4 min read

🤖 Nvidia Builds Humanoid

Plus: Robot Plays Tennis, Enter Chip Fabs, K1 Wins Soccer Cup

Good Morning, Roboticists!

Motion turned into data, coordination into inference, and labor into something that can be compiled.


Nvidia Builds Humanoid Robot Stack

TL;DR: Nvidia announced partnerships with European chipmakers Infineon, NXP, and STMicroelectronics to supply critical hardware for humanoid robots. Nvidia aims to provide the central computing “brain” through its Jetson Thor processors, while partners supply sensors, power management, motion control, and communications chips.


Humanoid Robot Plays Tennis

TL;DR: Chinese robotics company Galbot demonstrated software that enables a Unitree G1 humanoid robot to play tennis against a human engineer. The system uses imperfect human motion fragments to train the robot to track balls, react within milliseconds, and return shots with coordinated full-body movement. Researchers say the approach could extend to other complex physical tasks where high-quality human motion data is limited.


Humanoid Robots Enter Chip Fabs

TL;DR: STMicroelectronics plans to deploy more than 100 humanoid robots in its older semiconductor fabs across Europe to handle routine and physically demanding tasks such as wafer handling and machine interaction. The initiative aims to boost efficiency in aging factories that cannot easily be rebuilt or fully automated, helping the company remain competitive against newer highly automated fabs in China and elsewhere.


Booster K1 Robots Win Win Soccer Cup

TL;DR: B-Human, a 12-time RoboCup champion from DFKI and the University of Bremen, won Germany’s newly formed Humanoid Soccer League, defeating HTWK Robots 4–1 in the final. Using Booster K1 humanoid robots with 22 degrees of freedom, onboard AI, and 3D sensing, the team secured titles in both Middle and Large divisions, demonstrating coordinated play and autonomous decision-making.


Humanoids Learn to Navigate Safely

TL;DR: At Nvidia GTC, RealSense presented first-of-its-kind humanoid autonomous navigation system built with its depth cameras, Visual SLAM, and Nvidia’s cuVSLAM. In a demo with LimX Dynamics, a humanoid robot performed real-time localization, mapping, and navigation in 3D environments, including obstacle avoidance and terrain handling, showing how perception and vision software enable safe movement in human environments.


Robot Horse Carries 110 Pounds

TL;DR: DEEP Robotics unveiled a limited-edition robot horse based on its Lynx M20 platform, featuring bionic legs and hoof-like design. The quadruped weighs about 30 kg, can carry up to 110 lbs, and reaches speeds of 5 m/s. Equipped with LiDAR and SLAM, it supports autonomous navigation and operates in harsh environments like mud, rain, and dust.


China Trains Humanoids in Schools

TL;DR: China is establishing robot training centers across provinces to prepare humanoid robots for factory and logistics work. Dozens of robots are trained on tasks like carrying trays and folding clothes, generating millions of data points through repeated human-guided actions. Some centers simulate real environments such as assembly lines and homes, enabling robots to learn over 20 functions with reported success rates around 95%.


MLB Expands Robot Umpire Test

TL;DR: Major League Baseball is expanding its robot umpire challenge system to Triple-A, allowing players to appeal checked swing and ball-strike calls using automated tracking. Each team receives limited challenges per game, retaining them if successful. The move builds on earlier minor league trials as MLB continues testing rule changes alongside the system ahead of potential broader adoption.


READ MORE

Let the Future Come to Your Inbox

Stay ahead without drowning in information. We turn the most important signals across AI, tech, marketing, and future products into 5-minute reads you can actually finish.


TOGETHER WITH US

AI Secret Media Group is the world’s #1 AI & Tech Newsletter Group, reaching over 2 million leaders across the global innovation ecosystem, from OpenAI, Anthropic, Google, and Microsoft to top AI labs, VCs, and fast-growing startups.

We've helped promote over 500 Tech Brands. Will yours be the next?

Email our co-founder Mark directly at mark@aisecret.us if the button fails.