Are you interested in a custom automation solution? Let's talk! Contact Us

Industry Insights

Are We Ready to Call Our Robots Rosie?

October 22, 2026
Are-We-Ready-to-Call-Our-Robots-Rosie-

Meet Rosie

When The Jetsons first aired in 1962, Rosie the robot maid was pure fantasy. She rolled through the house cooking, cleaning and handling the jobs nobody else wanted, with a little attitude on the side.

Sixty-plus years later, that idea doesn’t feel so far-fetched. Humanoid robots are leaving the lab and showing up on factory floors. And here’s the fun part: Rosie rolled on wheels. Many of the humanoids headed into factories today do too.

So, are we ready to call our robots Rosie? Not quite. But we’re closer than most people think, and the reason has less to do with arms and legs than with a shift in how robots think.

 

The Old Rule: Perfect Process First

For decades, automation came with one big requirement: the process had to be perfect before a robot could touch it.

Same part. Same position. Same orientation. Every time. Traditional robots are incredibly fast and precise, but they do exactly what they’re programmed to do. If a part shows up slightly crooked, or a box lands a few inches to the left, the cell stops.

That’s why so many good automation ideas never got off the whiteboard. The robot could do the task. The process just wasn’t consistent enough to let it. So manufacturers spent the money on fixtures, guarding, conveyors and floor changes to make the world fit the robot.

That model still works, and it’s still the right answer for plenty of high-volume jobs. But it left a lot of work on the table.

What Physical AI Changes

Physical AI flips the old rule. Instead of making the world fit the robot, the robot learns to fit the world. Physical AI is artificial intelligence that can see its surroundings, make decisions and act in the real world, in real time. Think of it as the jump from a robot that follows a script to a robot that can read the room.

NVIDIA CEO Jensen Huang has made physical AI the centerpiece of his company’s next chapter. At CES 2026 he also named the challenge plainly: “The physical world is diverse and unpredictable.” That unpredictability is exactly what physical AI is built to handle.

This is where the humanoid shape starts to make sense. Our plants, aisles, workbenches, carts and tools were all designed around the human body. Siemens has argued that humanoids can step into those existing “brownfield” factories with minimal disruption, instead of forcing a costly redesign. Boston Dynamics product lead Aya Durbin put it simply: with a humanoid, “you can drop the technology into the brownfield work cell as it exists.”

The robot adapts to the plant. Not the other way around.

The New Question: What’s Mundane Enough for a Humanoid?

The conversation is shifting from “Is my process perfect enough to automate?” to “Which jobs are repetitive enough that a humanoid could just do them?”

That’s a big change. Every plant has work that’s too varied for a traditional robot cell but too repetitive, heavy or tiring for people to enjoy:

  • Moving totes and bins between stations
  • Loading and unloading carts
  • Feeding parts into a machine, hundreds of times a shift
  • Kitting parts for the next operation
  • Walking material across a crowded floor where fixed conveyors don’t fit

None of these jobs are glamorous. That’s the point. They’re the jobs that cause fatigue, injuries and turnover, and they’re the ones manufacturers struggle to staff. A humanoid doesn’t need to be brilliant to help here. It needs to be steady, safe and flexible enough to handle a little variation.

The goal isn’t to replace the skilled people on the floor. It’s to free them up for the work that actually needs their judgment.

Reality Check: Where Humanoids Are Today

Humanoids are real, but they’re early. Most deployments today are pilots, not full production lines.

A 2026 review in Science Robotics named three gaps the industry still has to close:

  1. Reliability. A great demo is one thing. Running a full shift, every day, with vibration, debris and people walking by is another.
  2. Dexterity. Walking has come a long way. Fine work, like picking a small fastener or seating a connector, is still hard.
  3. Data. Robots learn from real-world experience, and there isn’t enough of it yet at factory scale.

The same review estimated the field is one to three hardware generations from solving reliability and dexterity at production scale. That’s not a reason to wait. It’s a reason to start learning now, so you’re ready when the hardware catches up.

So no, Rosie isn’t clocking in on second shift next week. But the gap is closing fast.

What This Means for Manufacturers

The manufacturers who start learning now will be the ones ready when humanoids are.

Here’s where we’d start:

  • Walk your floor with fresh eyes. List the repetitive jobs you can’t keep staffed. Those are your future humanoid candidates.
  • Stop waiting for the perfect process. Some of the jobs you ruled out for automation five years ago deserve a second look.
  • Get the basics in place. Safe shared workspaces, clean data and a team that’s comfortable working next to robots will matter no matter which platform wins.
  • Find a guide. Humanoids still need integration: connecting to your systems, designing the workflow and keeping people safe.

That last one is where we come in. At Akeratos, we’re working toward integrating a humanoid right here in Wichita. We want to know what these robots can really do on a real floor, so we can give our customers straight answers about when, and if, a humanoid makes sense for them.

So, are we ready to call our robots Rosie? Not yet. But we’re getting her a name tag ready.

– Andy, General Manager, Akeratos

Sources

Back to top