⚛️量子位•Jun 2, 2026Stalecollected in 88m

LiOS: Bridging Cloud Models to Embodied Robotics

Post LinkedIn

⚛️Read original on 量子位

#embodied-ai #robotics #sim-to-reallioslios 招商局獅子山人工智能實驗室

💡Learn how LiOS bridges the gap between cloud-based AI models and real-world robotic manipulation.

⚡ 30-Second TL;DR

What Changed

LiOS enables seamless integration between cloud-based AI models and physical robotic hardware.

Why It Matters

This framework accelerates the deployment of general-purpose AI models into physical robotics, potentially lowering the barrier for household automation.

What To Do Next

Investigate the LiOS architecture to see if it can streamline your current sim-to-real pipeline for robotic manipulation tasks.

Who should care:Researchers & Academics

Key Points

•LiOS enables seamless integration between cloud-based AI models and physical robotic hardware.
•Supports complex flexible object manipulation in household environments.
•Provides a unified pipeline from simulation-based training to real-world deployment.
•Focuses on solving the 'sim-to-real' gap for embodied intelligence.

🧠 Deep Insight

Web-grounded analysis with 15 cited sources.

🔑 Enhanced Key Takeaways

•The Lion Mountain AI Lab (also known as Lion Rock AI Lab), responsible for LiOS, is a strategic initiative of the China Merchants Group (CMG), which has significantly increased its R&D investment in emerging technologies, including embodied intelligence, as part of China's 14th Five-Year Plan.
•The lab's establishment in Hong Kong Science Park underscores a regional effort to position Hong Kong as an international innovation and technology hub, with LiOS contributing to this goal by promoting industrialization and talent cultivation in AI.
•LiOS directly tackles the "sim-to-real" gap, a critical challenge in embodied AI where models trained in virtual environments often fail in reality due to discrepancies in physics, sensor noise, and environmental complexities, which physics engines struggle to perfectly replicate.
•The system's success in enabling full-chain embodied intelligence from simulation to multi-form real machine operation positions it within a global trend of advancing embodied AI from theoretical concepts to practical, real-world applications, particularly in China's rapidly growing robotics market.

🔮 Future ImplicationsAI analysis grounded in cited sources

LiOS and similar systems will accelerate the commercialization of embodied AI in diverse sectors, including household and logistics.

By effectively bridging the sim-to-real gap and enabling complex manipulation, LiOS facilitates the deployment of AI models from controlled simulations to real-world physical robots, which is crucial for practical applications in various industries.

The success of LiOS will strengthen Hong Kong's position as a leading hub for AI and robotics innovation.

The Lion Rock AI Lab's establishment in Hong Kong Science Park and its focus on industrialization and talent cultivation directly contribute to the city's strategic goal of becoming an international innovation and technology center.

LiOS will drive further integration of cloud computing resources with edge robotics for enhanced autonomy and scalability.

The system's core function of bridging cloud models to embodied robotics inherently promotes a hybrid architecture where powerful cloud AI can be leveraged by physical robots, enabling more sophisticated capabilities and fleet management.

⏳ Timeline

2021-2025

China Merchants Group (CMG) established the Lion Rock AI Lab and intensified investment in embodied intelligence during the 14th Five-Year Plan period.

2025-09

The Lion Rock AI Lab of China Merchants Group was officially inaugurated in Hong Kong Science Park.

2026-01

The Lion Rock AI Lab made breakthroughs in fields including quadruped robots and multi-modal algorithms.

2026-06-02

LiOS system launched by China Merchants Group's Lion Mountain AI Lab, enabling full-chain embodied intelligence from simulation to multi-form real machine operation.