Data Enabled Failure Management Process (DEFMP) across the Product Value Chain

Abstract

The continuously increasing amount of production data and the advancing development of digitization solutions promote advanced data analytics as a promising approach for failure management. Beyond the consideration of single units, examining the end-to-end value chain, including development, production, and usage, offers potential for failure in management-related investigations. Nonetheless, challenges regarding data integration from different entities along the value creation process, data volume and formats handling, effective analytics, and decision support arise. The CRISP-DM approach has become a widely established reference as a conceptual framework for data-driven solutions. However, the linkage between existing failure management procedures and the subsequent development of data-driven solutions needs to be specified. Accordingly, this paper presents a cross-value chain Data Enabled Failure Management Process (DEFMP). The central element is a process model to implement a cross-value chain data-enabled failure management, considering established quality management and data analytics approaches. Based on available failure, product, and process knowledge along the value chain, a path towards developing a comprehensive decision support system is shown. DEFMP combines a reactive failure process with a data-driven approach to incorporate data analytics for proactive improvements. Using DEFMP, the failure management process of a commercial vehicle manufacturer is adapted. With this, partial automation of failure management is made possible. In addition, the potential for improvements is identified and prioritized.