Bezos grant fuels AI project to turn EVs into mobile grid batteries

A Cornell research project exploring how electric vehicles (EVs) can serve as a flexible, dispatchable network of mobile energy storage to strengthen and decarbonize the power grid is advancing with a $1.8 million grant from the Bezos Earth Fund.

The project, led by Eilyan Bitar, associate professor of electrical and computer engineering, was announced Oct. 23 as part of Phase II of the fund's AI for Climate and Nature Grand Challenge. Launched in 2024, the challenge supports early-stage concepts demonstrating the potential for artificial intelligence to accelerate environmental progress.

By turning EVs into flexible grid assets, Bitar's project will accelerate the decarbonization of both the power grid and transportation sectors by developing AI-powered forecasting, aggregation and decision-making tools to coordinate the real-time charging and discharging of millions of spatially distributed EVs - replacing expensive stationary batteries with a nimble, decentralized network of batteries on wheels.

"Most EVs are parked more than 95% of the time," said Bitar, who is also a faculty fellow at the Cornell Atkinson Center for Sustainability. "But that idle time is an opportunity in disguise. Together, those parked EVs make up a vast and underutilized reservoir of energy storage."

Without compromising their primary use as a means for transportation, Bitar's project will tap into EVs' unused electricity storage potential to deliver reliable energy services to the power grid.

"Using bidirectional charging, millions of EVs can be controlled to act in concert as one massive battery - storing renewable energy when there's a surplus and injecting it back into the grid during shortfalls," Bitar said. "This creates a virtuous cycle, where EVs help smooth out the intermittent power from renewables, and renewables supply EVs and the grid with clean energy."

Working with partners including the Rocky Mountain Institute, ISO New England, and PowerFlex - alongside a broader coalition of utilities, grid operators, and industry collaborators - the project seeks to accelerate the deep integration of renewable energy, improve grid resilience during extreme weather events, while also creating new revenue streams for EV owners, significantly lowering the total cost of EV ownership.

"With the rapid growth in EVs, data centers and the overall electrification of the economy, the need for demand-side flexibility has never been greater," said Raphael Declercq, CEO of PowerFlex. "Partnering with Cornell on AI-driven forecasting and control complements PowerFlex's platform approach to smart charging and vehicle-to-everything, helping integrate more renewables and enhance grid resilience alongside leading utilities and system operators."

Bitar added, "This support from the Bezos Earth Fund will help us meet this challenge by developing the algorithmic tools and infrastructure needed to turn millions of EVs into a flexible energy storage network that strengthens the grid instead of straining it."

When the Bezos Earth Fund launched its AI for Climate and Nature Grand Challenge last year, more than 1,000 organizations submitted proposals. From that global pool, only 24 teams advanced to the Phase I development stage, including several projects led by Cornell researchers. The project, "Harnessing the Power of Electric Vehicles as Flexible Grid Assets through AI-Driven Prediction, Optimization, and Control," will now build on that early success through the multi-year Phase II award from the Bezos Earth Fund.

Syl Kacapyr is associate director of marketing and communications for Cornell Engineering.