A hybrid Decision Support System for automating decision making in the event of defects in the era of Zero Defect Manufacturing

Defects are unavoidable during manufacturing processes, and a tremendous amount of research aimed at improving defect prevention has been conducted by scholars. Zero Defect Manufacturing (ZDM) seeks to eliminate defects in production. In addition, technological advancements now allow the repair of defective products. This creates the need to re-schedule productions more frequently in order to take into account the actions necessary to fix defective parts. This study focuses on detection and repair-based ZDM strategies. It implements a newly developed, hybrid Decision Support System (DSS) that uses data-driven and knowledge-based approaches to detect defects and then automate the necessary decision-making processes. The system uses an ontology based on the MASON ontology in order to describe the production domain and enrich the available data with contextual information. Real time production data and past knowledge are utilized to analyse defects, identify their type and severity, and suggest alternative repair plans. Possible repair plans are evaluated using a dynamic multi-criteria approach that determines the plan most suited to production conditions at the time of defect detection. To test the efficacy of the DSS developed for this study, it was integrated with a dynamic scheduling tool and was also used in an industrial application in the semiconductor domain. The simulations and the real-world implementation both show that the proposed DSS system can efficiently detect defects and automate the post-detection decision-making process. The multi-criteria approach adopted by this study proves that the DSS can make well-adapted decisions, handle the dynamic nature of a production system, and help manufacturers move closer to Zero Defect Manufacturing.