The textile industry's digital transformation is moving from informatization to true intelligence, but the leap to decision-making AI remains fraught with hurdles. On July 27, a conference focused on AI-textile integration was held in Shaoxing Keqiao, where Zhongke Wenge officially launched its 'Industrial Decision Brain', targeting the critical gap between factory data and production execution.
Background and Pain Points
The event brought together representatives from the Institute of Automation of the Chinese Academy of Sciences, the China National Textile and Apparel Council, and several universities. The core question was how to move AI beyond buzzwords to solve real production decision-making problems in textiles.
Zeng Dajun, deputy director of the Institute of Automation, noted that decision-making is shifting from human-centric to human-agent collaboration, posing new challenges in data collection and human-machine coordination. Professor Liu Yuan from Zhejiang University emphasized that AI capabilities—data, computing power, and algorithms—must be deeply integrated with specific industry scenarios to generate real productivity.
Cao Jia, CTO of Zhongke Wenge, pinpointed the industry's dilemma: factories have equipment, systems, and data, but lack the intelligent brain to connect data to execution. He identified three key bottlenecks: data that is hard for AI to understand, models that are difficult to adapt, and expert experience that is difficult to codify.
Technology and Results
The Industrial Decision Brain is built on the DOMA (Data-Ontology-Models-Agents) architecture and a 'Base-Hub-Core-Brain-Edge' decision AI tree. It does not require enterprises to scrap existing systems but instead connects order, equipment, material, and process data to generate executable plans under constraints like delivery time, capacity, and cost.
In typical textile scenarios, the solution has achieved measurable results:
- Trade order processing: automated extraction of key information with over 95% accuracy, reducing time from 30-60 minutes to seconds.
- Intelligent weaving scheduling: AI allocates tasks considering equipment, capacity, and materials, compressing scheduling from hours to minutes.
- Intelligent color matching in printing & dyeing: reduces sample-making from 7-8 rounds to 3-4, cutting costs by 30%-50%.
- Intelligent dyeing vat scheduling: cuts scheduling time from about 2 hours to 2 minutes.
- Order completion feedback: cross-department coordination reduced from 1.5-3 hours to about 5 minutes.
Robots developed by Zhongke Juxi, including those for inspection, yarn hanging, and fabric unloading, operate under the unified brain, forming a closed loop of perception, decision, execution, and feedback. The inspection robot can identify over 20 types of textile defects with over 95% accuracy and an 8-hour battery life.
Industry Impact
This launch signals that textile AI is moving from single-point tools to full-process decision-making. Historically, AI in textiles was used for tasks like defect detection or process recommendation. The Industrial Decision Brain attempts to bridge the entire chain from data understanding to equipment execution.
For buyers and mills, the most immediate benefits are improved scheduling efficiency and cost reduction. A 30%-50% cost cut in color matching is particularly attractive for small-to-medium dyeing mills that rely heavily on veteran workers. Faster order processing and scheduling also help shorten delivery times, boosting flexibility for export orders.
However, the entry barrier remains. The system requires a certain level of IT maturity, including ERP and MES systems. For smaller factories still in the early stages of informatization, data governance and system integration may pose significant challenges.