Fu Hsun Fiber Industries has unveiled FabGoalie AI, an advanced AI-powered defect detection system, at booth L1109 during this week’s TITAS 2025 (14-16 October). Developed in collaboration with National Yang Ming Chiao Tung University (NYCU), the prototype represents a new milestone in intelligent textile inspection.

At the heart of FabGoalie AI lies an algorithm capable of detecting and classifying more than ten types of fabric defects, including oil stains, dye stains, holes, broken yarns, snags, and creases. The system also assesses colour consistency across the width of the fabric, identifying perceptual differences between the left, centre and right sections.

“The prototype is an important milestone for us. With FabGoalie AI, fabric producers can overcome the limitations of traditional manual inspection. I am glad that we have been able to push the boundaries once again,” said Tony Chen, Chairman and Board Member of Fu Hsun Fiber Industries.

FabGoalie AI supports roll-to-roll inspection, using seven cameras and an optical marking system to analyse fabric at speeds of 25–30 metres per minute. The compact hardware design allows easy integration into production lines, while a dedicated display and workstation enable real-time monitoring.

The system’s graphical user interface (GUI) includes preconfigured tables, charts and QR codes for quick access to dashboards. A multilingual web app further enhances accessibility and long-term analytics, catering to decision-makers in production environments.

Fu Hsun and NYCU are currently testing the system in live production settings and are seeking partners to broaden its deployment. Collaborating fabric manufacturers will help train the AI on a wider range of materials, structures and defect types, strengthening its accuracy and adaptability.

FabGoalie AI is not yet commercially available, but demonstrations and launch updates can be requested via the official website. Or you can watch the FabGoalie AI movie...

www.fabgoalie.ai