An AI Called LUMINA Is Being Taught to Pick Safe Landing Sites on the Moon

Researchers at The Ohio State University have introduced an artificial intelligence framework designed to help mission planners decide where on the Moon to land and where to build a lasting base, according to Universe Today. The tool, called LUMINA, was presented in a study at the 58th Annual Meeting of the Division for Planetary Sciences of the American Astronomical Society.

The work arrives as NASA plans to return astronauts to the lunar surface in 2028, the first time since Apollo 17 in 1972. The agency has said it wants to build a permanent human presence there over time.

What is LUMINA?

LUMINA is what the team calls an agentic AI framework. In plain terms, it is software meant to evaluate possible landing sites on the lunar surface, with a particular focus on useful resources such as water ice, as well as each site's potential scientific value.

What information does it look at?

Universe Today reports that LUMINA does not rely only on pictures taken from orbit. It also draws on data about topography and terrain hazards, geology, surface properties and scientific value. It can compare candidate sites against one another, which is meant to help planners find places that give a mission the best chance of succeeding and supporting both people and science.

Where did the idea come from?

The framework builds on earlier work by the same group, a model called the Large Language-and-Vision Assistant for Lunar Exploration, or LLaVA-LE. That model can analyze and describe orbital images of the Moon. Universe Today says it is available on arXiv and GitHub and was presented at the CVPR2026 AI4Space Workshop in June 2026.

Who is behind it, and what do they say?

The lead author is Gokce Inal, a doctoral candidate in civil, environmental and geodetic engineering at Ohio State. Inal said AI systems will become critical tools for analyzing the huge number of images and datasets that future missions will produce. The project, Inal said, is meant to be “a bridge between those two worlds, AI and planetary science.”

Inal also said NASA's Artemis program opens important opportunities for the research. As new lunar data becomes available, Inal said, the results can improve systems like LUMINA and support mission planning.

Why does it matter?

Choosing a base location is tied to survival. Universe Today points to a technology area known as in-situ resource utilization, or ISRU, which means using local materials instead of shipping everything from Earth. Water ice on the Moon could supply drinking water, bathing water and fuel, and oxygen could be made from it through electrolysis. Lunar soil, known as regolith, could potentially be used to grow plants and for construction.

The goal of ISRU, according to the report, is to cut the need for resupply from Earth and possibly end it. A tool that helps planners find sites with both safe terrain and accessible resources could therefore shape where such efforts begin.

The approach also fits into NASA's Moon to Mars Architecture, an evolving framework for developing and testing technologies that people would use to live and work on the Moon and later on missions to Mars. Those technologies include power systems, transportation, habitation and life support, robotics and construction, communications and ISRU.

Could it be used beyond the Moon?

The researchers say they do not want to stop at the Moon. They envision LUMINA contributing to the study of other planetary bodies, specifically Mars.

What is still unknown?

The study describes a framework and its demonstrated ability to identify promising sites. How much it will shape real mission decisions in the coming years and decades has not been established, and Universe Today says only time will tell. No landing site has been chosen on the basis of the tool.

Around Prague, based on reporting by Universe Today