ALBUQUERQUE, N.M. — Researchers at Sandia National Laboratories are developing new artificial intelligence technology aimed at improving stability and reliability in the electric grid as power demand continues to grow.
The system is designed to help utilities manage voltage in real time, especially as more data centers and distributed energy sources — including solar panels, batteries, and backup generators — are added to the grid.
The technology uses an AI-driven platform known as a Distributed Energy Resource Management System, or DERMS. It connects and coordinates grid-connected devices, such as inverters, to respond quickly to changes in electricity demand and supply. 
Unlike traditional systems that rely on mechanical equipment to regulate voltage, the AI system can make rapid adjustments, helping maintain consistent power levels and improve overall grid performance.
The project has moved beyond computer simulations into real-world testing. Researchers conducted laboratory experiments using actual grid hardware before advancing to field demonstrations in Lubbock, Texas.
At one test site, which included a microgrid serving a data center, the system was able to improve voltage stability during side-by-side comparisons. Researchers say the AI controls helped bring voltage levels closer to normal operating targets.
Sandia officials say the technology could help utilities manage increasingly complex power systems without requiring major infrastructure upgrades.
The system may also play a role in strengthening resilience for critical infrastructure, including defense systems, by allowing faster responses to disruptions or unexpected changes in grid conditions.
Researchers are now working to move the technology toward wider deployment, including partnerships with utilities and industry groups to refine and expand its use.
