The Chinese University of Hong Kong (CUHK)’s InnoHK Hong Kong Centre for Logistics Robotics (HKCLR) announced the establishment of the Hong Kong Embodied AI Lab, the city’s first full-stack laboratory of its kind. The lab will focus on advancing the development of frontier technologies in embodied AI and robotics, and will collaborate with 24 industry partners, investors, and innovation and technology enterprises to accelerate the translation of research into industrial applications. This marks a significant step forward in strengthening Hong Kong’s AI industry chain and research commercialisation, contributing to Hong Kong’s development into an international innovation and technology hub.

The launch ceremony of the Hong Kong Embodied AI Lab was held on 18 May. Professor Irwin King Kuo-chin, CUHK’s Pro-Vice-Chancellor (Education), said that AI would become a key driving force behind the transformation of research and education. In recent years, CUHK has integrated AI into teaching and research, with a commitment to nurturing talent equipped with AI knowledge, innovative thinking and practical capabilities. He added that the University would also continue to strengthen interdisciplinary education to cultivate high-calibre talent for Hong Kong and the rest of the nation’s innovation and technological development.

Professor Liu Yunhui, Director of HKCLR, said that as a research platform dedicated to AI and robotics applications, HKCLR has possessed capabilities from fundamental research to application translation. The establishment of the Hong Kong Embodied AI Lab, he added, would aim to drive embodied AI technologies from the laboratory into real-world applications and accelerate the commercialisation process.

Breakthrough technologies on display – AI robotics enter daily life and industries

Professor Dou Qi, Associate Director of HKCLR, introduced a vision-language model integrated with spatial intelligence, jointly developed by HKCLR and a CUHK research team. The research findings were recently published in the internationally renowned journal Science Robotics, demonstrating HKCLR’s extensive research capabilities in frontier areas, including AI, robotic perception and embodied intelligence, while further advancing the deep integration of AI with the physical world.

At the ceremony, several pioneering embodied AI robotic innovations were showcased, demonstrating vast application potential in logistics, manufacturing, and services:

• Hong Kong’s first AI-powered dual-arm robotic platform: Equipped with multi-model sensors and a mobile manipulation system, capable of spatial perception and dual-arm collaboration to handle complex automated tasks in household and service environments.
• “Millimetre‑level” lightweight force‑controlled robotic arm: Weighing only 6 kg, yet achieving millimetre-level precision, offering high adaptability across diverse scenarios, adaptable to different manipulation platforms,
• All-terrain quadruped robot: Equipped with a differential drive system that enhances structural efficiency and payload capacity, combined with reinforcement learning-based motion control and visual navigation algorithms, it is capable of autonomously traversing rugged and challenging terrain.

Strengthening industry-academia-research-application framework

The Hong Kong Embodied AI Lab collaborates with 24 industry partners, investors, and innovation and technology enterprises to deepen cooperation across research collaboration, technology transfer, industrial applications and investment support. The 24 partners include (in alphabetical order): AGIBOT Innovation (Shanghai) Technology, BPS Global Group, China Resources Research Institute of Science and Technology, CUHK Innovation, Cyan Technologies, DeepCybo (Beijing) Technology, Deep Robotics, Guangzhou Huashang College, Haptron Scientific (Shenzhen), Hong Kong Innovation Angel Investment, Lenovo Capital and Incubator Group, Lumistar (Shanghai) Robotics Technology, Mondo Technology, National and Local Collaborative Humanoid Robotics Innovation Center, Radiant Tech Ventures, Shanghai Houxue Capital, Shanghai Seer Intelligent Technology, Shenzhen Agricultural Technology Innovation Group, Shenzhen Wuji Technology, Universal Cosmos, VinMotion General Purpose Humanoid Robots Application Development and Research Joint Stock Company, Westwell Holdings (Hong Kong), Yushu Technology, and Zhejiang Humanoid Robotics Innovation Center.

About the Hong Kong Embodied AI Lab 

In alignment with the National 15th Five-Year Plan and global trends in AI and robotics development, the Hong Kong Embodied AI Lab is dedicated to building a full-stack research and development platform for embodied intelligence. Its research covers core technologies and systems ranging from joint modules and robotic hardware, control algorithms, manipulation algorithms and perception algorithms to humanoid robots, dexterous hands, quadruped robots, robotic arms and novel robot morphologies, with a focus on diverse real-world applications.

More details: https://www.cpr.cuhk.edu.hk/en/press/cuhk-establishes-hong-kongs-first-full-stack-embodied-ai-lab-to-advance-ai-driven-industrial-transformation/

In celebration of the 60^th Anniversary of The Chinese University of Hong Kong (CUHK), a special issue was published by the Advanced Materials in January 2025, with the theme on advanced materials research at CUHK: From Biomedicine to Electronics and Beyond.

