Physicl Raises $10 Million Seed Round, Launches 10,000 Robot-Simulation Environments
Updated
Updated · TechCrunch · Nov 6
Physicl Raises $10 Million Seed Round, Launches 10,000 Robot-Simulation Environments
3 articles · Updated · TechCrunch · Nov 6
Summary
$10 million in seed funding from Inovia Capital will help Physicl expand software that generates high-fidelity simulations for training physical AI and robots.
The startup repurposes technology from founder Alex de Vigan’s previous 3D-rendering company, aiming to turn messy real-world image and sensor data into environments with accurate geometry, materials and physics.
Studio, Physicl’s first product, launches with 10,000 3D environments and 100,000 assets so customers can create synthetic data for edge cases such as cluttered or shadowed objects.
SpAItial, an early customer, used Physicl’s data to improve fidelity for its Echo-3 world-model product, underscoring demand for tools that can generate training data on the fly instead of curating it manually.
Physicl enters a crowded robotics-data market but is betting ongoing synthetic-data generation—not just one-off datasets—will become core infrastructure as model builders scale to trillions of images and billions of training permutations.