Physicl raises $10 million to build realistic simulations for robot training

$10 million seed round led by Inovia Capital

Paris-based startup Physicl has closed a $10 million seed funding round led by Inovia Capital, with participation from U.S. Venture Partners. The company develops high-fidelity 3D simulations to generate synthetic training data for robots and physical AI systems.

Studio platform targets edge cases in AI training

Physicl’s flagship product, Studio, launched with 10,000 3D environments and over 100,000 assets. The platform allows researchers to recreate edge cases—unusual scenarios like poor lighting or cluttered spaces—where AI models often fail. By generating targeted training data, Studio helps improve model accuracy for robots, self-driving cars, and 3D vision systems.

According to The Next Web, SpAItial, a Munich-based startup, reduced its room-building model’s error rate by more than half after incorporating Physicl’s synthetic data into its training set. This demonstrates the platform’s potential to address specific weaknesses in AI models.

From e-commerce 3D modeling to physical AI tooling

Alex de Vigan, Physicl’s founder and CEO, previously built nfinite, a 3D modeling company for e-commerce that raised $100 million in 2022. De Vigan shifted focus to generative AI and simulation technology after recognizing a gap in AI’s ability to understand and manipulate physical spaces.

Physicl’s approach distinguishes itself by moving beyond static datasets. The company provides tooling for ongoing synthetic data generation, enabling customers to convert large volumes of real-world data—such as images or LIDAR captures—into dynamic 3D training scenarios. TechCrunch reports that this agentic workflow could automate data curation, reducing the need for manual intervention.

Growth plans and industry adoption

Physicl’s customer base includes major AI labs and universities such as Meta, Nvidia, Epic Games, Amazon Web Services, and the Technical University of Munich. The startup’s rapid adoption by industry leaders highlights the demand for high-fidelity synthetic data in robotics and AI training.

The $10 million funding will support team expansion and further development of the Studio platform. Physicl joins a growing wave of European startups working on physical AI, a sector gaining traction as robots and AI systems increasingly operate in real-world environments.

Related on Lazyfounder

• Meta
• Nvidia
• More startup news