Robots don’t care about your lobby.
They don’t care about your parking ratio, your break room, or your view of the skyline. But they care — deeply, it turns out — about your floor. Your ceiling height. Your power supply. Your HVAC system. And if you’re the building owner, developer, or corporate decision-maker who doesn’t understand that yet, you’re about to get an expensive education.
I’m in the middle of one right now.
An Assignment That Changed How I Think About Buildings
I’m working on a deal that I can’t say much about — and that’s the point. A client in the advanced robotics space is evaluating the Charlotte region for a manufacturing facility. I’ll leave the details there, because that’s all I should share.
But I will tell you this: when I sat down with their engineering team for the first time and they walked me through what their production process demands from a building, I realized I was in a different world. The power requirements alone would make a conventional spec building tap out. The floor tolerances are measured in fractions of an inch — because the robots calibrate against the surface they operate on. The HVAC system has to hold environmental conditions tighter than most hospital operating rooms. The building requirements are more similar to an Intel chip fabrication plant than an Amazon warehouse.
This isn’t a tenant looking for four walls and a loading dock. The building is part of the machine.
That’s a sentence I’d never thought about before this assignment. And it’s the sentence that should matter to anyone making real estate decisions in the next decade.
It’s Already Happening 75 Miles from Charlotte
If you think this is theoretical, drive southeast for an hour.
In Cheraw, South Carolina — a town of 5,000 that lost much of its textile industry after NAFTA and has watched factories close for decades — a humanoid robot named Digit is working an eight-hour shift at a Schaeffler auto-parts plant. A year ago, a person did that job. Now Digit walks baskets of bearing components from a stamping press to a washing machine conveyor, four hours on, recharge over lunch, four hours more. The worker who had the job was moved to a higher-skilled inspection role.
Digit doesn’t call in sick, is a great value, and doesn’t have opinions about politics that get under my skin. If I’m being honest, that describes half the….well, nevermind. I get threats when I pick fights with brokers.
Schaeffler’s Cheraw plant employs 750 people. Entry-level positions start at $20 an hour. The robot’s operating cost? $10 to $25 an hour over its lifetime — and its maker, Agility Robotics, says that could eventually fall to $2 or $3. McKinsey estimates fewer than 200 humanoids are working in factories worldwide today. By 2040, that number could reach 5 million. Schaeffler alone plans to have hundreds of humanoids deployed across its global plants by 2030.
Here’s what makes this a real estate story, not just a labor story: Digit currently works inside a Plexiglas cage because it can’t yet detect nearby humans — a federal safety requirement. A new model coming later this year will have that capability, meaning the robots can work alongside people without barriers. When that happens, the floor plan changes. The safety zoning changes. The power and data infrastructure changes. The building changes.
And good luck getting that building approved. Try walking into a public meeting in a town like Cheraw — a place that’s already been through economic upheaval — and explaining that you need to build or retrofit a facility for robots. Half the room doesn’t want another industrial building near their property. The other half is worried the robots inside it are coming for their jobs. NIMBYism on one side, automation anxiety on the other — and the company in the middle just needs a building with the right specs. That’s a real estate problem, and it’s one that doesn’t get nearly enough attention.
The Numbers That Explain Why This Matters
Let me give you the 30,000-foot view.
According to the International Federation of Robotics, 542,000 industrial robots were installed worldwide in 2024 — more than double the number from a decade ago. The global operational stock now exceeds 4.6 million units. And the IFR projects installations will surpass 700,000 units annually by 2028.
Those robots have to be manufactured somewhere. Tested somewhere. Calibrated, stored, and shipped from somewhere. And the “somewhere” isn’t a standard industrial box with dock doors and fluorescent lighting.
The facilities these companies need are complex, expensive, and extraordinarily specific. And here’s where it gets wild: the companies building robots are increasingly using robots — trained by artificial intelligence — to build more robots. Figure AI, the humanoid robotics company, recently unveiled a manufacturing facility designed to produce up to 12,000 humanoid robots per year, using their own AI-powered robots on the assembly line.
That’s not science fiction. That’s a lease negotiation.
When the Robot Changes, the Building Changes
Want proof at scale? Look at Amazon.
