FANUC and Google Develop AI Welding Robots That Can Read Blueprints

FANUC and Google have developed an AI-powered robotic welding system capable of reading engineering blueprints and generating the instructions needed to carry out welding work, reducing the need for conventional robot programming.

The system uses Google Gemini Enterprise together with FANUC’s industrial robot technology and is scheduled to begin deliveries in December 2026.

AI Turns Blueprints Into Robot Instructions

Traditional industrial robots generally need engineers or operators to programme the movements required for each manufacturing task.

FANUC and Google’s new system is designed to simplify that process by allowing AI to interpret blueprint information and determine the actions required for welding.

By reducing manual programming requirements, the technology could make robotic welding easier to deploy where production requirements change frequently or engineering resources are limited.

Gemini Enterprise Connects AI With Industrial Robots

The welding system builds on a wider collaboration between FANUC and Google around Physical AI.

FANUC announced the partnership earlier in 2026, combining Google’s AI technologies and development platforms with its industrial robot control systems.

At FANUC’s Open House, the companies demonstrated an autonomous kitting application in which an AI agent powered by Gemini Enterprise coordinated an industrial robot and a collaborative robot.

The blueprint-reading welding system extends that approach into another production process.

FANUC Pushes Open Platform for Physical AI

FANUC is developing its robots around a more open software architecture that can connect with external AI systems.

Its robots support technologies including ROS, Python and external high-speed communication interfaces, allowing AI applications to interact more directly with robot controllers.

The company has said these capabilities are intended to help robots interpret their surroundings, make decisions and perform tasks with greater autonomy.

Welding Could Become Easier to Automate

Welding has long been a major application for industrial robots, particularly in automotive and metal manufacturing.

However, programming robots for individual parts and welding paths can add engineering time when products or production runs change.

A system that can interpret engineering drawings directly could reduce some of that preparation work and make robotic welding more flexible.

FANUC already supplies collaborative and industrial robots for welding applications, while its CRX collaborative robot range supports functions including vision and simulation tools for automated production.

Deliveries Planned From December

The companies plan to begin delivering the new AI welding system in December 2026.

The launch reflects a wider shift in factory automation, where generative AI is beginning to move beyond planning and analytics into direct interaction with production equipment.

For manufacturers, the significance lies in reducing the amount of specialised programming required to introduce or modify robotic processes.

FANUC and Google’s system shows how AI agents, engineering data and industrial robots can increasingly be connected to automate tasks that previously required more manual setup.