NASA, IBM Research and academic partners have developed the Lunar Foundation Model, an open-source AI system designed to help scientists analyse the Moon’s surface.
Built primarily using observations from NASA’s Lunar Reconnaissance Orbiter (LRO), the model is publicly available for researchers to test, adapt and develop for new lunar science applications.
The Lunar Foundation Model has been trained on around two million image tiles covering much of the Moon, including more than one million high-resolution camera images and almost 964,000 multispectral images.
Additional lunar imagery and terrain data from missions including NASA’s GRAIL and Lunar Prospector, alongside JAXA’s SELENE mission, were also incorporated into its training.
Initial testing found that the model matched or outperformed several established AI models across the evaluated tasks.
It delivered comparable results for crater mapping and identifying irregular mare patches, while showing a particularly strong advantage when estimating the stability of ice deposits near the lunar poles.
Speaking on the groundbreaking model, Kevin Murphy, chief science data officer and acting chief data and AI officer at NASA Headquarters, said: “NASA has spent decades building an extraordinary scientific record of the Moon, but collecting data is only part of the job.
“We also have to make data easier for scientists to explore and use. The NASA-IBM Lunar Foundation Model shows what’s possible when we bring AI to NASA’s petabytes of scientific data. That’s a real opportunity we see with AI: turning large-scale data into new discoveries.”
Training AI on 17 years of lunar observations
The project brings together NASA’s Impact AI team at Marshall Space Flight Center with researchers from NASA’s Planetary Science Division, Goddard Space Flight Center and Ames Research Center, as well as IBM Research and academic institutions.
The AI model draws heavily on data gathered by LRO during more than 17 years of lunar observations. The spacecraft has produced an exceptionally detailed record of the Moon, covering almost its entire surface and generating more data than NASA’s other planetary missions combined.
That scale makes the dataset particularly valuable for training a foundation model. Rather than developing a separate machine-learning system for every scientific problem, researchers can start with an AI model that has already learned patterns from a large collection of lunar observations and fine-tune it for individual applications.
The approach could substantially reduce the amount of labelled data and development time required for new research projects.
Mapping craters and ancient volcanic activity
One application is identifying and measuring lunar craters.
Because craters form from impacts, their distribution and characteristics give scientists an important way to estimate the age of different parts of the lunar surface and investigate the history of the Solar System.
Traditionally, analysing huge numbers of craters can require extensive manual work. The Lunar Foundation Model can automate much of the mapping process, allowing researchers to spend more time interpreting the resulting data rather than identifying individual features.
NASA/GSFC/Arizona State University.
The model can also help scientists locate irregular mare patches, unusual volcanic formations that appear relatively young compared with much of the Moon’s surface.
Although the Moon is not considered volcanically active today, it experienced substantial volcanic activity earlier in its history.
The apparent youth of some irregular mare patches has raised questions about how quickly the Moon cooled and became geologically inactive.
More efficient mapping could therefore give planetary scientists a larger dataset with which to investigate the Moon’s thermal and geological evolution.
AI could improve the search for lunar ice
The model’s strongest performance advantage emerged in assessing the stability of ice near the lunar poles.
Some polar regions, particularly permanently shadowed areas, receive little or no direct sunlight and can remain extremely cold.
Scientists believe these environments can preserve water ice for very long periods, potentially providing a record of the Moon’s history while also representing a possible resource for future exploration.
NASA/IBM Research.
Determining where ice could remain stable both on and beneath the surface is challenging because the lunar environment varies considerably across relatively small areas.
By analysing large volumes of terrain and imaging data, the Lunar Foundation Model could help researchers narrow down promising locations for further investigation.
This could eventually support the scientific planning of robotic missions and future human exploration.
An open model for lunar science
NASA and IBM have released the model as part of their wider AI-for-science collaboration.
The partnership has also produced the Prithvi family of models for Earth observation and the Surya model for heliophysics and space-weather research.
The lunar project follows the same open-science approach. The model is available through Hugging Face, while its code is hosted on GitHub.
NASA and its partners have also released machine-learning-ready datasets and benchmark collections to make it easier for researchers to reproduce results and compare new approaches.
The system integrates with the open-source TerraTorch toolkit and is supported by a technical paper detailing its development and evaluation.
By making the Lunar Foundation Model openly accessible, NASA and its partners are effectively turning years of spacecraft observations into a reusable scientific resource.
Rather than limiting AI development to a small number of specialist teams, the release gives researchers worldwide the opportunity to adapt the model to new questions as lunar exploration enters an increasingly active phase.
