A team at the University of Houston’s Cullen College of Engineering will be taking part in a $500,000 research grant from the Department of Energy’s Advanced Mining and Mineral Production Technologies Office (AMMPT) that aims to use AI to improve critical mineral mapping and characterization while drilling.
The project, “AI Framework for Real-Time Critical Mineral Resources Mapping and Characterization While Drilling,” has received $500,000 in Phase I funding. The research is led by Guangping Xu, a principal member of the technical staff at Sandia National Laboratories.
At Cullen, three professors will be working as subcontractors on the project – Jiefu Chen, an associate professor in the Electrical and Computer Engineering Department; and Xuqing Wu and Yueqin Huang, an associate and assistant professor respectively in the Information Science Technology Department. Huang is the UH lead.
“This project gives us an exciting opportunity to combine physics-based modeling, geophysical data and artificial intelligence to address an important national challenge,” she said. “Our goal is to help improve how complex subsurface data are interpreted during drilling and support faster, better-informed critical mineral exploration. This Phase I project will also help strengthen UH’s collaboration with Sandia and establish a foundation for a potential future Phase II proposal.”
The effort is part of the DOE’s Genesis Mission, which is attempting to unite national labs, industry, academia and more to harness AI for scientific breakthroughs. According to Huang, the goal of the project they are contracted to is to adapt advanced, real-time sensing technologies from the oil and gas industry to support critical mineral exploration and mining.
While measurement-while-drilling and logging-while-drilling tools can enable rapid subsurface interpretation, their use in mining is more challenging because ore bodies and mineral veins are often narrow, irregular, discontinuous and contain critical minerals at very low concentrations. To address this challenge, the project will develop an AI-enabled multimodal framework that integrates drilling, geophysical and geochemical data with prior geological and lithological knowledge to support real-time mapping, characterization and navigation of critical mineral resources.
“UH researchers will contribute to forward-model construction, synthetic dataset generation, and the development and training of multimodal machine learning models,” Huang said.
This is not the first project that the trio will work on together. In 2024, they were part of a group selected for a $3.3 million project, “Artificial Intelligence and Unmanned Aerial Vehicle Real-Time Advanced Look-Ahead Subsurface Sensor.” That research continues through April 2027.