Development of a data mining system for continual process quality improvement

Quality issues are important to manufacturers so that they can ensure their products are up to the highest standard. However, due to various causes associated with the operational processes, it is hard to identify all the sources of quality problems in a timely manner. A product's defects may cause a very serious loss in market share and low profit margin in the competitive environment of short product life cycles. In order to address this, a systemized and well-designed approach has been formulated to perform the desired task of problem identification based on the specific process. A data mining system (DMS), which is incorporated into emerging technologies including on-line analytical processing (OLAP) and decision tree-based artificial neural networks (ANNs), is presented. With this DMS, essential support for users who wish to identify the cause and source of problems can be provided so that immediate action can be taken for their rectification. The system prototype has been implemented in the factory of a manufacturer of magnetic heads for hard disk drives (HDDs) in order to validate the workability of the proposed methodology. © IMechE 2007.