Application of non-conformity matrix to predict system interactions in complex quality problems

Abstract

Assuring the highest quality at a realistic cost is a permanent challenge for commodities manufacturing. Complex processes, demanding customer specifications and high production rates make the traceability of Non-Conformities (NCs) a defiant task, requiring a holistic approach and methods to complement the traditional quality control tools. This paper presents a new tool, that addresses the different NCs generated along a production line in a matrix way and apply the Design Structure Matrix (DSM) principles for a more comprehensive analysis. This tool was applied in an industrial setting with the purpose of improving the quality of three-piece tin plate aerosol cans. Due to the complexity of the problem, where many NCs can be generated along the production process eventually causing defective aerosol cans, the application of this Non-Conformity Matrix (NCM) tool showed promising results, by a clearer identification of interactions among NCs. Subsequently, DSM algorithms of clustering and sequencing were applied on this matrix, pointing out which are the most important clusters of NCs throughout the overall manufacturing process, thus targeting the efforts of quality control engineers at the critical activities of the production process. Linking this tool with other process improvement tools, such as design of experiments and failure mode and effects analysis will significantly improve the final product quality, thus reducing the overall costs.