Config, which is building a data layer for robotics foundation models, raised a $27M seed at a $200M+ valuation led by Samsung Venture Investment
Asia's push into physical AI is being fueled by the same manufacturing prowess that made the region a global industrial powerhouse.LinkedIn:ConfigLinkedIn:Config:We are excited to announce our $27M seed round, led by Samsung Ventures, at a valuation exceeding $200M. …
Context & Ripple Effects
Related coverage shows capital flowing into both general-purpose robotics foundation models and AI-powered industrial robots, while Covariant’s earlier rounds illustrate a longer-running effort to bring AI into industrial automation.
Config’s financing is notable because it targets the data-layer infrastructure around robotics foundation models rather than a single robot application, with Samsung Ventures providing a direct link to an established industrial technology investor.
First-order effects
- Config gains seed capital and backing from Samsung Ventures to build its data-layer product for robotics foundation models.
- Samsung Ventures becomes financially aligned with a startup positioned in the software and data infrastructure supporting physical-AI systems.
Second-order effects
- Robotics-model developers and industrial-robot vendors may face greater pressure to treat data infrastructure as a distinct part of their stack rather than an internal supporting function.
- Strategic investors with manufacturing ties may increasingly seek exposure to enabling layers of physical AI, alongside investments in robot builders and model developers.
Third-order effects
- If foundation-model approaches become more central to robotics, control of the data pipelines, formats, and feedback loops supporting them could become a consequential layer of industry power.
- The pattern points to physical AI developing as a stack: model companies, robotics applications, and specialized infrastructure providers may each attract separate funding and strategic partnerships.
The trend: Physical AI investment is broadening from AI-enabled robots themselves toward the data and software infrastructure required to train and deploy robotics foundation models.