Physicl raises $10M to make the digital world real enough for robots

Why it matters

Physicl, founded by Alex de Vigan, has secured seed funding from Inovia Capital to bring realistic simulation technology to the robotics and physical AI sector. De Vigan previously spent years building nfinite, a company that rendered 3D product imagery for brands and retailers, and he sees that same rendering pipeline as a solution to a persistent training problem: researchers building physical AI models have struggled with training on video and with the expense of gathering real-world data through remote operation. Simulated environments are considered promising, but making them faithful to real-world physics has proven hard. De Vigan believes nfinite's pipeline for turning unstructured real-world information into high-fidelity physical representations can help. Physicl enters a crowded, young market of startups supplying data and tooling to robot and foundation model builders, and it is releasing Studio, which lets users build scenes from a large library of environments and assets to generate synthetic training data. The company's focus is post-training models on difficult edge cases, positioning itself as ongoing tooling rather than a one-off data provider.