Shipbuilding is about to get a new kind of assistant — not human, but robotic.
Researchers from the University of Michigan Engineering and the Massachusetts Institute of Technology are working with Japanese partners on an AI-powered robotics project designed to cut down one of the industry’s most expensive problems: late-stage construction errors that lead to rework, delays, and spiralling costs.
The programme is backed by a $6.2m grant from Japan’s Ministry of Land, Infrastructure, Transport and Tourism and is being overseen by Nippon Yusen Kaisha’s Monohakobi Technology Institute. It will run through early 2027.
The issue it targets is well known inside shipyards. When pipes, cables or equipment are installed out of sequence, it often means later components no longer fit as intended. The result is time consuming redesign work and expensive project delays that ripple through the entire build schedule.
To prevent that from happening, the project will deploy autonomous robots fitted with LiDAR and high-resolution cameras to scan vessels while they are still under construction. These scans will then be continuously compared with the vessel’s digital twin — the original design model — to spot any mismatch between plan and reality.
Instead of reacting after mistakes are discovered, the idea is to flag them early, while they are still easy and cheap to fix.
The system is not meant to replace shipyard workers, but to support them. When the AI detects an issue, it will suggest possible fixes, explain the trade-offs of each option, and highlight cases where human judgment is still required.
“We want to build a co-pilot system that uses AI and robotics to take some of the detective work off workers’ shoulders,” said Alan Papalia, assistant professor of naval architecture and marine engineering at the University of Michigan.
The technology will be tested using a reconfigurable shipbuilding test block capable of simulating different construction and outfitting scenarios, allowing the researchers to evaluate how the system performs in realistic yard conditions before wider adoption is considered.
