Ready For Tomorrow #92

In partnership with

Hi, welcome back to Ready for Tomorrow.

Grab a coffee. Or pretend you grabbed one. This episode is short, light, and meant to wake up your brain without hurting it.

Let’s go!

TL;DR

• AGIBOT went from zero to No.1 in two years by shipping over 5,000 humanoid robots in a single year.

• Skild AI raised $1.4B because investors believe robot intelligence will matter more than robot hardware.

• Schaeffler partnered with Humanoid to deploy hundreds of humanoid robots directly into real factories.

• Siemens proved that humanoid robots can already handle basic logistics tasks in an active production environment.

• 1X is teaching robots to learn from their own experience, reducing the need for humans to train every move.

Two years old. Five thousand robots. No chill.

Chinese humanoid company AGIBOT just did something that makes the whole robotics world stop scrolling.

More than 5,100 humanoid robots shipped in 2025.

Founded in 2023.

Ranked No.1 globally in shipment volume and market share in 2025, according to Omdia.

That speed is wild.

AGIBOT showed up at CES 2026 with confidence. They also presented the A2 humanoid, plus Genie Sim 3.0, built on NVIDIA Isaac Sim. Simulation, data, evaluation, physics. All in one place.

They also talked about fleet updates. Vision, language, action models updated online across robots. That sounds boring until you realize it means learning at scale.

Here is the real question.

Can quality and reliability grow as fast as shipment numbers?

Shipping robots is one thing. Keeping them working for years is another story.

Humanoids are no longer “future tech”.

They are production numbers now.

The CRM that saves teams hours every week

It's not about working harder — it's about having a CRM that actually thinks ahead. HubSpot Smart CRM learns how your team operates and adapts to make everyone more effective. Streamline day to day tasks and track the activity that actually matters to your business. The result? Your team gets back hours every week to spend on growth instead of admin work. Start free. See the difference.

Get Started Free

Robots watching humans. Investors watching robots.

Skild AI just raised $1.4 billion in a Series C led by SoftBank.

Their valuation jumped to over $14 billion.

That happened in seven months.

Skild AI was founded in 2023. The idea is simple to explain and hard to build. Robots learn tasks by watching humans, not by endless reprogramming. One model. Many robots. Many jobs.

No weeks of training for every new task. No starting from zero each time.

So far, the company has raised more than $2 billion, with backing from NVIDIA, Macquarie Group, and 1789 Capital.

This timing makes sense.

Training robots is still one of the biggest blockers in robotics. Factory robots. Service robots. Even home robots. If robots can learn as they go, adoption suddenly looks much easier.

The interesting part is not just the money.

A $14B valuation sends a clear message. The real value is shifting away from metal, motors, and joints. It is moving into the intelligence layer that can sit on top of any robot.

SoftBank seems to agree. After years of mixed robotics bets, they are back. This time, they are betting on the brain, not the body.

Robotics is quietly changing shape.

And the market is already pricing that in.

When robot helpers leave the lab and head to the factory floor.

Schaeffler AG and UK-based robotics firm Humanoid just announced a multi-year partnership to bring humanoid robots into real factory work.

Here’s what’s happening:

• Schaeffler will integrate hundreds of humanoid robots into its global production facilities over the next five years.

• The deal covers actuator supply and joint development of key robot components.

• Initial beta deployments are expected in 2026–2027 to test real integration, safety, performance, and systems fit.

• After validation, robots may be offered under Robot-as-a-Service or direct purchase models for broader use.

This isn’t just another pilot. The goal is large-scale deployment in industrial environments, not demos locked behind ropes.

Factories may soon host robots that walk, move parts, and help with hands-on tasks as part of everyday production.

Humanoid robotics is stepping out of the lab and into real work.

Humanoids step out of the lab and into real logistics.

UK robotics company Humanoid and German giant Siemens just finished a proof of concept showing humanoid robots can handle real industrial logistics tasks.

Their robot, the HMND 01 wheeled Alpha, was tested inside a Siemens Electronics Factory in Erlangen, Germany. It wasn’t a demo on a stage. It worked inside active logistics, picking totes from a stack, moving them, and placing them on a conveyor for human workers.

The trial had two parts. First, teams built a physical twin to test and iterate quickly. Then the robot spent two weeks on site, performing real tasks with minimal supervision.

The results were solid:

• Autonomous tote movement from storage to conveyor.

• Continuous operation for more than 30 minutes at a time.

• Success rates above 90 % for picking and placing.

This is one of the clearest signs yet that humanoid robots can start joining logistics operations, not just research labs.

Humanoid and Siemens see the POC as a first step toward broader use in real industrial environments.

Robots are learning by watching, not being told.

1X is changing how humanoid robots get smart. Their robot, Neo, used to rely on humans in motion-capture suits or VR headsets to teach it physical tasks. Now the company says it’s moving away from that model and shifting to a new AI “world model” that lets Neo learn from its own video footage instead of humans actively training it.

This is a big deal because conventional humanoid training depends on people repeating tasks for hours to collect data, which is slow, expensive, and even physically tiring. With the world model, the robot uses video it captures itself to learn how to act in the real world.

Neo is still likely to be teleoperated early on, and it might not handle every task perfectly at first. But the plan is to reduce human involvement over time, letting the robot improve from its own experiences. It’s expected to ship this year and could generalize to new tasks faster than older approaches.

This shift echoes broader trends in the industry, where reliance on human-generated training data is slowly being replaced by systems that can learn from video and real interactions themselves.

Robots are starting to watch and learn the way people do… just without asking for snacks.

That’s it for this episode of Ready for Tomorrow.

And one question stays in my head.

Is this humanoid hype real. Or is it just a mirage that looks great from far away and disappears once you get closer.

Big numbers. Big money. Big promises.

But also big expectations.

We’ll see sooner than we think.