Artificial intelligence models are beginning to handle steelmaking, while welding robots inherit the nuanced skills of master craftsmen—a transformation is underway in Chinese factories. Recently, China announced its first list of 15 leading smart factories, recognized as the pinnacle of intelligent manufacturing in the country. A recent roundtable discussion brought together government officials and representatives from the selected enterprises to explore how these "new factories" are redefining manufacturing.

What sets these top-tier factories apart? According to Su Bo, Director of the National Expert Committee on Intelligent Manufacturing and former Vice Minister of the Ministry of Industry and Information Technology, "Leading smart factories represent the highest level of digital, networked, and intelligent development in manufacturing. China has established a four-tier system for smart factories: basic, advanced, exceptional, and leading. The leading tier sits at the very top, representing enterprises with the most advanced intelligent manufacturing capabilities."

What makes the selected enterprises stand out? Ao Li, Vice President of the China Academy of Information and Communications Technology, highlighted three core elements: leadership in manufacturing models, technological innovation, and developmental outcomes. He shared relevant data: "Among the 15 factories, the adoption, utilization, and integration rate of artificial intelligence exceeds 70%. On average, production efficiency has increased by 30%, while the defect rate of products has decreased by nearly 50%."

The core of intelligence lies in what is invisible. When AI-generated factory images were displayed, two business leaders agreed that true intelligence is not visible to the eye.

Yang Dongsheng, Party Committee Secretary and Chairman of XCMG Group and XCMG Machinery, noted, "The most defining feature of a leading factory is the data-driven upgrade of end-to-end value chain capabilities. We have achieved full visibility, transparency, management, and mobilization of XCMG’s global resources, enabling a globally customized agile delivery model. We have found the solution to balancing mass production and personalized customization—shifting from 'a thousand units looking the same' to 'a thousand units, each unique'."

Liu Baojun, Deputy Party Secretary and General Manager of Baosteel, a subsidiary of China Baowu Steel Group, expressed strong agreement: "The core of a truly smart factory is invisible. In the steel industry, the essence of intelligent manufacturing lies in model-driven approaches based on data." Ao Li added, "Baosteel’s intelligent decision-making hub for silicon steel serves as the true brain of the smart factory."

Deep integration of AI: Coexistence and progress, not replacement. A high penetration rate of artificial intelligence is a strict criterion. Su Bo emphasized, "In practical applications, it must reach at least 60%. All enterprises must progressively achieve digital, networked, and intelligent manufacturing through phased cultivation. This is not only the future trend of manufacturing development but also a fundamental requirement for survival."

Yang Dongsheng stated, "This is not a matter of willingness but of actively embracing new technologies. Among 38 scenarios across five major segments, AI is currently applied in 25. After implementing smart factories, the delivery cycle for global mobile cranes has been shortened by 55%, reflecting the intelligent upgrade of the entire value chain."

Liu Baojun remarked, "If we fail to leverage digital and intelligent technologies, we will become outdated tomorrow." In the production of silicon steel, often referred to as "the art of steel products," Baosteel has embedded numerous AI models to manage 1,080 key process parameters. Liu shared, "With the application of artificial intelligence, silicon steel inventory has decreased by 50% compared to before."