Haier Chief Urges China to Build National Embodied Intelligence Open Innovation Platforms

(Yicai) March 5 -- Zhou Yunjie, chairman and chief executive of Chinese home appliance giant Haier Group and deputy to the National People's Congress, has called for the establishment of national open innovation platforms for embodied intelligence to speed the effective development of new technologies and the in-depth application of related products.China should build national-level embodied intelligence open innovation platforms, initiate a special national data project for embodied intelligence, and accelerate the formation of a standard system for industrial-grade embodied intelligence robots covering the entire chain from design, production, and testing to application, Zhou told Yicai yesterday ahead of the NPC annual meeting, which kicked off today. It should also promote the setup of standards for robot system integration and process adaptation in specific scenarios, he added.The meeting of China's parliament is one of the country's two most important annual political conferences and runs through March 12, while the Chinese People's Political Consultative Conference, the top political advisory body, opened yesterday and ends March 11.China should also set up a national key research and development project on embodied intelligence that is dedicated to making breakthroughs in core algorithms, including multimodal perception and understanding, environment adaptive learning, and bionic dexterous manipulation, and support the R&D and industrialization of high-precision sensors, intelligent joints, real-time control systems, and other key hardware components, Zhou noted.He suggested China should experiment with the in-depth application of embodied intelligence in specific application links, including complex assembly, precise inspection, high-risk operations, and personalized customization, and create several replicable and scalable benchmark cases.China's embodied intelligence industry still urgently needs to break through multiple development bottlenecks, including that artificial intelligence models, precision mechanisms, drive control systems, and other relevant technologies have not yet formed an efficient collaborative mechanism, there are no industry-level data sharing mechanisms, high-fidelity simulation training platforms are in short supply and with not deep enough application, and there are no standards for unified hardware interface and safety and performance, Zhou said, adding that implementing his suggestions will help overcome these difficulties.Editors: Tang Shihua, Martin Kadiev