Key Takeaways:

• Physical AI moves robots beyond pre-programmed tasks into real-time, on-the-fly decision making.
• Companies like FANUC, Rockwell Automation, ABB, and NVIDIA are already deploying it at scale across industries.
• The shift is happening now, from warehouse floors to hospital wards, and it represents a meaningful evolution in how the global economy runs.
• For most of its history, a robot’s job description was pretty simple: do this thing, in this order, every time. No improvising. No adapting. Definitely no thinking. That’s changing fast.

Physical AI is the idea that robots don’t just execute instructions anymore. They perceive their environment, process what they’re seeing in real time, and make decisions on the fly. At NVIDIA’s 2026 GTC Conference, CEO Jensen Huang put it plainly: every industrial company will become a robotics company. That’s not a prediction. It’s a description of what’s already underway.

On the Warehouse Floor: Robots That Read the Room

Warehouse automation has been around for a while. But the older version of it was rigid. A robot knew its route, knew its task, and couldn’t do much if something unexpected showed up. That’s the automation of the past.

At MODEX 2026 in April, FANUC America demonstrated what the next generation looks like. Their CRX-30iA mobile manipulator works alongside the OTTO 600 autonomous mobile robot from Rockwell Automation in a fully integrated system built for warehouse and distribution environments. What makes it different from traditional automation: AI-driven perception. The system uses payload estimation, box-locating vision, and barcode-based decision logic to automatically adapt to varied box sizes, pallet conditions, and routing requirements without being told what to do in advance.

Sources: CRX-30iA by Fanuc. OTTO by Rockwell Automation.

Think about what that means in practice. A pallet of mixed-size boxes shows up. The robot scans, assesses, figures out what it’s dealing with, and adjusts. No human required. That’s the Physical AI difference.

And it’s not just FANUC and Rockwell. Across the broader robotics landscape, the pattern is the same: AI perception layers being added to hardware that used to run on fixed logic. The warehouse is becoming one of the most interesting testing grounds for what Physical AI can do at scale.

In the Hospital: Less Legwork, More Care

Hospitals are facing a serious nursing shortage, and a big part of the problem isn’t clinical work. It’s the physical grind of logistics: delivering supplies, transporting specimens, restocking rooms. Important tasks, but not exactly why someone went to nursing school.

ABB has been working directly on this problem. The company deployed its collaborative YuMi robot at Karolinska University Hospital in Sweden to automate manual lab tasks, reducing physical strain on staff and improving the overall working environment for clinical teams. It’s a practical, unglamorous application of robotics that quietly makes a real difference. Other players in the space, like Diligent Healthcare with its Moxi robot, are tackling similar challenges in U.S. hospitals, navigating hallways autonomously and handling supply runs so nursing staff don’t have to.

Sources: YuMi by ABB. Moxi by Diligent Healthcare.

What makes this category of robot possible at this level is the AI infrastructure underneath. NVIDIA’s Isaac for Healthcare platform provides the real-time sensor processing, simulation, and training capabilities that allow robots to operate safely and intelligently in complex, unpredictable environments like a hospital ward. It’s the kind of foundational plumbing that tends to get overlooked until you realize nothing works without it.

The common thread across both industries: Physical AI isn’t replacing human judgment. It’s taking over the tasks that never required human judgment in the first place.

What This Means for Investors

Physical AI isn’t a concept waiting on a breakthrough. The infrastructure is being built, the deployments are live, and the companies leading the charge are already operating at scale. FANUC has over one million robots installed globally. Rockwell Automation’s AMR systems are running in warehouses right now. ABB’s collaborative robotics are deployed across healthcare and industrial settings in more than 100 countries. NVIDIA’s Isaac platform is the AI backbone powering a growing number of these deployments across industries.

For investors looking for exposure to this shift, the VanEck Robotics ETF (IBOT) provides targeted access to the companies at the center of it: industrial robotics leaders, AI hardware providers, and the automation platforms connecting them. As Physical AI moves from demo floor to factory floor to hospital corridor, the companies driving that transition are where the opportunity sits.