Under the editorship by Prof. Mao Chuanbin, Professor (Global STEM Scholar) from the Department of Biomedical Engineering at CUHK, this special issue has drawn international attention from the materials/device community. Advanced materials research is actively pursued on the CUHK campus by professors from various departments. This special issue featured various research articles contributed by members CUHK Faculty of Engineering and Faculty of Science, as well as some scholars from School of Science and Engineering at CUHKSZ.

Special Issue: 60th Anniversary of The Chinese University of Hong Kong: Advanced Materials: Vol 37, No 2

A research team from The Chinese University of Hong Kong (CUHK)’s Faculty of Engineering has developed a Vision-Language Model (VLM) integrated with spatial intelligence. This breakthrough enables robots to comprehend 3D spatial information like humans do, featuring scalability for visuo-tactile fusion[1], allowing them to autonomously perform complex, long-horizon manipulation tasks with various objects and further enhancing AI’s analytical capabilities. The findings have been published in the renowned international journal Science Robotics.

Although current VLMs allow robots to accurately understand human language instructions, they still lack a deep understanding of the 3D spatial relationships among objects, making it difficult to generate accurate plans for long-horizon manipulation tasks. To enhance the spatial understanding of VLMs, the CUHK team proposed a novel method called Retrieval-augmented Manipulation (RAM). This approach allows robots to simultaneously answer two critical questions during the planning process: what action to take at each step and how such actions can be executed feasibly in 3D space.

The team constructed a structured 3D object knowledge base for the robot, cataloguing the 3D geometries, stable placement configurations and graspable affordances of a variety of everyday objects. When generating a manipulation plan, the VLM retrieves relevant geometric and manipulation records from the knowledge base in real time. It evaluates physical feasibility to determine action sequences and intermediate states, while grounding abstract instructions in explicit spatial constraints. This equips the AI robots with the capability to handle long-horizon task manipulation.

The research deeply integrates vision-driven spatial intelligence with the long-horizon task planning capabilities of VLMs. By constructing a structured 3D object knowledge base, the VLM can dynamically retrieve the geometric and manipulation records of objects when planning long-horizon operations. This approach effectively extends the VLM’s language-level understanding and reasoning capability to complex 3D physical manipulation scenarios.

Professor Dou Qi, Associate Professor from CUHK’s Department of Computer Science and Engineering, who led the study, said: “Spatial intelligence is key to unlocking long-horizon manipulation, and visual perception is a crucial pathway to achieving it. Our method marks a breakthrough in bringing spatial understanding together with VLM reasoning.”

Professor Dou added that the proposed robot spatial intelligence technology scales effectively across tasks and platforms. In 14 manipulation tasks requiring spatial perception and covering 31 different objects, RAM enabled robots to accurately follow spatial language instructions, reason about 3D spatial relationships and perform adaptive manipulation conditioned on the scene’s physical context. RAM works seamlessly with leading VLMs and can be readily deployed on general-purpose humanoid robot platforms for fine-grained long-horizon manipulation.

Furthermore, CUHK’s newly developed system features scalability for visuo-tactile fusion, leveraging tactile feedback for more adaptive manipulation. Professor Liu Yun-hui, Choh-Ming Li Professor of Mechanical and Automation Engineering at CUHK, and Director of the Hong Kong Centre for Logistics Robotics (HKCLR), said: “This research demonstrates the potential of AI to advance robot manipulation, with promising applications across scenarios from industrial to household settings, which will ultimately help to improve human life.”

This research was supported by the HKCLR. Founded by CUHK, the centre is driven by a research team comprising professors from CUHK and the University of California, Berkeley. It is funded by the Innovation and Technology Commission of the HKSAR Government under the InnoHK Research and Development Platform. Its mission is to advance robot intelligence across perception, interaction, manipulation and mobility. Working closely with academic and industry partners in Hong Kong, the Greater Bay Area and the Chinese Mainland, the centre helps to translate cutting-edge AI and robotics research into real-world applications.

For the full research, please visit:  https://www.science.org/doi/10.1126/scirobotics.aea2092

[1] Visuo-tactile fusion is the process of combining visual data and tactile data to create a comprehensive understanding of an environment or object, enabling robots to perform complex, contact-rich manipulation tasks with human-like dexterity.

More details: CUHK develops VLM with spatial intelligence to improve AI robotic manipulation in complex tasks | CUHK Communications and Public Relations Office