Comau Partners with NVIDIA to Drive Next-Generation Physical AI in Industrial Robotics

• Comau is integrating NVIDIA Physical Al simulation libraries and accelerated computing into its products and system design processes to accelerate the development of increasingly intelligent and autonomous robotics solutions
• The integration unites Comau's industrial robotics and automation expertise with NVIDIA edge Al, physical Al simulation and accelerated computing
• Concrete results include the integration of NVIDIA technology with Comau next-generation robotic products, and simulation tools

Comau is boosting the development of Al-enabled robotics and automation solutions for industrial applications with NVIDIA technologies. By combining Comau's deep experience in robotics with NVIDIA physical Al simulation libraries and accelerated computing, the companies are paving the way for the deployment of increasingly intelligent and autonomous robotic systems. As part of this collaboration, Comau is integrating NVIDIA technologies across its product portfolio and design processes, specifically featuring:

• Comau industrial robot controllers powered by NVIDIA Jetson edge compute;
• The integration of the Comau robot portfolio into NVIDIA Isaac Sim;
• The utilization of the NVIDIA software stack to optimize vision-based features and control strategies for intralogistics products.

The next generation of Comau industrial robot controllers will be combined with NVIDIA technology to provide the robust computing power required for advanced Al applications. The new architecture is designed to support demanding, Al-driven workloads, enabling smarter, more adaptive, and highly autonomous robotic solutions tailored to diverse applications and customer needs.

Furthermore, Comau is integrating its industrial, collaborative, and mobile robot portfolio into NVIDIA Isaac Sim - an open source framework built on NVIDIA Omniverse libraries for robotics simulation, testing, and synthetic data generation.

Comau is also contributing SimReady OpenUSD robot assets to SimReady Central Library on Hugging Face. The library helps developers discover and reuse validated assets, reducing preparation time for building and testing robotic applications.

NVIDIA's simulation and Al libraries are also supporting the development of intelligent, vision-based applications and the optimization of control strategies for intralogistics equipment. This includes training Al models for Comau's proprietary Ml.RA vision system family, utilizing reinforcement learning for advanced palletizing and depalletizing tasks. By using NVIDIA Isaac Sim's high-fidelity physics, Discrete Event Simulation (DES) for logistical workflows, and NVIDIA Isaac Lab's advanced learning capabilities, Comau is creating ultra-precise virtual models. These models are used to train Al-driven control strategies before they ever reach the factory floor. Developing and validating capabilities within these photorealistic, physics-based simulated environments allows Comau and its customers to drastically accelerate deployment in real-world industrial applications.

Andrew Lloyd, Chief Engineering Officer of Comau, states: "The combination of NVIDIA Physical Al simulation libraries and accelerated computing with Comau's platforms and automation expertise will help us develop intelligent solutions faster and make advanced automation easier for our customers globally to deploy across different applications and industrial sectors."

"Physical Al is bringing new levels of intelligence and adaptability to industrial robotics," said Paolo Guglielmini, vice president EMEA at NVIDIA. "By combining its robotics expertise with NVIDIA Isaac Sim and Isaac Lab, Comau can simulate, train and validate robotic applications before deployment, helping manufacturers worldwide accelerate the adoption of Al-powered automation."

www.comau.com