Key Highlights
- NVIDIA boosts humanoid robots with advanced AI computing, enhancing motion, vision, and decision-making.
- Next-gen AI-powered humanoid robots, such as the Galbot G1 Premium robot, are set to transform healthcare, logistics, and industrial applications.
- Advances in Sim2Real robotics mark a major leap in bridging robotics with human-like adaptability and intelligence.
Humanoid robots are taking a major step forward, thanks to NVIDIA robotics technology designed to enhance their intelligence and real-world performance. This development will accelerate how robots learn, adapt, and interact with real-world environments.
Robotics company Galbot is set to integrate Nvidia’s latest Jetson Thor physical AI platform into its G1 Premium humanoid robot. The aim is to have robotics performance boost and create higher levels of autonomy.
This development creates opportunities across industries such as artificial intelligence, research, manufacturing, healthcare, logistics, and customer service.
Smarter, Faster, and More Capable
The G1 robot, designed for use in retail, healthcare, and logistics sectors, was recently showcased at the World Robotics Conference in Beijing. With Jetson Thor onboard, the humanoid robot is set to deliver faster, smarter, and more efficient operations.
By integrating NVIDIA’s powerful computing platforms, AI-powered humanoid robots can process data at incredible speeds.
A Step Toward Human-Like Intelligence
NVIDIA says Jetson Thor offers over seven times the AI computing power of its predecessor, Jetson Orion, and three times greater energy efficiency. This jump in capability allows Galbot’s robots to handle complex planning and motion tasks with far greater accuracy and reliability.
A key factor behind this leap is Galbot’s Sim2Real approach, which uses NVIDIA’s Isaac Sim and Isaac Lab frameworks combined with open-source robotics simulation.
Based on this base, Galbot has developed its vision-language-action (VLA) models, including GraspVLA, TrackVLA, and GroceryVLA. These enable robots to perform advanced tasks such as zero-shot grasping, autonomous navigation, and dense retail operations with near-human-like performance.
Professor Wang He, Galbot’s founder and CTO said, “Our G1 Premium, now running on Nvidia Jetson Thor, has already shown remarkable improvements in speed and real-time reasoning. This early adoption empowers us to take our VLA models to an entirely new level of real-world capability.”
The G1 Premium is already proving its value, with 10 smart pharmacies in Beijing currently powered by the robot. Galbot is planning to expand to 100 locations nationwide by the end of the year.
NVIDIA’s Sim2Real robotics frameworks reduce training times and improve scalability for developers.
Looking Ahead
With the rapid evolution of AI-driven robotics, the combination of NVIDIA’s advanced chips and software with innovative humanoid robot designs will set the stage for a new era of automation.
The company is collaborating with Bosch and other manufacturers to deploy embodied AI-powered robots on production lines. The goal is to replace traditional, rigid robotic arms with more adaptable, human-like machines, making factories more efficient and flexible.