Roboselect360

@HardwareSpeed You're right. Capturing the "how" but does not directly capture the planning. But the quantity and quality of data will speed up this transition

@roboselect360 15,000 training points per day is serious scale. But every trajectory assumes someone already decided what task to run. Who planned the sequence? Teleop captures HOW to move. It doesn't capture WHAT to do, in what order, and why. That layer is still a human with a spreadsheet.

Manual jobs will never be replaced ? Watch this. China just opened the world’s largest humanoid robot data factory in Zigong, Sichuan. Dozens of operators in VR headsets tele-operate UBTECH Walker S2 robots in real time. Every movement is captured and sent to the cloud to train the AI. 6,000 m² 15,000 training data points/day 3 million high-quality trajectories/year A new job is born: AI Robot Trainer. While the West debates, China is scaling data + production (10k units/year). The robot revolution is learning by imitation and it’s already here. Must-see video

@JulianSaks China has a sizeable lead….

Chinese companies in 2025 outshipped the US in humanoids by 28x 🇨🇳🇺🇸 humanoids.fyi/industry/geopo…

I’ve put together a more detailed breakdown here: roboselect360.com/en/robots/agib… @AGIBOTofficial

AGIBOT A2 highlights something important: humanoid robots are no longer just prototypes. Some companies are already scaling production. The next question isn’t capability, it’s real-world deployment at scale.

@AGIBOTofficial Impressive scale. I recently added AGIBOT to my tracking, curious to see how this translates into real deployments.

Today, the 10,000th AGIBOT humanoid robot officially rolled off the line. This isn't just a production milestone; it’s the moment humanoid robots transition from technical validation to global, large-scale deployment. The most exciting part? Our velocity. 0 to 1,000: 2 years 1,000 to 5,000: 1 year 5,000 to 10,000: Just 3 months We’ve accelerated production efficiency by 4x, proving that humanoids aren't just a future concept—they are ready for the workforce today. From logistics to industrial manufacturing, AGIBOT is now scaling at the speed of global demand. #AGIBOT #HumanoidRobots #EmbodiedAI #MassProduction #HumanoidNo1 #HumanoidShipmentsNo1 #SmartManufacturing

@xmaquina @physical_int Big bet on general-purpose robotics. The real question is whether intelligence alone will be enough, or if execution and deployment will decide the winners.

Breaking: @physical_int is in talks to raise $1B at a valuation above $11B, doubling in the past four months. Investors continue to back the idea that general-purpose robot intelligence will be a winner-takes-most market. More on the robot brain layer ↓ xmaquina.io/blog/inside-th…

The humanoid robot market is becoming harder to read. More players, more announcements, but very little standardized data. Comparing robots isn’t just about specs. It’s about understanding: deployment, economics, and scalability. That’s the gap I’m trying to address. Open to feedback.

@JulianSaks Step by step

coming 🔜 humanoids.fyi

Still an open question. Likely early use cases: manufacturing and logistics, where tasks are repetitive and environments more controlled. The real challenge is finding where it creates enough value to scale.

Many humanoid robots look impressive. The real question: where do they actually create value? Manufacturing? Logistics? Services? Innov8 Haguenau - France - 27/03/2026

If you’re curious, I wrote more about this here: roboselect360.com/en/robots/figu…

Figure 03 is not just a better robot. It’s designed for manufacturing from day one. Fewer parts, new supply chain, production-ready processes. That’s a major shift: from prototype → to industrial product. Most humanoids aren’t there yet.

We launched RoboSelect360. A platform focused on robotics, humanoid robots and physical AI. Breaking down what actually matters: scaling, manufacturing, unit economics and real-world deployment. Not hype — execution. #robotics #AI #humanoids

La robotique entre dans une nouvelle phase : des démos au déploiement. Échelle. Économie. Impact réel. C’est pourquoi nous avons lancé RoboSelect360. Une plateforme centrée sur la transition du proof-of-concept au déploiement réel avec un regard affûté sur l’économie unitaire, la scalabilité et l’impact mesurable. Disponible en anglais et en français.

The robotics industry is crossing a threshold, from impressive demos to actual deployment. RoboSelect360 was born from that transition, with a focus on what truly matters: scale, unit economics, and tangible impact. Open to connecting with anyone thinking seriously about where this goes next. RoboSelect360 is now available in English and French.

@JulianSaks You're doing a tremendous amount of work, it's very impressive. I could show you what I've accomplished since January 2026. I've started building a database of several thousand rows.

The future of Atlas could quickly turn into: > access data feeds > one shot humanoid designs > have agents procure all components Humanoid designs will become so accessible over the next year got to love it

@BostonDynamics Well said. The journey from research to real-world deployment is where most of the value is created.

You don't get from the research lab to customer sites overnight. As our first Atlas customers put their robots to work this year, the evolution of Atlas tells a clear story, mirroring our journey with Spot and Stretch. 🤖 Real-world impact is driven by solving problems that add value, not flash 🤖 Development take patience, persistence, and collaboration 🤖 Software and hardware go hand in hand - a useful humanoid is smart and autonomous, but also robust, service, and scalable 🤖 The humanoid form factor is a jumping off point, not a limitation bosdyn.co/47AEtGK

Did you put a viability score for each robot? I know that few industrial deployments have taken place so far, but setting up a type of scoring for each robot would make sense, and to be developed as the data comes to us. The industry is only in its infancy, BUT starting to build a viability model would allow us to have constant monitoring. In any case, that’s what I did on my side.

Understand key supply chains across all humanoid components & suppliers

Physical AI is clearly accelerating. The real test now is moving from capability to large-scale deployment.