This week in robotics brought us five different stories.

A robotic hand that wants to be more human than most humans before coffee.
Agility changed its name, because apparently even robot companies need a rebrand moment.
A shelf-scanning robot got certified to safely live among shoppers.
CMU opened a huge robotics center, because the future apparently needs more square meters.
And Xiaomi is already testing humanoids in its EV factory, because of course they are.

Let’s go!

ChatGPT is a superpower if you know how to use it correctly.

Discover how HubSpot's guide to AI can elevate both your productivity and creativity to get more things done.

Learn to automate tasks, enhance decision-making, and foster innovation with the power of AI.

Download the free guide

The first item is not a breaking business headline. It is a research story, but a very interesting one. The video points to work presented at RoboSoft 2024 called “Replicating Human Anatomy with Vision Controlled Jetting. A Pneumatic Musculoskeletal Hand and Forearm.” The team behind it includes researchers from ETH Zurich, Inkbit, and MIT.

What makes this project special is not only that it looks human. It is that the researchers tried to copy the actual structure of the human hand and forearm much more closely than usual. The system includes a printed rigid skeleton, soft joints, tendons, touch sensors, and 22 independently controlled pneumatic artificial muscles. According to the paper, the hand showed human like finger motion and could grasp objects from a coin to a 272 gram can.

Why does this matter?

Because robotic hands are still one of the hardest problems in robotics. Moving a robot arm from point A to point B is one thing. Grabbing a random object gently, precisely, and reliably is another story. Human hands still win by a mile. This kind of research reminds us that the future of robotics is not only about bigger AI models or humanoid bodies. Sometimes the real battle is in the fingers.

My take is simple. We all get excited by full humanoids walking around factories. But if we do not solve dexterity, many of those robots will still be limited. A good hand is not a detail. It is the job.

Agility Robotics has rebranded. The company is now called simply Agility. On paper, this is just a name change. In practice, it is the kind of move companies make when they want to tell the market: we are not an experiment anymore. We are becoming a platform.

The company says the new name reflects its wish to grow beyond current deployments and expand into more industries and use cases. It also says its humanoids have already been deployed in real industrial settings for more than three years, and that it is aiming to deliver the first cooperatively safe humanoid in 2026.

This is where branding becomes strategy.

When a company drops part of its original name, it often means it wants more room. More products. More services. More markets. Less “we are a robotics startup,” more “we are building a category.”

And honestly, this makes sense. The humanoid race is getting crowded. If you want to stand out, it is no longer enough to say you have a humanoid robot. You need to show that you can deploy, support, scale, and survive the messy real world.

My take? This is not the biggest robotics story of the week, but it is a useful signal. Humanoid companies are slowly moving from technical identity to commercial identity. That sounds boring. It is not. It is what happens when an industry starts trying to grow up.

This may be the most underrated story on the list.

Simbe announced that its shelf scanning robot Tally achieved UL 3300 certification, and the company says it is the first retail robotics company to do so. The certification is for service and information robots working in public facing environments. In plain language, it is about proving that a robot can operate safely around normal people in normal places, not only in a controlled lab or behind a safety fence.

According to Simbe, Tally passed more than 40 safety tests. These covered things like navigation around people, control system safety, charging safety, stability, and electrical resilience. UL Solutions said this standard addresses the special safety needs of robots that work in public spaces with constant human interaction.

This matters a lot.

Why? Because the future of robotics is not only in factories. It is in stores, hospitals, airports, malls, and all the places where robots will move close to people who did not sign up for a robotics demo.

We often talk about AI, speed, autonomy, and labor savings. But trust may be even more important. A robot in a supermarket cannot feel like a risk. It has to feel normal.

That is why I think this story is bigger than it looks. Tally scanning shelves is not as flashy as a humanoid carrying boxes. But this is how robotics really enters daily life. Quietly. Through standards. Through certification. Through boring proof that the machine will not do something stupid near shoppers.

And yes, boring is good here.

Carnegie Mellon University officially opened its new Robotics Innovation Center on February 27. The center is 150,000 square feet, houses more than 50 labs and working groups, and is meant to support work in robotics, automation, and AI. FieldAI is already the first corporate tenant, and Pennsylvania Governor Josh Shapiro announced $1.5 million in state support for physical AI.

This is not just another university building.

This is infrastructure for the next chapter of robotics. CMU says the center will support research on robots for land, water, air, space, and even inside the human body. Other reporting on the center notes features like a 50,000 square foot indoor robot test area, a 6,000 square foot drone cage, and a 75,000 gallon water tank for testing submersible systems.

What I like about this story is that it reminds us where a lot of future robotics still begins. Not on LinkedIn. Not in product hype. In places where researchers, students, startups, and industry can work close together for years.

We like to imagine robotics progress as one big product reveal. Real progress is usually slower. It comes from test spaces, hard engineering, repeated failure, and people sharing tools and ideas under one roof.

My take is this. If humanoids are the loud part of robotics right now, places like CMU are the quiet engine behind what comes next. And those engines matter more than —-headlines.

And now to the story that probably got the most attention.

Xiaomi has begun trial operations of humanoid robots at its automobile factory. TechNode reported that CEO Lei Jun said the robots are already doing tasks such as loading self tapping nuts at assembly stations and transporting material boxes, and that Xiaomi wants large scale deployment within five years.

CNBC also reported that Xiaomi President Lu Weibing said two humanoid robots could complete about 90% of the work in three hours in a factory test, including installing nuts and moving materials. He added that the robots were not doing an “official job” yet, but were more like interns. CNBC’s reporting also said Xiaomi’s car factory pushes out a new vehicle every 76 seconds, and that keeping pace with the line is one of the biggest challenges.

That “intern” line is probably the most honest sentence in the whole humanoid market right now.

Because that is exactly where many of these robots are. Not useless. Not fake. But not ready to replace stable industrial systems at scale either. They are learning. They are being tested. They are entering the factory, but they still have to earn their place.

At the same time, this is serious. Xiaomi is not a tiny lab. It is a giant company with manufacturing ambition, hardware experience, and real pressure to improve production. When a company like that starts testing humanoids on actual tasks, it is worth watching.

My take? We should not overreact, but we also should not laugh it off. Humanoids in factories are no longer a pure concept story. They are now in the trial phase at major companies. That does not mean mass rollout is here. It means the door is open.

So yes, another week in robotics.

Some people are building better robot hands.
Some are changing logos and names.
Some are proving robots can survive public spaces without causing chaos.
Some are building giant research centers.
And some are already sending humanoids into factories like interns with no lunch break.

Business as usual. Just in 2026.

Cheers and have a nice week, Jacek !