BEIJING, July 5 (Xinhua) -- When picturing a steel factory, words like "heavy," "smoky," and "arduous" often come to mind. But a technological breakthrough is proving that the iron and steel industry can be smart and sophisticated, thanks to the power of artificial intelligence (AI).

Shougang Group, one of China's largest steelmakers, is leading the way with an AI-powered visual system that is transforming the entire steelmaking process from labor-driven to model-driven. With this cutting-edge technology in place, the steel giant has boosted production efficiency by over 20 percent and reduced defects by 35 percent.

Exemplifying the shift toward digital and intelligent industrial transformation, the AI-powered system was named one of the top 10 benchmark applications at the 2025 Global Digital Economy Conference, which wrapped up in Beijing on Saturday. During the event, business leaders and industry representatives discussed the growing impact of AI across industries, highlighting China's broader push toward smart manufacturing.

"For a company of our size, the top priority in the face of the ongoing AI revolution is to explore how this technology can deliver better solutions to our challenges, and how it improves quality, increases efficiency and reduces cost," said Jiang Xingqun, senior vice president of BOE Technology Group Co., Ltd., one of the world's largest display panel manufacturers.

"Traditional visual technology has been widely used for quality checks, but it still falls short in addressing some issues," Jiang said. "This is why we employ AI for data labeling, model training, and detection of defects and abnormalities." He noted that in recent years, BOE has also developed its own automated decision-making system, which is already being used in production.

The company has been actively seeking ways to integrate AI into display panel manufacturing, a comprehensive process that encompasses quality inspection, monitoring and analysis, defect repairs, and equipment maintenance. With the help of AI, BOE has effectively improved graphic processing efficiency, shortened defect handling cycles, and reduced labor costs.

Similarly, numerous companies across the country are accelerating their digital transformation efforts. According to the China Internet Development Report 2024, the country now has nearly 10,000 digitalized workshops and intelligent factories. Of these, more than 400 have been recognized as national-level benchmark factories in smart manufacturing, utilizing technologies such as AI and digital twins.

Industry insiders highlighted that the costs of large models have dropped sharply over the past year, creating favorable conditions for the application of AI technologies. Ruan Yu, vice president of Baidu, noted that as costs continue to fall, large AI models are becoming a core productivity tool for an increasing number of enterprises.

According to Zhang Di, vice president of Kuaishou Technology and head of Kling AI, AI video generator technology will continue to evolve. By providing interactive and lifelike environments, it will further boost the development of the industrial internet and accelerate the digital transformation of manufacturing and other traditional industries, Zhang said.

In recent years, the Chinese government has introduced a wide range of measures to accelerate AI innovation and promote its application, intending to support new industrialization and the development of the industrial sector.

"The industrial sector is a cornerstone of China's economy, characterized by its large scale, wide scope and diverse scenarios. It is a primary arena for AI applications," said Zhao Chaofan, an official with the Ministry of Industry and Information Technology. Zhao added that the ministry will continue to actively advance AI adoption, organize regular initiatives to integrate AI with new industrialization, and promote model solutions for AI applications. These efforts aim to ensure that the technology more effectively supports the real economy, Zhao said.

Looking ahead, Gu Weixi, researcher at the China Academy of Industrial Internet, noted that several challenges remain for AI applications in the industrial sector. For instance, AI service providers lack sufficient channels to reach customers, while manufacturing companies have limited knowledge of cutting-edge AI technologies.

To address these challenges, several analysts have called for better alignment between supply and demand, which would help AI products enhance performance while reducing costs and increasing efficiency in the industrial sector.

"China is equipped with a comprehensive range of industrial categories, which provide access to diverse application scenarios and abundant data," said Mao Shengbo, partner at MiraclePlus, a startup incubator. "By fully leveraging China's strengths in its industrial chain, application scenarios and market size, we can create an entirely new AI supply chain and ecosystem." ■