FANUC CORPORATION announced a collaboration with Google on May 13 to integrate Gemini Enterprise, Google's enterprise-grade generative AI, into its Physical AI Robot System. The partnership enables an AI agent to interpret natural-language instructions and direct industrial robots, with collaborative and non-collaborative units operating together as a single cell.
The system builds on FANUC's existing open-platform architecture. FANUC robots already support ROS through open-source drivers, and Google contributes to ROS through its Intrinsic robotics AI group. The new arrangement extends full Intrinsic platform support across the entire FANUC lineup, from 3 kg payload units to 2.3-ton machines, including the CRX collaborative series. Intrinsic's Flowstate development environment, interoperable with ROS, is designed to speed AI solution development using FANUC's open interfaces.
What's new here is the cell-level coordination model. An AI agent built with Gemini Enterprise can receive a plain-language instruction and dispatch tasks across mixed robot cells simultaneously, rather than relying on fixed-path programs for each unit. Mike Cicco, President and CEO of FANUC America, said the collaboration aims to let customers "take on more complex, variable production while maintaining the reliability and performance that production environments demand."
FANUC is also participating in Google DeepMind's "Gemini Robotics Trusted Tester Program," a research effort targeting foundational robotics models for AI. The company has shipped more than 1,000 robots for Physical AI-related applications since the system debuted at the International Robot Exhibition in Tokyo last December, with demand continuing to grow.
FANUC demonstrated the Gemini-integrated system at its New Product Open House Show in May, where visitors could issue natural-language commands and observe autonomous robot execution in real time. The partnership reflects mounting pressure on robot manufacturers to reduce programming overhead in high-mix production environments, where task-specific code slows changeover. How the Gemini Enterprise agent performs under production-floor conditions, including latency and uptime requirements, has not been detailed in public materials.
