Clothes manufacturing companies have a strong need for improved quality of their products in order to keep their market interest against the keen competition of third countries. DOUROS S.A. is one of the most important companies in Greece manufacturing high quality clothes and wishes to develop a reliable, real-time quality control system for the inspection of both the incoming raw fabric and the finished garments. At present quality control is performed by human supervisors and thus the results are subjective and cannot cover all the defect types. Therefore, the reliability of the final product is reduced. Typical defect types of the incoming fabric are holes, colour shadows or oil spots, colour differences and thick yarns. Some common faults of finished garments are the incorrect sizing, the different distances of collar ends from the garment centre line, as well as the wrong pocket location or orientation. . . The present project aims at the incorporation of advanced visual and data acquisition techniques, as well as neural network technologies into the production line of a garment manufacture, in order to improve the real-time quality control of both the incoming fabric and the finished clothes. The new automated control system is expected to achieve the followings : at the first stage, the detectable size of the raw fabric defects will be reduced from a few mm now to less than 0 5 mm . At the second stage the error in measured size will be reduced from 10 mm to 1 mm.After the above improvements, the overall customers rejection percentage of finished goods is expected to be reduced at least one order of magnitude. The consortium consists of a manufacturer of high quality clothes (DOUROS S.A ) as the end user of the new quality control system, a company with expertise in industrial applications and automated vision systems (EDIXIA), two designers and manufacturers of mechanical systems and industrial equipment (ASATEC and TTE), an R&D Company specialized in parallel processing, Neural Network and expert systems (TECNOTESSILE), a research centre with expertise in textile industrial application and quality control processes (ITF), a rusk industry interested in applicating the proposed quality control practice in its production and packaging lines (ELITE S.A.), and a company developing automated control mechanical and electronic systems (General Automation Ltd). Brite/Euram III Areas covered by the proposal are: 1 1 and 2.3 of thc CRAFT Programme. In particular the research tasks of the proposed project comply with the short-term tasks decribed in 184.108.40.206, 220.127.116.11, 18.104.22.168, and the medium tasks in 1.1 3 M, 2 3 3.M.
Funding SchemeCRS - Cooperative research contracts
19009 Raphina (Pikeermi)
50046 Poggio A Caiano Prato