In December 2025, the Beijing Key Laboratory for Industrial Embodied AI, led by the Institute of Automation, Chinese Academy of Sciences, and jointly built by Beijing Zhongke Huiling Robotics Technology Co., Ltd. and State Grid Beijing Electric Power Company, was officially approved upon review and accreditation by the Beijing Municipal Science and Technology Commission and the Zhongguancun Science Park Administrative Committee.
The establishment of the Laboratory marks a key strategic move by Beijing to lay out the industrial embodied AI track. It signifies the formation of a complete capability system for embodied AI in real industrial environments, covering theoretical exploration, system integration, and practical application deployment.
Targeting core scientific challenges in complex and open environments — including unified spatial perception of embodied agents, generalized embodied control, and efficient heterogeneous collaboration — the Laboratory conducts research on theoretical methodologies and key technologies in the field of industrial embodied AI. It aims to make breakthroughs in critical technologies such as world model-driven 3D representation and autonomous navigation for complex dynamic scenarios, embodied large models and policy optimization oriented to human–robot interaction and motion transfer, as well as the design and optimization of multi-terminal collaborative architectures based on multi-agent heterogeneous cooperation. In response to national strategic demands in mining, manufacturing, energy transmission and other key industries, the Laboratory carries out technological integration in priority industrial sectors. It builds integrated software and hardware embodied platforms and systems, and launches demonstration applications in three major industrial scenarios: embodied quality inspection, embodied mining, and embodied power inspection. The Laboratory supports the rapid integration, effective verification and iterative evolution of embodied systems in practical engineering scenarios, and promotes the formation of an application paradigm for embodied intelligent systems with strong universality and high engineering practicability.
