- Step 01
The challenge
Manual quality inspection caught fabric weave faults, stains, and holes too late — often only after packaging. This resulted in a high batch rejection rate and costly customer returns.
- Step 02
What we built
We deployed a high-speed computer-vision line camera array integrated with a custom-trained defect detection convolutional neural network (CNN) model. The system watches fabric at full line speed and flags defects instantly on a live dashboard.
- Step 03
The result
Cut batch rejections significantly and reduced material wastage by catching loom defects at the source within seconds. Established a digital quality-assurance trail for every roll shipped.
We caught more defects in the first week of running LetinAI's cameras than our inspection team could catch in a month. Batch waste is down significantly and our export clients now get a quality log with every shipment.