专家观点

Data And Skills, Not Algorithms, Hold Back AI In Textiles, Finds New ITMF And ITA White Paper

Textile World 新闻2026-10-16 22:23查看原文 ↗

✦产业研判

产业概括：ITMF与ITA白皮书：数据与技能制约纺织AI落地

ITMF与亚琛工业大学ITA联合发布纺织价值链AI战略指南白皮书，基于2024至2026年十场网络研讨会及2025年ITMF大会讨论。白皮书指出，AI在纺织服装业落地的主要障碍并非算法本身，而是数据整合不足、技能短缺与治理缺口，行业呈现“数字化碎片化”特征。AI已在织物纱线视觉检测、预测性维护、能耗优化和需求预测等场景产生可量化价值，例如某非织造产线应用因果AI使废料减少10%。报告认为端到端AI工作流的现实周期约为两到三年，信任与数据控制是应用前提。

面料服装方向中性强度中等周期中等

关键观察变量

纺织企业数据整合与结构化程度、行业AI技能培训供给与覆盖、AI治理与数据安全规范建立进度、因果AI与代理系统在产线中的实际部署案例数、端到端AI工作流从试点到规模化的时间、白皮书建议在各国产业集群中的落地跟进

判断边界白皮书为行业战略指南，未披露纺织服装全行业AI渗透率、投资规模或分区域数据；10%废料减少仅为单一非织造产线案例，不能外推为全行业普遍成效；端到端AI两到三年为专家估计，非已实现事实；未涉及批发市场商户、客流、租金或库存等流通环节数据。

Textile World

Breaking News

International Textile Manufacturers Federation (ITMF)

TPA

TWA

ZÜRICH, Switzerland / AACHEN, Germany — October 15, 2026 — The International Textile Manufacturers Federation (ITMF) and the Institut für Textiltechnik of RWTH Aachen University (ITA) today published “AI in the Textile Value Chain: A Strategic Guide to Artificial Intelligence for the Global Textile Industry”.

The white paper finds that the main barriers to AI adoption in the textile and apparel industry are poor data integration, skills shortages and gaps in governance, not the technology itself. It argues that the industry’s progress will depend on coordinated, industry-wide capability-building.

The paper draws on the ITMF & ITA webinar series AI in the Textile Value Chain (2024–2026). Across ten sessions, more than 20 experts from companies, research institutes and policy bodies spoke in the series. The paper also draws on a World Café discussion with practitioners at the ITMF Conference 2025 in Yogyakarta, Indonesia.

According to the authors, AI is becoming relevant because of structural pressures rather than technological hype. These pressures include demographic decline, persistent skills shortages, volatile markets, and rising demands for quality, traceability and sustainability. Decades of digitalization have left the sector with abundant data but fragmented systems. As a result, the paper describes the industry as “not under-digitalized, but digitally fragmented.”

Key findings include:

AI is already delivering measurable value in well-defined applications. Examples include computer-vision fabric and yarn inspection, predictive maintenance, energy optimization and demand forecasting. In one example cited from a nonwovens line, causal AI reduced scrap by 10%.

Data and skills are more limiting than algorithms. Participants warned that AI does not fix poor data but amplifies its flaws. Many companies hold far less usable, structured data than they believe.

AI is moving from decision support to action. Agentic systems are starting to coordinate multi-step processes such as sourcing. Causal AI is beginning to adjust production settings directly, in both cases under human supervision. The realistic horizon for end-to-end AI workflows is estimated at two to three years.

Trust is the decisive currency. Industry participants named clear conditions for AI use: control over sensitive data, strong cybersecurity, transparency about system limits, and preserved human accountability.

The paper recommends that companies invest in people before scaling technology and treat data as shared infrastructure. It also advises proceeding incrementally, starting with bounded use cases. It highlights a strong role for industry associations and pre-competitive collaboration, particularly for SMEs. Examples include shared data standards, data spaces that keep ownership with the data originator, and sector-specific tools such as a controlled “Textile GPT.”

“AI is no