CarbonSix Inc., a developer of deploy-ready robotic artificial intelligence solutions, announced Wednesday it raised $40 million in early funding to change how manufacturing works.

The company focuses on delivering robotic intelligence hardware and software that manufacturing lines can implement immediately in real-world operations.

Getting AI robotics up and running successfully inside manufacturing requires massive amounts of data and learning capabilities. To remain competitive in this industry, CarbonSix uses what it calls a “data flywheel” rather than generalized datasets for robotic AI foundation models.

As robotics and intelligence work together in the field, high-quality, task-specific data is captured back into the system. The company said this creates a powerful self-reinforcing loop that drives data generation and refinement, delivering smarter automation tools on the factory floor.

“From the very beginning, our goal has never been about building technology for technology’s sake — it has been about creating practical, field-ready physical AI that drives measurable bottom-line results for manufacturers,” said Chief Executive Tae-yeon Terry Moon.

Physical AI, the integration of AI into hardware, enabling machines to perceive, reason about and physically interact with the real world, is changing how intelligent software is deployed in the real world. Common examples include autonomous vehicles, humanoid and industrial robots, drones and medical devices.

This shift in technology is swiftly moving onto factory floors and into warehouses.

In keeping with that transformation, CarbonSix has assembled deep domain expertise to build its platform, including Moon, formerly of SuaLab, an industrial AI vision firm acquired by the American maker of machine vision systems, Cognex Corp. Chief Technology Officer H.J. Terry Suh, a Ph.D. graduate from the Massachusetts Institute of Technology, leads the development of the company’s robotic intelligence framework, and Chief Hardware Officer Je-hyeok Kim, a former Yale postdoc, provides expertise in state-of-the-art hand and manipulator hardware design.

Together, they have developed SigmaKit, a collection of tools that includes AI software, robotic hands, teaching devices and sensor modules. It can be deployed easily in factories and manufacturing lines without deep expertise in robotics or AI. It’s designed to be simple and requires no complex coding. Users can quickly teach tasks to robots intuitively by having robots capture operator know-how with minimal effort.

Users repeatedly demonstrate target tasks to robots using teleoperated motions, captured as image and motion data. This builds training datasets. Depending on the task, a model can be generated and deployed in less than a day. SigmaKit makes training efficient and allows users to quickly filter, review and edit the collected data.

Once trained, skills can be deployed directly to unstructured tasks. The company said that even when scaling across production lines or additional sites, skills require only minimal adjustments, allowing for mass deployment.

The Series A funding round was co-led by DSC Investment and LB Investment. New investors joining in the round included IMM Investment, Korea Development Bank, SV Investment, Cortentia and A Squared U.S. All existing seed-round investors fully participated in the follow-on investment, including Foothill Ventures, Storm Ventures, Zeitgeist Capital, Xquared and CarbonBlack Fund.

The company said it plans to use the new funding to hire new talent, scale its infrastructure and expand globally as more factory floors are transformed with smart machines.

Image: CarbonSix