GR00T N1: An Open Foundation Model for Generalist Humanoid Robots

2025-03-18

AI Summary

Open foundational models for general-purpose humanoid robots must address multi-task generalization in real-world scenarios, real-time motion generation, and cross-modal instruction understanding. This paper introduces the first dual-system Vision-Language-Action (VLA) architecture: System 2 (vision-language understanding) and System 1 (diffusion Transformer-based action generation) are jointly trained end-to-end to unify modeling of real robot trajectories, human demonstration videos, and synthetic heterogeneous data. Departing from conventional staged pipelines, our approach enables language-conditioned, high-fidelity, low-latency bimanual manipulation. It significantly outperforms existing imitation learning methods across multiple simulation benchmarks and is successfully deployed on the Fourier GR-1 humanoid robot. With only a few demonstrations, it accomplishes complex tasks—demonstrating strong generalization, robustness, and data efficiency.

Technology Category

• Intelligent Robots: Manipulation
• Computer Vision: Language and Vision
• Humans and AI: Human-Aware Planning and Behavior Prediction

Application Category

• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Search and Retrieval-Augmented AI: Multilingual and cross-lingual Web search

Abstract

General-purpose robots need a versatile body and an intelligent mind. Recent advancements in humanoid robots have shown great promise as a hardware platform for building generalist autonomy in the human world. A robot foundation model, trained on massive and diverse data sources, is essential for enabling the robots to reason about novel situations, robustly handle real-world variability, and rapidly learn new tasks. To this end, we introduce GR00T N1, an open foundation model for humanoid robots. GR00T N1 is a Vision-Language-Action (VLA) model with a dual-system architecture. The vision-language module (System 2) interprets the environment through vision and language instructions. The subsequent diffusion transformer module (System 1) generates fluid motor actions in real time. Both modules are tightly coupled and jointly trained end-to-end. We train GR00T N1 with a heterogeneous mixture of real-robot trajectories, human videos, and synthetically generated datasets. We show that our generalist robot model GR00T N1 outperforms the state-of-the-art imitation learning baselines on standard simulation benchmarks across multiple robot embodiments. Furthermore, we deploy our model on the Fourier GR-1 humanoid robot for language-conditioned bimanual manipulation tasks, achieving strong performance with high data efficiency.

Problem

• Develops a versatile foundation model for humanoid robots
• Enables robots to handle real-world variability and learn tasks
• Improves performance in language-conditioned manipulation tasks

Innovation

• Vision-Language-Action model for humanoid robots
• Dual-system architecture with real-time motor control
• Trained on diverse datasets including real-robot trajectories