Comau Integrates NVIDIA Technology Into Industrial Robot Products and Design Process

On October 21, 2026, Comau announced it would integrate NVIDIA technology into its products and system design process. Its next-generation industrial robot controllers will use NVIDIA Jetson edge computing, and it will bring its industrial, collaborative and mobile robots into NVIDIA Isaac Sim. For logistics equipment, it will also use NVIDIA’s software suite to optimize vision capabilities and control strategies.

Comau described a setup for training and validating robot control strategies in a virtual environment before applying them in factories. It says reusing SimReady assets can reduce development preparation time. For those considering automation with industrial robots and logistics equipment, the effort connects simulation before deployment on physical systems with product design, although no quantitative figures for its effects have been disclosed.

A notable feature of this setup is that NVIDIA technology will be used not only in robot controllers but also in design, training and validation in virtual environments. Jetson will provide edge computing for next-generation controllers, while Isaac Sim and Isaac Lab will be used to work with robots and logistics processes in virtual environments. The effort also includes training for MI.RA’s vision capabilities and applications such as palletizing, connecting computing infrastructure, simulation and control development in a single process. However, the announcement did not specify the Jetson model, controller performance figures or validation results on physical systems. A related article from this newspaper, Agility Robotics Standardizes Digit 5’s Onboard Computing Platform on NVIDIA IGX Thor, covered the onboard computing platform of a specific humanoid robot. Comau’s announcement, by contrast, pairs the use of Jetson in industrial controllers with efforts to integrate multiple robot categories into a simulation environment. The two efforts involve different robots and computing products, and there is no information with which to compare their performance. The shared use of NVIDIA technology alone does not mean they use the same implementation approach. Compared with AWS Releases a Toolchain for Robot Development, the distinction is between making development tools available and a robot maker integrating computing and simulation into its products and design process. Comau says it will contribute SimReady assets to a shared library. The announcement describes asset reuse and reduced preparation time, but does not quantify a reduction in the time needed to deploy systems in the real world. The accuracy of transferring work from virtual environments to physical systems, and the scale of any additional adjustments required, will need to be assessed separately. The broader industry implication is that preparing simulation assets and training and validation environments could become part of the automation deployment process, alongside selecting the robot itself. Comau says a larger pool of reusable assets could reduce preparation burdens, but it did not specify which assets will be included in the library or how widely they can be reused across different robots and processes. Further details to watch include the specific Jetson configuration, the products targeted and their application timelines, real-world evaluations of control strategies developed in simulation, and performance metrics after deployment.

Comau will integrate its robot lineup into the open-source Isaac Sim, which is used for robot simulation, testing and synthetic data generation. It will contribute OpenUSD robot assets in SimReady format to Hugging Face’s SimReady Central Library. Comau says the aim is to make it easier for developers to find and reuse validated assets. In logistics, the company will develop AI models for its MI.RA vision system and use reinforcement learning for palletizing and depalletizing. It says it will use Isaac Sim’s physics simulation, discrete-event simulation (DES) for logistics workflows and Isaac Lab’s training capabilities to train and validate control strategies in virtual environments before applying them in factories.