NVIDIA and Robotics Leaders Advance Physical AI for Real-World Deployment

NVIDIA is working with the global robotics ecosystem, including robot brain developers, industrial robot makers and humanoid specialists, to advance production-scale physical AI. The company also introduced new NVIDIA Isaac simulation frameworks, along with NVIDIA Cosmos and NVIDIA Isaac GR00T open models, to help the industry develop, train and deploy next-generation intelligent robots.

Industry players using the NVIDIA platform include ABB Robotics, AGIBOT, Agility Robotics, FANUC, Figure, Hexagon Robotics, KUKA, Skild AI, Universal Robots, World Labs and YASKAWA.

The company highlighted that physical AI is becoming central to industrial operations, with robotics expected to play a key role across factories, logistics, transport and infrastructure. Its full-stack platform, covering computing, software frameworks and open models, is designed to support this transition and connect global robotics players.

Advancing robotic simulation and validation

As robotics becomes more AI-driven, manufacturers require accurate and high-fidelity simulation to design and optimise systems before deployment. With more than two million robots installed globally, companies such as FANUC, ABB Robotics, YASKAWA and KUKA are integrating NVIDIA Omniverse libraries and Isaac simulation frameworks into their virtual commissioning tools. This enables them to develop and validate complex robotic applications and entire production lines using digital twins.

They are also integrating NVIDIA Jetson modules into controllers to enable real-time AI processing at the edge.

Developing general-purpose robot intelligence

Robotics is shifting from task-specific machines to more adaptable systems capable of handling multiple functions while maintaining industrial precision. Developers such as Field AI and Skild AI are building generalised robot intelligence using NVIDIA Cosmos models and Isaac simulation tools. These technologies support data generation and allow robots to learn new tasks with limited retraining.

NVIDIA also introduced Cosmos 3, a new world foundation model that combines synthetic data generation, vision reasoning and action simulation to support advanced robot intelligence in complex environments.

Supporting humanoid robot development

Building humanoid robots remains a major challenge, requiring advanced AI, perception and real-time control. Companies including Boston Dynamics, Figure, Hexagon Robotics and others are using NVIDIA’s Cosmos models and Isaac platforms to accelerate development and validation.

NVIDIA launched Isaac Lab 3.0 in early access, enabling faster large-scale robot training using DGX-class systems. It includes improved physics simulation and better support for complex robotic movements. The company also expanded its GR00T models, designed to provide robots with general-purpose skills, supporting real-world industrial deployment.

Expanding into healthcare robotics

Healthcare is emerging as a key area for physical AI, particularly in surgery and medical systems. CMR Surgical is using NVIDIA simulation tools to train robotic systems before clinical use, while Johnson & Johnson MedTech and Medtronic are applying these technologies to improve robotic-assisted procedures and ensure safety.

Strengthening global robotics collaboration

NVIDIA is building an open platform to support the design, training and deployment of robotics systems, helping to scale adoption across industries. Companies such as Skild AI are collaborating with ABB Robotics and Universal Robots to deploy shared robot intelligence across applications.

Other collaborations include Foxconn using AI-driven robotic systems for precision manufacturing, Samsung improving robotic assembly processes, and PTC enabling design-to-simulation workflows for robotics development.

Logistics provider KION Group is working with NVIDIA and Accenture to develop AI-powered warehouse systems, including autonomous forklifts. Meanwhile, cloud providers such as Microsoft Azure, Nebius, CoreWeave and Alibaba Cloud are integrating NVIDIA’s physical AI technologies to support robotics development.

Supporting innovation and startups

Through its Inception programme, NVIDIA is supporting over 40,000 startups by providing access to its AI tools, computing resources and partner ecosystem. It has also partnered with Hugging Face to integrate robotics tools into open-source platforms, connecting millions of developers worldwide and accelerating innovation in robotics.