Physicl has raised US$10 million to help robots perform better in real-world environments, using technology originally developed to create realistic 3D product images for online stores.
The funding will support the company’s efforts to improve how robots are trained before being deployed in warehouses, factories and other environments.
The problem Physicl is trying to solve is simple: robots can perform tasks successfully in computer simulations but often struggle when faced with real objects and unpredictable conditions.
For example, a robot may learn to pick up a box in a virtual environment but fail when the actual box has a different weight, surface or position. This makes robot training expensive, as developers must spend time testing machines in real-world conditions.
Physicl believes more realistic simulations can help reduce this problem.
From Online Shopping to Robotics
The company’s technology has an unusual starting point. It builds on 3D visualisation tools originally developed for online retailers, allowing customers to view products in realistic digital settings.
Physicl is adapting that technology to create virtual environments where robots can practise handling objects and responding to different situations.
Instead of building every training environment from scratch, developers can use detailed digital objects and settings to test how their AI systems respond. This could reduce the amount of physical testing needed and help companies develop robots more efficiently.
Expanding Its Platform
Alongside the funding announcement, Physicl introduced Studio, a platform that allows developers to create customised training environments.
It includes more than 10,000 3D environments and 100,000 assets. Developers can adjust details such as lighting, camera angles and object placement to create different training scenarios.
The company has also worked with SpAItial, an AI company developing technology that turns images into interactive 3D environments. The collaboration demonstrated how Physicl’s training data could help improve the accuracy of AI-generated environments.
Building for the Future of Robotics
The US$10 million funding round was led by Inovia Capital, with participation from US Venture Partners.
As more businesses explore robotics and physical AI, the need for reliable training technology is growing. Robots must be able to recognise objects, understand their surroundings and respond to situations that may not have been included in their original training.
Physicl is positioning its platform to support that development by making realistic digital training environments easier to create.
Its approach brings together two industries that do not traditionally overlap: e-commerce visualisation and robotics. The company now aims to turn that combination into a practical tool for businesses developing the next generation of intelligent machines.