SDAS - A KNOWLEDGE-BASED FRAMEWORK FOR ANALYZING DEFECTS IN APPAREL MANUFACTURING

The ability to produce high quality, defect-free garments is critical for the commercial success of any apparel manufacturing enterprise. Su ch a capability, however, requires the tools to analyze the defects, i dentify their causes and record the level of incidence of such defects . The Sewing Defects Analysis System (SDAS) is a knowledge-based syste m that has been designed and implemented for the classification and di agnosis of such garment manufacturing defects. The user's description of a defect based on visual characteristics and location is employed t o classify a defect and identify the manufacturing operation at which the defect was caused. SDAS also suggests remedies for the identified defect. The basis of the classification used in the SDAS, i.e., nature of the defect and its location, is applicable for all garment types. The SDAS diagnosis information is kept in text files which facilitates easy customization of the diagnosis output. SDAS also provides the us er with detailed military garment specifications of the construction f eatures pertinent to the currently observed defect. Its linkage to a d atabase is useful in order to track any trends which may be evident in the defects occurrence. SDAS has been developed to assist personnel r esponsible for inspecting finished garments. Personnel in charge of ga rment manufacturing can also use the SDAS for identifying causes of fr equently occurring defects.