AI deals and embodied robotics push factory automation into a new era

Fanuc and Google announced a collaboration on May 19 to build smarter, more adaptive robots for factory applications. The two companies are working to apply Google's latest AI technologies across Fanuc's robotics lineup and advance its open platforms, according to Manufacturing Dive reporter Nathan Owens. The deal is one of several high-profile physical AI partnerships that emerged in May 2026 as manufacturers look to embed intelligence directly into production systems.

Partnership activity signals a strategic shift

The Fanuc-Google announcement was not an isolated move. Manufacturing Dive also reported that Kawasaki opened a Silicon Valley center dedicated to expanding physical AI collaboration between the U.S. and Japan. Separately, Stellantis is planning to work with Accenture and Nvidia on digital twin technology. Together, these moves point to a broader realignment: automakers, robotics companies, and technology giants are all converging on the factory floor.

Digital twins, which create virtual replicas of physical production systems, allow manufacturers to simulate and optimize operations before committing to hardware changes. Pairing that capability with AI from Nvidia positions Stellantis to iterate faster on production design without the cost and time of physical trials.

Rethinking how robots learn

While the headline deals focus on partnerships, a quieter shift in robot training methodology is also gaining momentum. Imitation learning is changing how industrial robots acquire new skills, moving away from hand-coded instructions toward learning by observing and mimicking human actions. Robotics Tomorrow published analysis from Anders Billesø Beck explaining that success with the technique depends heavily on data quality, force sensing, and the use of production-grade hardware rather than lab-grade equipment.

The distinction matters because many early robotics AI demonstrations took place in controlled settings that do not reflect the variability of a real factory. Getting imitation learning to work reliably at scale requires the robot to process rich sensory input, including force and torque feedback, not just visual data. That hardware requirement raises the integration bar but also narrows the gap between a trained model and a dependable production asset.

Embodied AI targets the margin problem

A separate thread in the industry conversation centers on what Robotics Tomorrow describes as the "great margin squeeze," the pressure manufacturers face from rising costs, labor shortages, and increasing product variety. Embodied AI-enabled robots, which combine perception, reasoning, and physical action in a single system, offer a path through that pressure by supporting high-mix manufacturing with faster changeovers.

The practical appeal is that fewer exceptions stop the production line, and the changeover burden that once required dedicated engineering time can be absorbed by the robot itself. Robotics Tomorrow noted that the approach allows manufacturers to take on more product variety without adding engineering headcount, a meaningful advantage when skilled labor remains difficult to source.

Sensing and integration round out the picture

On the sensing side, time-of-flight technology is drawing attention as a cost-effective option for 3D machine vision applications. Robotics Tomorrow reported on the technical rationale behind indirect time-of-flight sensors for cost-sensitive deployments, with IDS product manager Patrick Schick explaining where the technology fits relative to other 3D imaging approaches.

Integration complexity is a running theme across all of these developments. Robotics Tomorrow highlighted that combining robots with high-frequency welding machines, for example, requires careful planning around electromagnetic interference, safety interlocks, and cycle synchronization. The principle extends broadly: adding AI or advanced sensing to a production line only delivers results when the surrounding system is engineered to support it.

Automate 2026 is serving as a near-term showcase for many of these technologies. Companies including DESTACO are presenting robotic gripping, tool changing, and workholding solutions designed to improve integration flexibility, according to Robotics Tomorrow's preview coverage of the event.