Kawasaki and Nvidia’s New Robotics Center Aims to Turn Physical AI Into Real Machines

Kawasaki Heavy Industries and Nvidia are expanding their collaboration with a new robotics development center in San Jose, California, designed to accelerate the creation of AI‑powered robots that operate in the real world. The initiative combines Kawasaki’s robotics hardware with Nvidia’s software and computing stack for what the companies call “physical AI.”

The center will focus on developing practical robotics systems—especially in medical and mobility applications—while using advanced simulation and digital‑twin tools to train and test robots before they are deployed in physical environments.

The partnership: robotics hardware meets physical AI

The planned Silicon Valley facility will serve as a joint research and development hub where Kawasaki engineers work with Nvidia’s AI and simulation platforms to design next‑generation robotic systems.

The core concept behind the collaboration is physical AI, which refers to AI systems capable of perceiving, reasoning, and acting in the physical world through machines such as robots or autonomous devices.

By integrating AI directly with robotics hardware, the partnership aims to create machines that can understand their surroundings and carry out complex tasks autonomously in real environments.

Simulation and digital‑twin technology at the center of development

A major part of the initiative is Nvidia’s robotics development stack, which includes simulation and digital‑twin tools that allow developers to train robots in realistic virtual environments before deploying them in the real world.

Training robots in simulation has become a key technique in modern robotics because it allows developers to:

• generate large volumes of training data
• test robots safely in virtual scenarios
• accelerate development cycles before hardware deployment

This “sim‑to‑real” approach is a central strategy in Nvidia’s robotics platform, which includes tools for simulation, robot learning, and AI models designed specifically for embodied machines.

Testing AI on Kawasaki’s Corleo mobility robot

One early example connected to the collaboration is Corleo, a four‑legged personal mobility robot under development by Kawasaki.

Nvidia’s simulation technology is expected to be applied to Corleo to help train and validate its behavior in virtual environments before real‑world testing.

Corleo itself represents Kawasaki’s vision for future mobility systems. The robot uses a hydrogen‑powered energy system and is designed to traverse rugged terrain while responding to a rider’s body movements and providing feedback through an AI‑driven interface.

While reports link Corleo testing to Nvidia’s simulation platform, available information does not describe a detailed role for the robot inside the San Jose center beyond its use as an example of the technology integration.

Initial focus: medical and mobility robotics

The partnership will initially target robotics applications in healthcare and mobility, two areas where autonomous machines must safely interact with humans and complex environments.

Potential use cases include:

• medical support robots capable of handling materials or assisting clinical workflows
• mobility robots designed to transport people or operate in difficult terrain
• intelligent machines that can perceive and navigate dynamic environments

Developing these systems requires tightly integrating AI perception, control systems, and robotics hardware—one of the main reasons the companies are combining Kawasaki’s robotics expertise with Nvidia’s AI software stack.

Part of Nvidia’s broader push into physical AI

The collaboration also fits into Nvidia’s wider strategy to extend AI beyond digital software and into physical machines such as robots, vehicles, and industrial systems.

At recent industry events and product announcements, Nvidia has emphasized robotics platforms that combine:

• simulation environments
• AI models for robot learning
• high‑performance compute hardware

The goal is to enable developers and industrial partners to design robots in virtual worlds and deploy them faster in real environments.

Nvidia has already partnered with numerous robotics and industrial companies to scale this approach across manufacturing, logistics, and autonomous systems.

The broader robotics context in Japan

Kawasaki’s participation highlights Japan’s continuing role as a major robotics powerhouse. Japanese manufacturers have decades of experience building industrial robots and automation systems.

Recent surveys show that about one‑third of Japanese companies are already using or considering AI‑powered robots, with transportation and manufacturing companies among the most active adopters.

By combining Japan’s robotics engineering with Silicon Valley AI infrastructure, initiatives like the San Jose center aim to accelerate the transition from traditional automation to intelligent machines powered by physical AI.

What remains unclear

Some discussions around the project have mentioned additional technology companies and ecosystem partners. However, publicly available reports describing the San Jose center do not confirm specific contributions from companies such as Analog Devices, Microsoft, or Fujitsu to this particular initiative.

For now, the confirmed elements of the collaboration center on:

• a joint robotics development center in San Jose
• integration of Kawasaki robotics with Nvidia physical‑AI technology
• use of simulation and digital‑twin platforms for training robots
• initial development focused on medical and mobility applications.

As robotics development accelerates globally, partnerships like this one highlight a major trend: the merging of advanced AI software with physical machines capable of operating autonomously in the real world.