Amazon has deployed more than 1 million robots across its operations network since it acquired robotics company Kiva Systems in 2012. The early fulfillment centers were warehouses with mobile robots scooting shelves to human pickers. Then Amazon introduced robotic arms — systems named Robin, Cardinal, and Sparrow — that sort, lift, and route packages using AI and computer vision. They deployed Proteus, their first fully autonomous mobile robot, which navigates alongside human workers using “safety bubbles.” These technologies had existed in manufacturing and automotive plants for over a decade — companies like Intel and GM have long operated factories with comparable robotic integration. But Amazon is deploying them at a scale and pace that’s rewriting the playbook for industrial buildings across the country.
The result? The newest facility in Shreveport, Louisiana is a 3-million-square-foot, five-story building with eight different robotic systems working together. It uses ten times more robots than previous designs. Amazon is building a similar next-generation facility down Highway 74 in Wilmington — over 3 million square feet, opening in 2026.
The generation of the robot dictated the generation of the building. And according to Amazon’s own research division, they’ve built AI models that allow one generation of robots to train the next. The robots generate data. The data trains smarter AI. The smarter AI trains better robots. And those better robots need a different building.
Why Charlotte Is in the Conversation
One of the reasons this assignment landed in our market — and one of the reasons I picked up the phone and called Dr. Srinivas Akella at UNC Charlotte — is the talent pipeline.
Dr. Akella is a Professor of Computer Science in UNC Charlotte’s College of Computing and Informatics. His Ph.D. is from Carnegie Mellon’s Robotics Institute — one of the top programs in the world — and his research spans motion planning, multi-robot coordination, and manufacturing automation. So I asked him a simple question: does the building actually matter?
His answer reframed the whole conversation. He told me that layout planning is part of the robotic system itself — the building is the outer skin. The process inside dictates what the building needs to be. Do you need wide-open space, or distinct zones? Is the facility all-robot, or are humans working alongside the machines? How those questions get answered changes the floor plan, the ceiling height, the utility infrastructure, the safety design — everything. A 2024 academic review of 80 published studies confirmed what Dr. Akella described: buildings have traditionally been designed around human needs, with no consideration for robotic integration. That’s the gap. And it’s closing fast.
“I hadn’t really thought about it until you asked, but the building is like the robot’s skin.
The inputs that went into the robot layout will dictate the building specifications.”
~ Dr. Srinivas Akella, Professor of Computer Science, UNC Charlotte.
Ph.D., Robotics, Carnegie Mellon University.
The market data backs up the Charlotte story. Lucid Bots, a Charlotte-based robotics company, makes autonomous cleaning drones that have washed the exterior of Bank of America Stadium. They’re Y Combinator–backed, raised $9.1 million in Series A funding, and were named one of the top three fastest-growing companies in the city by the Charlotte Business Journal. FANUC, one of the world’s largest industrial robot manufacturers, operates its Southeast U.S. regional office in Huntersville. DBR77, a European robotics and AI platform company, recently planted its U.S. headquarters here. Site Selection Group ranked North Carolina #1 in the nation for manufacturing in 2024.
“But Won’t They Just Build Their Own Facility?”
I get this question a lot. And the answer surprised me.
Most robotics companies are engineering companies, not real estate companies. They can design a machine that navigates a warehouse autonomously, but they can’t tell you whether a build-to-suit lease makes more sense than buying land and building from scratch. They don’t know how to structure a state incentive package. They don’t know that the power utility timeline to upgrade service to the level they need can run 9 to 14 months — and that timeline has to sync with their construction schedule or the whole project slips a quarter.
This is the kind of work we do — the deals that don’t fit neatly into a category, where the real estate itself is part of the problem to solve. It requires an advisor who understands both the real estate and the operations inside it. And frankly, those deals are the ones we enjoy the most.
What I Want You to Take Away
Whether you own industrial real estate, advise clients who do, or are a corporate leader planning your next facility move — the playbook is changing.
The next generation of industrial tenants won’t just occupy your building. They’ll interrogate it. They’ll ask about the power capacity before they ask about the rent. They’ll want to know the floor flatness before they look at the lobby. And if your building was designed for yesterday’s tenant, you’re going to lose them to the developer who built for tomorrow.
The buildings that win the next decade of industrial demand will be the ones designed from the inside out — starting with what the machines require, not what the market has always supplied.
If you’re evaluating an advanced manufacturing facility — or advising a client who is — give us a call or shoot us an email. We’ll walk you through our Deal Canvas™ framework and help think through the site selection, build-vs.-lease analysis, and incentive strategy before anyone signs anything.
