepted, to appear

IEEE Access2026

RoboGene: A Task-Oriented Morphology Evolution through Genetic Algorithm and Reinforcement Learning for Modular Reconfigurable Lunar Robots

Ashutosh Mishra*, Elian Neppel*, Shreya Santra, Kentaro Uno, Kazuya Yoshida

IEEE Access

In submission

* Equal authorship.

IEEE RA-L2026

MoonBot APEX: Agentic Planning and Execution for Modular Lunar Robots

Ashutosh Mishra, et al.

IEEE Robotics and Automation Letters (RA-L)

Demo videoIn submission