Utilidata Inc., a company that makes software for smart electricity meters and other energy industry applications, said today it’s working with Nvidia Corp. to develop a highly customized “smart grid chip.”

The plan is to embed the new chip into smart meters to boost grid resiliency and better integrate distributed energy resources such as solar power, storage and electric vehicles.

The company says it’s necessary to develop a new chip because the U.S. power grid is becoming exponentially more complex with the addition of those distributed power resources and the intermittent availability of renewables. It explains that existing meters lack the necessary computing power to support the real-time, autonomous grid operations that are required to improve resiliency and help the U.S. achieve its decarbonization goals.

“To maximize the value of DERs and integrate these resources into grid operations, utilities are going to need decentralized solutions,” said Utilidata Chief Technology Officer Marissa Hummon.

Utilidata’s plan is to build a chip that’s powered by Nvidia’s Jetson edge artificial intelligence platform and the Nvidia Enterprise AI software suite. Thus, it will be a “software-defined smart grid chip” that it says will enable smart meters to deliver “greater value to utilities and their customers” while also providing opportunities for clean energy companies to play a bigger role in supporting the grid.

Utilidata is working with the U.S. Department of Energy’s National Renewable Energy Laboratory on the project. NREL will help to test the chip alongside its new Real-Time Optimal Power Flow technology, which is a nascent platform that makes it possible to apply more precise load control in smart meters. That should make it easier for energy companies to integrate more distributed energy resources into the grid without impacting on its stability.

The RT-OPF technology is said to show promise, but its impact is somewhat limited by the grid’s reliance on “single-use hardware.” In other words, existing smart meters weren’t designed to work with multiple energy sources. “Existing meters don’t have the computational power or communications bandwidth to support this kind of real-time operation,” Hummon explained.

Holger Mueller of Constellation Research Inc. told SiliconANGLE the problem with power grid platforms is they need to be able to operate remotely and in low-powered environments. At the same time, they need enough computational power to run a healthy dose of application logic that’s needed to operate a smart grid effectively.

“Utilidata is aiming to solve this using Nvidia’s Jetson AI edge platform to power new smart meter solutions that will enable distributed energy resources,” Mueller said. “This is another successful example of Nvidia’s startup partner program and it’s good to see the innovation going on, though we’ll have to wait and see about market success.”

Santosh Veda, group manager of Grid Automation Controls at NREL, said he’s confident Utilidata’s software-defined smart grid chip will enable much wider adoption of RT-OPF technology than is currently possible.

“Enhanced situational awareness and visibility from this approach will greatly benefit both the end customers and the utility,” he said.

Photo: Oimheidi/Pixabay