As the wave of large language models sweeps across industries, the digitalization of the textile sector is accelerating in Keqiao. The 2026 Textile Industry Digitalization Development Conference on May 7 sent a clear signal: AI applications in textiles have moved from labs to workshops, from concept validation to large-scale deployment. The data and technical paths disclosed at the conference paint a more actionable roadmap for the industry.
Event Background
Yan Yan, Vice President of the International Textile Manufacturers Federation and Vice President of the China National Textile and Apparel Council, outlined three directions: promoting deep AI deployment, enhancing full-chain collaborative efficiency, and deeply integrating green manufacturing. The combination of a 'textile intelligent large model' and vertical-scenario small models was put on the agenda, indicating the industry is no longer satisfied with point-based digitalization but aims to build an intelligent base covering common knowledge and specialized capabilities.
Keqiao, as the 'International Textile Capital,' has formed a systematic digital layout. Sun Weigang, Deputy Secretary of the China Textile City Party Working Committee, introduced the '1+4+N' comprehensive intelligent agent system and the 'live streaming + platform + cross-border e-commerce + overseas warehouse' trade framework. These moves show Keqiao is extending digitalization from production to transactions, aiming to open up the entire digital trade chain.
Industry Impact
The TDSD® (Textile Digital Smart Dyeing) low-carbon digital intelligent dyeing process released by Hangzhou Huanyu Digital Technology Co., Ltd. became a technical highlight. The process achieves nearly 99% water saving, 33% carbon reduction, and 21% chemical reduction, using fully self-developed inkjet equipment and AI color management systems for 'print-and-dye' production. For dyeing companies facing EU carbon tariffs, these figures represent a viable path to cross green entry barriers.
AI fabric inspection technology was another major focus. Shanghai Kaiqian Intelligent Technology and Nantong Julian Digital Technology respectively demonstrated self-learning AI vision systems and multi-scenario inspection solutions. Traditional manual inspection suffers from recruitment difficulties, high omission rates, and delayed feedback. The core of AI inspection lies in 'teachable' systems that integrate with production management systems to achieve dual improvements in efficiency and accuracy. However, Luo Yucheng, Deputy General Manager of Shaoxing Keqiao Weaving and Dyeing Industry Brain Operation Co., Ltd., pointed out that challenges remain: algorithm adaptation for complex fabrics, system instability in complex environments, and high deployment costs for SMEs.
Hu Song, Director of the China Textile Information Center, proposed a five-step digitalization method in his report: 'diagnose current status, select scenarios, supplement data, run pilots, and expand capabilities.' He emphasized that SMEs should adopt low-cost, fast-result paths, while large enterprises should shift from system construction to data ecosystem leadership. This differentiated strategy addresses the long-standing dilemma of 'big companies dare not invest, small companies cannot afford it.'
Long Fangsheng, General Manager of Zhejiang Meixinda Textile Printing and Dyeing Technology Co., Ltd., shared a three-stage evolution path of 'shaping, soul-casting, and intelligence-enabling,' proposing that smart factories are not cost centers but order-creation centers. This view repositions digitalization from 'forced transformation' to 'active creation of incremental value,' offering insights for both buyers and factories.