New approaches for online visual inspection of products with multiple-characteristics and known tolerances

Abstract

The increasing customer requirements regarding better quality products and lower costs are challenges that should be resolved. Also, lean manufacturing, as an effective approach to increase efficiency of production systems, that is based on eliminating wastes and reaching excellence, makes it necessary to find and apply non-traditional approaches in manufacturing and quality control. Costs of product quality control are among those costs that do not add value to the product and should be maintained within limited ranges in order not to affect the final product cost. Either classical inspection methods (measuring methods) or even traditional quality control approaches are costly; and, unfortunately, do not guarantee defect-free products especially for products with multiple quality characteristics. In this work, an efficient online 100% visual inspection approaches are introduced. Those approaches are based on digital imaging of products and analysing their data using advanced mathematical approaches to extract the product features. Statistical tools are then applied to compare the extracted features with a master feature for the product. The developed approaches give a good performance in inspecting multi-dimension products at low costs and very little inspection errors.