GM, Ionate deploy smart grid transformer at Michigan plant

General Motors and Ionate, a UK-based deeptech company, are deploying Ionate’s Hybrid Intelligent Transformer (HIT) technology at GM’s manufacturing plant in Romulus, Michigan. The collaboration marks Ionate’s entry into the US market, following earlier projects in Portugal, Spain, and Austria.

The HIT technology is designed to identify and resolve power quality issues in real time, providing visibility into power conditions at sub-millisecond resolution while operating at more than 99% efficiency, according to Ionate. The technology will be installed in the coming months at GM’s roughly 900,000-square-foot Romulus facility, which produces 10-speed transmissions for the automaker’s full-size trucks and SUVs.

GM has invested a combined $600 million in two tranches since November 2025 to expand transmission capacity at the Romulus facility.

Ionate said grids using its technology can accommodate 33% more distributed energy resources, carry 25% more power without upgrading poles and wires, and reduce wasted power by 6%, based on the company’s own modeling of average deployment scenarios.

“Maximizing efficiency and resiliency is crucial in GM’s manufacturing system, and it’s why we’re implementing innovative technology to enhance our energy management and the overall grid,” said Luis Cervantes, GM’s executive director of manufacturing excellence.

Each HIT unit also functions as a control node coordinated through Ionate’s Aurora software platform, the company said. Ionate said this allows individual deployments to contribute to broader grid capacity and resilience as networks scale. Ionate was founded in 2019 and has previously partnered with EDP/E-REDES in Portugal and Spain and with Wien Group/Smartworks in Austria.

Grid connection bottlenecks of the kind Ionate says its technology can ease have also slowed battery storage deployment in markets including Germany, the United Kingdom, and the United States, where developers increasingly face multi-year interconnection queues.