Using artificial intelligence “agents” to enhance operational efficiency to take Chinese manufacturing to new levels
At a textile mill in Suzhou, Jiangsu, the rhythmic rattle of the loom is in harmony with the new non-sound: the quiet and rapid firing calculation (AI) of artificial intelligence.
China Telecom's Starweave Textile AI Agent works here. The thread tension is adjusted in real time, scrutinizing the fabric as it appears, and in the process it changes old ships.
The results speak to a 99% on time, a 20% surge in productivity, and a near-perfect amount of defect detection, China's Telecom said.
This is not an isolated experiment, but gives us a glimpse into the drastic changes that have rapidly become an essential engine and drive the evolution of China's industry.
Beyond the vast landscape of China's manufacturing industry, the wave of “industry and AI” is being driven by high-tech companies in the country rushing to develop AI agents to solve tasks.
An AI agent is a software system that uses AI to pursue goals and complete tasks on behalf of a user. Unlike traditional AI, which simply supports or provides recommendations, AI agents can run complex workflows without continuous human input. These agents are dedicated to performing professional roles within the company, including customer support, data analytics, compliance checks and even performing financial tasks, experts said.
Minister of Industry and Information Technology Li Lecheng chaired a conference held in Beijing in June, outlining a comprehensive strategy to accelerate AI innovation and integrated applications, thereby positioning AI as a core factor in China's new industrialization.
Xie Shaofeng, chief engineer of the Ministry of Industry and Information Technology, the country's top industrial regulator, said at a press conference that they have coordinated with the state government to develop 11 national AI innovation pilot zones.
Meanwhile, central and local governments are jointly building specialized manufacturing innovation centres for embodied AI robots, humanoid robots and other next-generation technologies to promote the development of industrial clusters.
Hong Kung Liang, a researcher at the Academy of Macroeconomic Research, National Development and Reform Committee, said: “Manufacturing is the focus of competition between key forces and technological hegemony. The fostering of future strategic advantages essentially depends on the quality development of manufacturing.
“Promoting AI-driven industrial upgrades is paramount to China's real economy over the next five years. And achieving critical technology breakthroughs is not an option. It is essential as we still face challenges in these areas.”
“We are pleased to announce that we are committed to providing a range of services to our customers,” said Ryoji Sekido, co-CEO of Accenture Asia Pacific.
“There is no suitable place to witness AI advances. In China and Shanghai, the rapid evolution of ecosystems is impressive.”
Sekido highlighted how Chinese companies are moving from isolated AI experiments to reinventing a comprehensive enterprise-wide. He pointed out electric car giant BYD as a case of a paradigm shift.
“This is not about simple AI use cases. It leverages the end-to-end AI from battery R&D and design automation to manufacturing and autonomous driving systems. It fundamentally drives market-to-market and reinvents operations to gain dominant market share,” Sekido said.
He stressed that this approach represents a “reinvention of a company” in which AI becomes a strategic engine rather than a productivity tool.
Statistics also check for surges.
Market Research Company International Data Corp (IDC) has revealed earthquake changes. The percentage of Chinese industrial companies deploying large models and intelligent agents has skyrocketed from 9.6% in 2024 to 47.5% in 2025.
Importantly, people implementing these technologies in multiple operational areas have exploded from 1.7% to 35%.
“We are moving rapidly beyond isolated pilots,” said CUI Can, senior research manager at IDC China. “High-value intelligent agent applications are landing and scaling.”
IDC forecasts China's industrial AI spending to reach 90 billion yuan ($12.53 billion) by 2028, spurring this accelerated adoption.
This is more than just automation.
AI allows for a fundamental evolution of how Chinese industries operate, moving from fragmented point solutions to interconnected, increasingly autonomous systems. “Industrial AI agents represent important levers,” CUI said.
“They drive the innovation cycle of key industrial assets (datasets and large-scale language models) and create a self-enhancing ecosystem for progress.”
Applications are rapidly diversifying across the industrial environment.
In May, Lenovo unveiled a manufacturing-focused AI agent designed to combine AI capabilities with decades of industrial know-how. The “Supply Chain Control Tower” agent offers incredible efficiency. Decisions increased by 50-60%, order fulfillment rates increased by 5%, manufacturing and logistics costs reduced by 20%.
The momentum is clear.
Huawei took part in the contest in FusionPlant 3.0 in June. FusionPlant 3.0 is a strategic platform explicitly designed to accelerate the creation and deployment of industrial AI agents and smart applications.
The message from these companies is clear. Industrial AI is no longer speculative. It is operational and provides concrete value.
At the same time, companies like Shanghai Wisdom Information Technology Co Ltd are pioneering “industrial AI agents co-driven by large language models” and are creating new digital applications and multimodal AI collaborations specifically designed for manufacturing environments with a focus on end-to-end optimization.
A sector-specific revolution has also been developed and exemplified by a subsidiary of Dongfang Precision Group. It focuses on AI-driven conversions for the packaging industry, offering Taylor Manufacturing Industrial Internet Platforms and Digital Upgrade Solutions.
Despite impressive advancements, the path forward demands more than isolated technical victory. Industry leaders and analysts likewise emphasize that the next stage in industrial AI maturity rests on robust ecosystem collaboration to overcome sustained hurdles.
However, the challenges are multifaceted and deeply interconnected. He said that breaking down these barriers is still a major barrier and that breaking down these barriers requires industry-wide standards and collaboration, as long as there is inconsistent data that completely hinders AI's potential.
Meanwhile, critical technical hurdles persist, including the limited generalization capabilities of large-scale models, while tackling highly specific and complex industrial tasks and critical bottlenecks of computing power.
“The lack of computing power essentially constrains development. However, training and inference utilization is often paradoxically low,” said Cao Kai, director of Smart Industrial Solutions at Baidu Smart Cloud.
The conversion cannot be denied. From Lenovo's streamlined global supply chain to the perfect fabric rolling from AI-supervised looms, Intelligent Agents clearly increase productivity, reduce costs and enable unprecedented accuracy.
As these agents expand and adapt learning and adapt within complex industrial ecosystems, experts added, they have evolved from sophisticated tools to the highly central nervous system of modern Chinese manufacturing.
“Innovation is a pivotal force driving the quality advancement of the manufacturing sector, and companies increasingly assume a central role in driving innovation,” says Li Jinghong, a scholar at the Chinese Academy of Sciences and a professor at Tingia University.
“We have great potential for growth in China's AI. With its vast markets and diverse application scenarios, China provides a unique space for AI innovation,” said Ma Jun, Senior Vice President of Volvo Group and President of Volvo (China) Investment.
“We look forward to seeing more creative AI solutions in the transportation and infrastructure sector, further reducing logistics costs and supporting sustainable development in China and the world.”
Please contact the author (masi@chinadaly.com.cn)
