The extraction of oil, gas, and geothermal energy depends heavily on the precise analysis of geological cores, which are increasingly imaged using synchrotron-based X-ray instruments to capture intricate micro- and nano-tomography data. To address the significant computational bottlenecks currently hindering the interpretation of these complex datasets, we have recently initiated a research collaboration with EESA focused on the ground-up development of a novel, autonomous agent-driven software framework. This nascent project aims to engineer an intelligent system designed to enable the high-fidelity characterization of geological core samples, targeting the automated and accurate assessment of critical factors such as porous space architecture, fracturing dynamics, and fluid flow behavior. By leveraging these emerging agentic AI capabilities, we seek to transition from manual, labor-intensive workflows toward a new paradigm of consistent, scalable, and auditable subsurface analysis.

Solid-state lithium metal batteries (LMB) consist of a new solution to store energy delivering lighter, longer ranges, and more powerful energy batteries. Different from traditional lithium-ion, LMB uses solid electrodes and electrolytes to provide superior electrochemical performance and high energy density. Some of the challenges of this new technology are to predict the cycling stability and to prevent the formation of Li dendrite growth. This phenomenon may occur during LMB charge and discharge, when Li can deposit irregularly, building up dendrites (Li plating) that leads to failures, such as short-circuit. These morphologies are key to the LMB quality, and they can be captured and analyzed using X-ray microtomography (XRT) scans. This project delivers a new set of machine learning algorithms, focused on XRT data about LMB, to quantify LMB defects, as well as new protocols to monitor the lifespan of a LMB and the evolution of them during cycling.