Summary This repository provides a comprehensive, structured learning guide for Embodied AI, targeting newcomers and practitioners. It curates a path from foundational concepts and overviews to advanced research, practical simulation tools, and real-world robotics frameworks, aiming to accelerate project implementation and open-source contribution.

How It Works The guide breaks down Embodied AI into detailed learning roadmaps covering Reinforcement Learning, Vision-Language-Action (VLA), Simulation-to-Reality (Sim2Real), and robotics control. It integrates theoretical foundations, classic algorithms, state-of-the-art research, and hands-on experience with popular simulation platforms (Isaac Lab, MuJoCo, PyBullet) and real-world frameworks (LeRobot), emphasizing clear paths and engineering implementation.

Quick Start & Requirements Installation uses standard Python package management (pip, conda) and potentially Docker. Prerequisites include foundational knowledge in Python, linear algebra, probability, control theory, and Linux/Git. Specific tools like Isaac Sim require NVIDIA GPUs with substantial VRAM. The guide links extensively to external resources, papers, code repositories, and documentation for detailed setup.

Highlighted Details

• Holistic Curriculum: Covers Embodied AI from overview to SOTA research and industry landscape.
• Actionable Roadmaps: Detailed plans for sub-fields like RL, VLA, Sim2Real, and Control.
• Tooling Integration: Practical guidance on simulation platforms (Isaac Lab, MuJoCo, PyBullet) and robotics frameworks (LeRobot).
• SOTA & Classics: Curated content on cutting-edge research and foundational algorithms.
• Community & Industry Insights: Lists key researchers and companies in the field.

Maintenance & Community Maintained by Xbotics community volunteers, the project uses GitHub Issues for Q&A and PRs for contributions. A contribution guide and versioning strategy are in place, fostering community engagement.

Licensing & Compatibility Code is MIT licensed, documentation CC BY 4.0. External assets retain their original licenses. These permissive licenses generally allow broad compatibility, including commercial use.

Limitations & Caveats This resource serves as a "roadmap and resource navigation," not an exhaustive tutorial. Users require foundational technical knowledge. Some simulation tools have specific hardware requirements (e.g., NVIDIA GPUs). Coverage depth varies, directing users to external resources for deeper dives.