Formic raises $50 million to scale physical AI across US manufacturing
Mercato Partners led the Series B for the robots-as-a-service platform, which runs hundreds of robots across more than 100 US factories.
Formic Technologies has raised $50 million in growth capital to scale its physical AI platform across US manufacturing. The financing, announced on October 22, combines a Series B equity round led by Mercato Partners with debt financing from S2G Investments. Light Ray Ventures, Lightscape Partners and Citi joined the equity round, alongside other new and existing investors.
The Chicago-based company sells robots as a service: equipment, deployment software, field operations, maintenance, financing and performance monitoring bundled into a single recurring monthly payment, with no upfront capital required from the customer. Formic reports more than 750,000 hours of live production across hundreds of robots in more than 100 facilities in over 30 states, and says it now deploys robots into production about three times faster than it did a year ago.
The money will fund the move from task-specific automation into general-purpose robotics, strengthen Formic's AI and robotics capabilities, and open new applications such as metal fabrication, while deepening its presence in aerospace, space, defense, medical and energy. Formic reports 99.6% fleet uptime and a 97% customer renewal rate. Ryan Sanders, a managing director at Mercato Partners, joins Formic's board as part of the financing.
The round resurfaced on November 2, when Humanoid Global Holdings, a minority shareholder in Formic, highlighted the financing in a release of its own. Formic is also the deployment partner rolling out Physical Intelligence's Pi robot foundation models across factories. The bet is on the deployment layer: factories want robots without the capital cost or the integration work, and a 97% renewal rate says the model sticks once it is in. The open question is whether a services business can fund the jump to general-purpose robots, where the models, not the deployment, are the hard part.