ICVS - Intelligent Control of weaving machines using Vision Systems (2010 - 2013)

This project investigates how knowledge on fabric quality can be used to control looms.

To continuously improve productivity and production costs of fabrics, different feedback methods are essential.

To this end, modern looms are equipped with different feedback mechanisms. By means of online measurements, machine settings are continuously optimized. The optimization aims at three parameters

  • Production speed
  • Energy cost
  • Fabric quality

Quite a lot of research has already been done into feedback methods for production speed and energy costs, e.g. in the framework of previous research projects at our department: AUTOSPEED I en AUTOSPEED II, ARVD, ARVPS

The current feedback method for the third group, fabric quality, is fabric inspection. This is mainly done in a visual way, at best by the technician operating the machine. He can adjust the machine settings almost immediately. Chances of a successful adaptation depend largely on the experience of the technician involved.

Generally, feedback on fabric quality is a lot slower, usually occurring at an inspection table after weaving. The dead time between weaving and inspection results in loss of production and rapid feedback is out of the question.

In this project, fabric quality is optimized by means of an interactive vision system. The system will complement the feedback methods for production speed and energy costs.

The vision system is based on a powerful image analysis. The research focuses on:

·         Determining deduced parameters usable as feedback

·         Studying the parameter sensitivity

·         Relating to machine parameters and existing control algorithms

·         Controlling fabric quality

Supported by IWT

Duration: 6/2010 to 5/2013

Contact: Prof. Dr. ir. Lieva Van Langenhove