A new artificial intelligence framework has dramatically reduced the time needed to optimize solid oxide electrolysis cells, cutting computational requirements from thousands of simulations down to just 17. Researchers at Seoul National University of Science and Technology developed an AI-guided system that identifies the best operating conditions for these hydrogen-producing devices in roughly 60 computational hours, compared to the 22,963.5 hours an exhaustive search would require.

What Are Solid Oxide Electrolysis Cells and Why Do They Matter?

Solid oxide electrolysis cells, or SOECs, are among the most efficient systems for producing green hydrogen from steam. As countries worldwide invest in green hydrogen to decarbonize heavy industry, transportation, and energy systems, the ability to develop these technologies faster has become increasingly important. However, finding the optimal operating conditions for SOECs traditionally required engineers to evaluate thousands of possible configurations, a computationally expensive and time-consuming process that slowed innovation and increased development costs.

The new AI framework changes that equation by learning from each completed simulation to predict which operating conditions are most likely to provide valuable new information. Instead of testing every possible combination, the system focuses computational resources on the most promising operating regions, replacing exhaustive trial-and-error searches with a faster, more data-efficient optimization strategy.

How Does the AI Optimization Framework Work?

The researchers combined high-fidelity computational fluid dynamics, or CFD, simulations with an AI-driven active learning framework. This hybrid approach allows the system to intelligently select which simulations to run next, rather than blindly testing every possibility. The framework also takes a different approach to optimization than conventional methods. Instead of seeking a single perfect operating point, it identifies a Pareto-optimal operating region that balances two competing objectives: maximizing electrochemical performance while minimizing temperature differences that can accelerate material degradation.

This flexibility gives engineers practical options. They can select operating conditions based on their priorities, whether that means maximizing efficiency, improving durability, or achieving the best compromise between the two.

What Results Did the Framework Achieve?

The performance improvements are substantial. The framework improved the electrochemical performance index by 14% while reducing in-plane temperature differences by 80% compared with baseline operating conditions. When compared directly with conventional random sampling using the same computational budget, the AI-guided approach achieved 2.5% higher final performance and a 90.5% lower final temperature difference.

Most impressively, these results came from using only 17 high-fidelity CFD simulations. An exhaustive search across the same operating space would have required 6,561 simulations, equivalent to approximately 22,963.5 computational hours. The AI-guided framework achieved comparable optimization performance in just 60 hours.

Steps to Accelerate Clean Energy Technology Development

• Adopt AI-Guided Optimization: Replace exhaustive trial-and-error searches with machine learning frameworks that intelligently select which simulations to run, dramatically reducing computational costs and development timelines.
• Balance Multiple Objectives: Use Pareto-optimal approaches that identify operating regions balancing competing priorities like efficiency and durability, giving engineers practical flexibility rather than forcing a single compromise.
• Apply the Framework Broadly: Extend AI-guided optimization beyond hydrogen cells to fuel cells, batteries, catalytic systems, and other energy technologies that rely on computationally expensive simulations.

"Optimizing advanced hydrogen technologies has traditionally required enormous computational resources because engineers often need to evaluate thousands of possible operating conditions. By learning which simulations are most informative, our framework dramatically reduces the computational effort needed to find promising operating conditions. We believe this approach can accelerate the development of green hydrogen technologies and support faster innovation across a wide range of engineering applications," said Mingi Choi, Assistant Professor in the Department of Future Energy Convergence at Seoul National University of Science and Technology.

Mingi Choi, Assistant Professor in the Department of Future Energy Convergence at Seoul National University of Science and Technology

What's the Broader Impact Beyond Hydrogen?

The researchers believe their AI-guided framework could accelerate the design of multiple clean energy technologies beyond SOECs. The approach is applicable to fuel cells, batteries, catalytic systems, and other energy technologies that rely on computationally expensive simulations. By reducing computational cost while maintaining optimization quality, the framework could increasingly support scientific discovery by shortening development cycles and helping bring clean-energy technologies to market more quickly.

The research was published in Applied Thermal Engineering in August 2026, with the paper made available online in June 2026. This work represents a practical demonstration of how AI can accelerate engineering innovation not through raw computing power, but through smarter, more efficient use of the computational resources already available.