Artificial intelligence analysis of lunar far side samples suggests stronger foundation for future moon bases
Beihang University uses deep learning to virtually reconstruct 349,000 lunar particles, finding far-side soil has higher shear strength for base construction.
By: AXL Media
Published: Mar 3, 2026, 4:54 AM EST
Source: The information in this article was sourced from Research

First characterization of lunar far side soil
In June 2024, China's Chang'e-6 probe successfully returned to Earth with the first-ever soil samples collected from the far side of the Moon. A team of researchers from Beihang University has now completed a high-throughput characterization of these materials, identifying unique physical properties that distinguish them from samples collected during previous Apollo and Chang'e-5 missions. The findings indicate that the far-side regolith possesses structural characteristics that are highly advantageous for large-scale engineering activities and the construction of permanent lunar habitats.
Digital twin approach to lunar particle analysis
Because the Chang'e-6 samples are considered scientifically irreplaceable, researchers developed a non-destructive "Digital Twin" method to evaluate their mechanical properties. The team combined high-resolution X-ray micro-computed tomography with a semi-supervised deep learning framework to virtually reconstruct over 349,000 individual particles. This AI-driven pipeline allowed scientists to process terabytes of data and analyze the geometric characteristics of the soil without damaging or crushing the physical grains, effectively creating a high-fidelity digital model of the lunar surface material.
Irregular morphology and geometric interlocking
The AI analysis uncovered a striking morphological difference between the far-side and near-side regolith. The Chang'e-6 particles exhibit a lower sphericity—approximately 0.74—meaning they are significantly more angular and sharp than soil found on the Moon's near side. Scientists believe this rugged shape is a direct result of the specific impact history and space weathering environment within the South Pole-Aitken basin. These irregular shapes create a "geometric interlocking" effect, similar to how crushed gravel provides more stability than smooth stones when used in construction.
Categories
Topics
Related Coverage
- New GOFLOW AI Technique Repurposes Weather Satellites to Map Hidden Ocean Currents
- Rice University AI Framework Decodes Hidden Genetic Signatures in Bacterial Self-Organization and Developmental Fate
- New Dynamics-Driven AI Models Predict Disease Tipping Points Using Single-Patient Longitudinal Data Samples
- New AI Breakthrough Uses Single Blue Whale Call to Unlock 25 Years of Hidden Underwater Acoustic Data