The textile industry has long talked about intelligent transformation, but tools that truly cut into production decision-making have been scarce. On July 27, Zhongke Wenge launched the 'Industrial Decision Brain' in Keqiao, Shaoxing, directly targeting three chronic problems: data silos, model misalignment, and experience loss. Public data shows the product has closed loops in trade order processing, weaving scheduling, dyeing color matching, and dye vat scheduling. Weaving scheduling time has dropped from hours to minutes, and dyeing sample-making rounds have been reduced from 7-8 to 3-4, saving 30%-50% in costs. These figures mark a critical shift for AI from a 'display tool' to a 'decision engine.'
Background: AI Moves Beyond Concepts to Tackle Three Transformation Bottlenecks
The textile industry's IT foundation is not weak. ERP, MES, and PLC systems have been widely deployed for years, accumulating vast data. But as Zhongke Wenge CTO Cao Jia noted at the launch, data remains at the recording and query level, not yet transformed into analysis and decision support. The company summarizes three core obstacles: data is hard for AI to understand, models are hard to adapt to industrial contexts, and experience is hard to preserve. Factories don't lack equipment, systems, or data; they lack a decision-making brain connecting data to execution.
Zeng Dajun, deputy director of the Institute of Automation at the Chinese Academy of Sciences, highlighted in his keynote that with the rapid development of industrial robots, intelligent agents, and embodied intelligence, decision-making is shifting from human-centered to human-agent collaboration. This means future factory decisions will be driven by intelligent agents deeply involved in the entire process. Professor Liu Yuan from Zhejiang University emphasized that AI's technical capabilities do not automatically translate into industrial empowerment; they must deeply integrate with specific scenarios, business processes, and institutional environments. These insights point directly to the cognitive gap the textile industry must bridge.
Industry Impact: From Order Processing to Dye Vat Scheduling, Six Key Links Upgraded
The Industrial Decision Brain is not a conceptual framework; it has already been deployed in multiple production and management links. In trade order processing, a foreign trade order once required 30-60 minutes of manual cross-referencing across multiple files. After deployment, key information such as fabric specs, colors, quantities, processes, and delivery dates is automatically identified in tens of seconds, with accuracy exceeding 95%. This directly lowers the labor threshold for trade follow-up, meaning a leap in efficiency for small and medium textile trading companies.
Weaving scheduling is another high-frequency pain point. Traditional scheduling relies on manual experience to balance equipment, capacity, processes, inventory, and delivery, often taking hours. With the Decision Brain, AI automatically assigns tasks under multiple constraints and recalculates in real time as conditions change, compressing scheduling from hours to minutes. For weaving factories frequently facing rush orders, equipment anomalies, and order adjustments, this real-time responsiveness directly improves delivery performance.
The transformation in dyeing and printing is particularly noteworthy. Color matching traditionally depended heavily on veteran technicians' experience, requiring repeated sampling and adjustments. The Decision Brain generates color suggestions by integrating historical formulas, spectral data, fabric information, and dye properties, then optimizes based on sample results. This reduces sampling rounds from 7-8 to 3-4, cutting costs by 30%-50%. More importantly, veteran knowledge is continuously digitized into reusable assets, no longer lost when personnel change. Dye vat scheduling improvements are equally significant: single scheduling time dropped from about 2 hours to about 2 minutes, greatly enhancing the factory's ability to handle rush orders and equipment changes.
In the production completion and order delivery stage, a finished batch once required manual cross-checking of production records, quality inspection results, and original orders, involving sales, warehousing, quality control, and finance departments, taking 1.5 to 3 hours. The Decision Brain automatically verifies data and triggers subsequent processes, compressing cross-departmental collaboration to about 5 minutes. For textile foreign trade companies, this end-to-end efficiency gain translates into shorter lead times and lower operational costs.