

CyberRobo
4K posts

@CyberRobooo
One morning in 2050, humanoid robots and humans stand at a crossroads.About humanoid robots: products,use cases, business ,HRI 📮[email protected]




An engineer from 1X explains why NEO's tendon system is more natural and safer: When a human arm makes contact with a robotic hand equipped with geared joints, the hand continues to operate,potentially causing injury. NEO, however, contains no gears; it is driven by tendons. Upon contact with a human, it yields and adapts to the person's movements,responding just as naturally as a human would. (Friendly HRI)



"This is the secret room that nobody’s allowed to come in." Figure CEO Brett Adcock walks through every generation, from Figure 01 → hints of Figure 04 We started in 2022, & the goal was, how do we get humanoid robots to our AI & software team as fast as possible?" Evolution highlights: → Cost ↓ ~90% (Figure 2 → Figure 3) → Built for speed, not cost (early generations) → Compute ↑, battery ↑, sensors ↑ each iteration → “Ozempic” effect — slimmer, lighter, same power → From fragile prototypes → scalable manufacturing “Figure 4 will be the biggest step up we’ve ever made by far.” @adcock_brett @Figure_robot

Collaborating like an ant colony Blue Worm Robotics from Xi'an has released a wheeled humanoid robot solution for warehouse automation. The wheeled humanoid robots move freely and quickly among warehouse shelves and workstations, using their lifting bodies to move, transfer, and sort goods. The robots collaborate with each other, and multiple robots can even work together to assemble another robot. As the collaborative tasks become more complex, swarm intelligence will emerge.

A unit of 15 smart traffic robots hit the streets of Hangzhou on day one of the May Day holiday, quickly becoming one of the city's hottest sights. Deployed across key areas, from scenic spots to busy commercial zones and major roads, they team up with officers to direct traffic and help people find their way, marking a new stage in the city's smart traffic management.

Hangzhou has deployed humanoid robot traffic police at West Lake to ease congestion caused by heavy tourist traffic. These robots, part of a recent batch of 15, can work continuously for 8–9 hours a day on repetitive tasks like directing traffic, shouting instructions, and guiding crowds. This frees up human officers to focus on accidents, serious violations, and other complex incidents.

GR00T-VisualSim2Real is now open source! VIRAL and DoorMan are now available with training code, simulation assets, and the full recipe for bringing visual sim-to-real loco-manipulation skills to your own humanoids. Repo: github.com/NVlabs/GR00T-V…






X-Humanoid released Embodied Tien Kung 3.0: a More Open and Practical Humanoid Robotics Platform, and felt like a preview of large-scale deployment. For years, humanoid robots have proven their mobility. Now, they're beginning to prove their maneuverability. -» Jump off a 1-meter high stage(Imagine a construction site) -» Stable recovery on cluttered terrain. (Outdoor exploration, special operations) -» Millisecond-level ball catching. (Precise perception-motion loop) -» Precise operation of electrical cabinet buttons and block arrangement. … Even the control of a ping-pong paddle allows for continuous fine-tuning under disturbance (with immense potential for fine-tuning applications in warehouses/workshops and service scenarios). Thanks to hardware evolution, it boasts a 1.69m humanoid size, 43DOF, 62kg, 400 Nm peak torque, and integrates highly dynamic whole body control and tactile sensing. At its core lies autonomy. Wise Kaiwu constructed a perception-decision-execution loop (VLM + world model + VLA + autonomous navigation), achieving adaptive action, multi-robot collaboration, and mobile control. It incorporates early brain-computer interface integration, pre-programmed for intention-level human-robot collaborative interaction. The highly open technology stack, including open-source models, control frameworks, datasets, and the ROS native SDK, is rapidly narrowing the gap between the capabilities and deployment of individual developers and startups. It can be said that humanoid robots may transition from the demonstration phase to the infrastructure phase faster than expected. With the infrastructure in place, large-scale deployment enters another level.