The quiet revolution of warehouse robots, 2025

The humanoid robots get all the attention. The flashy videos. The funding rounds. The headlines.

Meanwhile, the most successful robot deployment in human history is happening in warehouses and nobody cares.

Amazon has over 1 million robots in its warehouses. One million. Little orange platforms that carry shelves. Arms that sort packages. Vision systems that identify items. They process a significant percentage of the packages that arrive at your door.

Ocado’s automated grocery fulfillment system is in 8 countries. Symbotic is deploying in Walmart. Locus Robotics just passed 2 billion units picked.

These robots don’t walk. They don’t do backflips. They’re not on magazine covers. They’re on warehouse floors. Working. Right now. While you read this.

The lesson

The most successful robots are the least humanoid. They’re purpose-built for specific tasks. They don’t try to do everything. They do one thing, relentlessly, at scale.

There’s a tension in the robotics industry between the generalist dream (a humanoid that can do anything) and the specialist reality (a purpose-built machine that does one thing very well). Right now, the specialists are winning. By a landslide.

A million warehouse robots versus maybe a few hundred humanoid robots in commercial deployment. The scoreboard isn’t close.

Why humanoid robots still matter

The specialist robots work because warehouses are designed for them. Rails, shelves, standardized bins. The environment is structured.

Homes aren’t structured. Construction sites aren’t structured. Disaster zones aren’t structured. For unstructured environments, a generalist (humanoid) form factor makes sense because the world was built for humans.

Both approaches will coexist. Specialists in structured environments. Generalists everywhere else. The warehouse robot and the humanoid robot aren’t competing. They’re solving different problems.

But the warehouse robots are the ones generating revenue today. And revenue is what turns a technology from a research project into an industry.

I think the humanoid robot’s moment is coming. But while we wait for it, the unglamorous robots are quietly building the world’s first large-scale robotic workforce.

A million robots. Working. Right now.

That’s not the future. That’s the present. And almost nobody noticed.

Related thinking: