Nvidia and Kawasaki Heavy Team Up to Build an AI-Powered Shipyard in Japan

Kawasaki Heavy Industries and Nvidia are partnering to build what the two companies describe as a next-generation, AI-powered shipyard in Japan, aimed squarely at a labor shortage that has left Japanese shipbuilders short of the skilled welders needed to keep production on schedule.

The Technology Behind the Plan

The centerpiece of the partnership is Nvidia's digital twin technology, which lets Kawasaki simulate and refine fabrication and assembly workflows virtually before deploying them on the shipyard floor. That approach is meant to compress the trial-and-error period that normally comes with introducing new automated processes into a live industrial environment, where mistakes are costly and downtime is expensive.

On the robotics side, Kawasaki is adapting technology from its Corleo quadrupedal robot platform into an AI image-recognition welding robot, first unveiled separately in April, that the company says can double welding productivity in shipbuilding while improving worker safety. Deploying a legged robotic platform in this context matters because shipyard welding often happens in tight, uneven, or elevated spaces, including scaffolding and confined hull sections, where wheeled equipment struggles.

A Broader Physical AI Push

The shipyard project sits inside a larger collaboration the two companies previously announced: a Kawasaki Physical AI Center in San Jose, California, built to develop and test next-generation robotics using Nvidia's simulation and AI tools. That center's initial focus areas span healthcare and nursing care robotics, mobility applications, and shipyard automation, with shipbuilding treated as the most immediate commercial priority given Japan's demographic pressures on its skilled trades workforce.

Why It Matters

Japan's shipbuilding industry has struggled for years with an aging, shrinking pool of certified welders, a problem that has become more acute as global shipbuilding demand, including for LNG carriers and next-generation vessels, has climbed. Automating welding in a shipyard is a materially different engineering problem than automating welding on an automotive assembly line, since ship hull geometry varies far more than a standardized car body, which is part of why Kawasaki's approach leans on AI image recognition to adapt weld paths in real time rather than relying on fixed, pre-programmed robot arms.

If the partnership delivers on its productivity claims, it would give Japanese shipbuilders a domestically developed alternative to simply shrinking output or losing contracts to shipyards in South Korea and China, both of which have invested heavily in yard automation over the past decade. The project also reinforces a broader theme in Japanese industrial policy this year: a roughly $65 billion public-private commitment to physical AI investment by 2040, aimed at helping Japanese manufacturers close a widening automation gap with Chinese and European competitors.

This report is based on company statements and reporting from Nikkei Asia.

This article is for informational purposes only and reflects publicly available information at the time of writing.