logo

Lightning Talk: Demystifying Challenges/Learnings in Converting an Old-school Textile Inspection Machine into a Smart System Using AI/ML

2022-06-23

Authors:   Neethu Elizabeth Simon, Scott Thomas


Summary

Converting an old-school textile inspection machine into a smart system using AI/ML is effective and affordable even in the commodity fabric manufacturing industry.
  • Textile inspection is traditionally labor-intensive and error-prone.
  • Computer vision-based AI/ML solution using open source tools was developed for textile defect detection during the fabric inspection process.
  • Old-school manual fabric inspection machine was successfully integrated with cameras and open source AI/ML tools running on high-performance compute device.
  • Reasonably priced system was affordably applied to a much lower cost labor-intensive industry without expensive retooling or excessively high-priced technology.
  • Implementation and integration challenges encountered during design and development of this unique solution were resolved.
  • Model worked but was not scalable enough and was sensitive to folds and creases.
  • Inferencing was good but the system was not robust enough to handle high motor speed.
The textile industry has very tight margins and can't afford a lot of defects. The state of the art of textile inspection involves a person standing there for a maximum of two hours a day because they become brain dead by the end of two hours. This is ripe for automation. The team successfully integrated cameras and open source AI/ML tools with an old-school manual fabric inspection machine to create a reasonably priced system that can be applied to a much lower cost labor-intensive industry without expensive retooling or excessively high-priced technology.

Abstract

Textile inspection has traditionally been very labor-intensive and error prone. Because labor costs are very low in most fabric manufacturing locations, any automation must be extremely cost-effective. Team, comprising of members from multiple teams across multiple Geos, developed a computer vision based AI/ML solution using open source tools for textile defect detection during the fabric inspection process. Team successfully integrated the eyes, i.e cameras and open source AI/ML Tools running on high-performance compute device with an old-school manual fabric inspection machine. Results show that the inferencing capabilities offered through hardware and software solutions are effective and affordable even in the commodity fabric manufacturing industry. Hence a reasonably priced system was affordably applied to a much lower cost labor-intensive industry without expensive retooling or excessively high-priced technology. This presentation will cover the implementation & integration challenges encountered during design & development of this unique solution. It will also cover the resolution adopted to enable developers to better design and implement IOT, AI/ML & Computer Vision based solutions.

Materials:

Post a comment