Galbot’s Humanoid Robot Integrates Nvidia Jetson Thor with Potential for Manufacturing Use
AI and robotics company Galbot is integrating Nvidia Jetson AGX Thor into its G1 Premium humanoid robot. With Jetson Thor, the Galbot G1 Premium has 7.5x the AI compute of the previous generation Nvidia Jetson Orin and 3.5x greater energy efficiency.
Unveiled at the World Robotics Conference, this partnership positions Galbot among the first companies to use Thor’s computing capabilities in commercial robotics applications. These advancements were demonstrated at the World Humanoid Robot Games, where the G1 Premium operated autonomously and completed tasks such as retrieving medications from multi-tiered shelves.
Galbot said its success with humanoid development stems from its Sim2Real methodology, which uses Nvidia Isaac Sim and Isaac Lab built on the Omniverse 3D graphics platform. This approach uses pre-training on synthetic datasets with minimal real-world data refinement, reducing costs while improving generalizability. Galbot has developed proprietary VLA (vision-language-action) foundation models — GraspVLA, TrackVLA and GroceryVLA — creating what it says is “the world's first end-to-end embodied AI model family.” These models enable zero-shot grasping, autonomous navigation and operation in high-density environments.
Real-world deployment of Galbot’s G1 Premium is already underway, with more than 10 smart pharmacies in Beijing currently using the robots, with expansion to more than 100 locations planned by the end of the year. The company has also partnered with Bosch Group to deploy its AI-powered platform in smart manufacturing to replace traditional pre-programmed robotic arms.
More robotics coverage from Automation World:
