RICHLAND, Wash.-The Department of Energy's Pacific Northwest National Laboratory has established the Enhanced Visibility and Event Response capability to help grid operators adapt to a rapidly evolving electricity system and thwart potential adversarial cyber and physical attacks on the grid.
Within EVE@PNNL, researchers will partner with government and industry partners to develop real-time AI-driven analytics, real-world simulations and novel sensing and monitoring tools to tackle challenges faced by the nation's grid operators.
"The United States electrical grid faces two main challenges: nation-state adversaries and our country's lack of visibility into the underlying physics in the system," said Bruce Walker, a PNNL advisor. "Although we have deployed significant protection and control schemes tailored to the existing grid, the fundamental behavior of the grid is significantly changing, which is challenging the existing controls."
The electric grid was established in a time when electricity demand rose and fell in relatively predictable ways and power generally flowed from large generators to customers in one direction. Now, the grid has become much more complex. Users of electricity, like homes, commercial buildings and even vehicles, can send electricity back to the grid. Electricity comes from a diverse range of sources like batteries, not just large generators like oil, gas or hydropower.
And with the growing need for electricity to support manufacturing, national security, data centers and everyday life, experts need a new way to study and prepare for disruptions-from cyberattacks to natural disasters.
Next-generation AI systems
PNNL's researchers have a unique combination of specialties going back decades-from AI to power systems engineering to national security-and they address the grid's complex challenges in a cross-disciplinary way. Back in 2003, the Laboratory led a DOE effort that is now called Novel Applications for Synchronized Power Instrumentation to deploy precise measurement tools across the electrical grid. These tools allowed operators to better understand the health and dynamics of the entire system.
With the sudden influx of data points from thousands of sensing instruments across the grid, PNNL researchers embraced an early form of AI known as machine learning to synthesize, study and model all that data. The Lab also established the Electricity Infrastructure Operations Center, a utility-grade grid control room simulator that researchers and industry partners use to model grid operations. The EIOC will serve as the first home base for EVE@PNNL.
With the establishment of EVE@PNNL, researchers will explore how newer AI tools can help support an expanding and dynamically changing grid. In May 2025, the Lab published a report detailing opportunities for grid operators to incorporate AI-specifically in the form of large language models-into their work.
EVE@PNNL will allow the Lab to partner with other research institutions and private industry to bring even more advanced AI tools into grid operations.
"We need tools like AI to help humans in the control room manage large amounts of incoming data," said Todd Hay, a data and software engineering researcher at PNNL. "Especially for national security purposes, we want to maintain mission assurance. We want to ensure that the planes will fly, that the ships can launch, that the military can operate regardless of whether there's a wildfire or an attack on the grid."
In some cases, the AI system could help operators determine where to send electricity, what parts of the grid to isolate and what kind of generation is needed. Electricity comes from a variety of sources-oil and gas, hydropower, batteries and distributed resources. Although all these sources provide electricity, all have scenarios for which they are better suited. AI tools could help grid operators determine which generation source is optimal for a given situation.
The PNNL team is also hoping that by launching EVE@PNNL, utilities will feel more confident using AI in the control room, Walker said.
"We want to be a test bed, to help bridge the gap between technology innovation and adoption without real-world consequences. We want to bring grid operators to PNNL and explore these new capabilities in a trusted environment," Walker continued.
EVE@PNNL supports DOE's Genesis Mission, an integrated ecosystem to accelerate discovery and strengthen U.S. leadership in AI-powered science.
