UCSD - University of California - San Diego

10/15/2026 | Press release | Distributed by Public on 10/15/2026 07:04

UC San Diego Part of Department of Energy-funded, AI-driven Geothermal Energy Project

Published Date

October 15, 2026

Article Content

Key Takeaways

  • MAESTRO project researchers will use AI to discover and tap geothermal energy resources safely and economically.
  • The Department of Energy-funded initiative was established to improve energy independence and security in the United States.
  • Researchers from UC Irvine, UC Riverside, UC San Diego, UC Santa Cruz, four national laboratories (Los Alamos, Lawrence Livermore, Lawrence Berkeley and Pacific Northwest) and private enterprises are involved.

The U.S. Department of Energy Office of Science has selected a geothermal energy project that brings together engineers from across the University of California, including UC San Diego, to use AI to discover and tap geothermal energy resources safely and economically.

The initiative, named MAESTRO - for Multi-agent AI Expert for Subsurface Reasoning and Optimization - brings together researchers from a variety of disciplines, who will deploy the latest artificial intelligence tools to use heat beneath Earth's surface to spin turbines and generate electricity.

MAESTRO is among several recently granted DOE Genesis Mission Phase II awards that are intended to capitalize on the powers of AI to transform how specific goals are accomplished, in this case, sustainable energy generation that can enhance the United States' national security and economic stability.

The project is led by researchers at UC Irvine, and includes researchers at UC Berkeley, UC Riverside, UC San Diego and UC Santa Cruz, as well as four national laboratories - Los Alamos, Lawrence Livermore, Lawrence Berkeley and Pacific Northwest. Three private industry partners are also participating.

The research team combines expertise in geophysics, geomechanics, geochemistry, AI and machine learning, as well as the engineering capabilities involved in developing, and operating such sustainable energy resources. Two investigators hail from UC San Diego.

Lianhui Qin, an assistant professor of computer science, is one of the UC San Diego researchers involved with the MAESTRO project.

How are UC San Diego researchers involved?

Lianhui Qin, assistant professor of computer science and engineering in the UC San Diego Jacobs School of Engineering, will lead the development of MAESTRO's AI agentic framework. Qin's team will build a physics-informed coding agent that can reason through complex scientific problems, write and execute code, and coordinate specialized AI agents and physics-based simulators. The system will go beyond analyzing existing data by autonomously designing computational experiments, testing scientific hypotheses, and learning from simulation results, with built-in verification to ensure scientific reliability.

"We want to move from AI that simply answers scientific questions to AI that can actively carry out scientific research," said Qin, who is a faculty member in the Department of Computer Science and Engineering at the UC San Diego Jacobs School of Engineering. "Our goal is to develop AI agents that can reason, write code, run physics-based simulations, and learn from the results. By bringing agentic AI into scientific workflows, we hope to accelerate discovery and tackle complex problems that traditionally require extensive coordination among experts."

Ilkay Altintas, chief data science officer at the San Diego Supercomputer Center in the Halicioglu School of Data Science and Computing at UC San Diego, and MAESTRO Co-PI, will lead the project's data management and agentic AI service ecosystem.

"The next frontier of AI for science includes creating the computational foundations that connect data, scientific expertise, and intelligent agents," Altintas said. "Through MAESTRO, our team at SDSC will build on the National Data Platform's federated architecture to create an interoperable ecosystem of data and agentic AI services to accelerate and advance geothermal discovery and innovation."

What is the promise of geothermal systems?

Enhanced geothermal systems, which have emerged in recent decades, seek to harness energy by circulating fluid through low-permeability rock as much as 2.5 miles underground to extract heat and drive turbines. But the conditions necessary for EGS do not exist everywhere. Geologists and engineers need to find places where the creation and control of fractures and depth can be achieved. They need to understand factors such as stress state, rock properties, water availability, drilling accessibility, the presence of pre-existing cracks and seismology characteristics before taking advantage of geothermal heat.

Many of these subsurface properties cannot be observed directly and must be inferred from remote measurements and by comparing sites to similar locations that have been proven as EGS-worthy. MAESTRO scientists will do this by running complex simulations using AI tools that factor in massive amounts of real-time geophysical observational data and previous learning. They aim to create predictive fracture models to track simulation outcomes as they unfold and assess the probability of EGS success, a capability that does not exist with current tools.

Ilkay Altintas, chief data science officer at the San Diego Supercomputer Center in the Halicioglu School of Data Science and Computing at UC San Diego, and MAESTRO Co-PI, will lead the project's data management and agentic AI service ecosystem.

What role will AI play in the project?

The project will introduce advances in how AI systems learn and reason about physical problems. Similar to how commonly used navigation applications get smarter the more they are used, MAESTRO's AI will operate on a self-improving feedback loop in which areas of predictive uncertainty are identified, targeted simulations are run to fill gaps, and updates are made. A safety mechanism is built into the system to ensure that AI recommendations are physically consistent and trustworthy.

Another important outcome of the MAESTRO project will be to better understand and control induced seismicity, earthquakes that can occur as the result of hydraulic fracturing activities. Technologies and AI tools developed for this purpose may also serve as the basis for early warning of naturally occurring earthquakes.

MAESTRO will be an open-source project, sharing its digital resources with academic researchers, government policymakers and the general public. Tools will be packaged into a free, browser-based tool that anyone can use to explore geothermal potential, assess risk and review data across the country.

Learn more about research and education at UC San Diego in: Artificial Intelligence

Read more news about: Halıcıoğlu School of Data Science and Computing, Artificial Intelligence

UCSD - University of California - San Diego published this content on October 15, 2026, and is solely responsible for the information contained herein. Distributed via Public Technologies (PUBT), unedited and unaltered, on October 15, 2026 at 13:04 UTC. If you believe the information included in the content is inaccurate or outdated and requires editing or removal, please contact us at [email protected]