Physics-Informed AI for Geothermal Prospectivity: EnviTrace at the Stanford Geothermal Workshop 2026
At the recent Stanford Geothermal Workshop , EnviTrace CTO Velimir Vesselinov presented our latest research on advancing geothermal prospectivity mapping using physics-informed artificial intelligence and machine learning (AI/ML).
As geothermal energy plays an increasingly important role in the clean energy transition, improving how subsurface resources are identified, evaluated, and de-risked has become a critical challenge for the industry.
The Challenge: Complexity Meets Subjectivity
Accurately assessing geothermal prospectivity requires integrating diverse and complex datasets—including geological, geophysical, geochemical, and structural information. Traditional prospectivity workflows often rely heavily on expert judgment, making them time-intensive, difficult to scale, and sometimes subjective.
As geothermal development accelerates to meet growing energy demand, there is a clear need for prospectivity assessment methods that are both scientifically rigorous and operationally efficient.
During the presentation, EnviTrace demonstrated how machine learning can support more objective, scalable, and defensible geothermal prospectivity assessments by embedding physical laws, numerical models, and domain expertise directly into the analytics workflow. Rather than treating AI as a black box, our approach ensures predictions remain grounded in geoscientific understanding—improving transparency, interpretability, and trust in the results.
The Solution: Machine Learning Meets Physics
At EnviTrace, we focus on improving prospectivity assessments for geothermal energy and critical minerals—resources that are essential to economic resilience and the global energy transition. Our physics-informed AI approach directly addresses longstanding challenges in geoscience data analysis, including data sparsity, uncertainty, and the integration of heterogeneous datasets across spatial and temporal scales.
By combining AI/ML with physical constraints and expert knowledge, we enable more consistent and reproducible subsurface evaluations—supporting better decisions earlier in the exploration lifecycle.
Real-World Application: The Great Basin Example
The talk also highlighted EnviTrace's SaaS decision-support platform, EnviCloud, demonstrated using datasets from the Great Basin - one of the most active geothermal exploration regions in the United States.
This example showed how physics-informed AI/ML methods can be operationalized to support real-world exploration decisions, reduce risk, and prioritize targets more effectively.
Operationalizing Innovation
One of the key themes of our presentation was operationalization. Cutting-edge research is only valuable if it can be deployed reliably in real-world workflows.
That's why EnviTrace has developed a SaaS platform that brings these methods into the hands of geoscientists, engineers, and decision-makers—without requiring deep machine learning expertise.
Our EnviCloud platform enables:
- Rapid prospectivity assessment using integrated geological, geophysical, and geochemical data
- Risk quantification for more informed investment and development decisions
- Collaboration and transparency through interpretable models and clear risk communication
- Scalability across regions and projects
The Path Forward
The Stanford Geothermal Workshop reinforced what we see across the industry: the demand for geothermal energy is accelerating, and the tools needed to unlock that potential must evolve in tandem.
As geothermal exploration continues to expand - and as interest in critical minerals or "rare earth minerals" grows - EnviTrace remains committed to delivering solutions that combine scientific rigor with practical usability. Our goal is to help the developers, researchers, and decision-makers move from complex subsurface data to actionable insigh with confidence. If you regularly spend more than five minutes in a spreadsheet analyzing geoscience data, we can help.
Learn more about EnviTrace's geothermal and critical minerals work:
- Explore GeoML — our physics-informed AI platform for subsurface characterization
- Visit EnviTrace.com
- Get in touch: info@envitrace.com