Physicl raises $10M seed round to build simulations that robots can learn from
Physicl, founded by Alex de Vigan, raised $10 million led by Inovia Capital and launched Studio, a tool for making synthetic training data for robot AI.
Physicl, a startup run by founder and CEO Alex de Vigan, has raised $10 million in seed funding. Inovia Capital led the round. The company wants to help robot builders and makers of world models train their systems in digital scenes that follow real-world physics more closely.
De Vigan previously spent about ten years at nfinite, which turned product goods into 3D models for online shops. When generative AI took over that market, he saw that the same pipeline, which turns messy real-world information into 3D with correct geometry, materials and physics, could serve physical AI instead. Physicl's first product, Studio, offers 10,000 3D environments and 100,000 object assets. Users build scenes to create synthetic data for narrow problems, such as a robot that struggles to spot a cup on a cluttered counter or in shadow.
SpAItial, a world model company, is an early customer. Its latest model, Echo-3, came out last week.
Why it matters
Robot training is held back by costly remote-operation data and by video that teaches too little about handling objects. Simulation could help, but only if it matches reality. Physicl is entering a crowded field, and de Vigan himself calls it young and blurry. The company has not said how it will prove its data improves real robots, or what it charges. Watch for named robotics customers and results from real-world tests.
