Data-Driven Approaches to Smart Manufacturing Quality Management

Abstract

The comprehensive digitalization of manufacturing has triggered a paradigm shift in quality management, increasingly moving towards data-intensive methodologies. This paper critically examines various data-driven methods currently employed or planned for use in smart manufacturing systems, focusing particularly on their integration with quality assurance workflows. We systematically review literature from multiple interdisciplinary fields, revealing significant and fragmented phenomena in the current research landscape. Consequently, this paper identifies some persistent implementation challenges. Furthermore, translating these academic findings into practical and economically viable industrial solutions presents considerable challenges, often underestimated in purely methodological research. Therefore, this paper does not merely list existing technologies but attempts a comprehensive analysis, structurally examining their interdependencies, known limitations, and potential for synergistic combinations. We argue that the future of quality management in smart manufacturing lies not in pursuing a single, holistic solution, but in developing adaptive hybrid architectures that can fully leverage the advantages of different data-driven paradigms based on specific circumstances. This prompts us to further consider the necessity of new evaluation benchmarks and the crucial yet often overlooked role of human-machine interaction in these increasingly automated decision-making cycles. Further research is urgently needed to bridge the gap between methodological innovation and industrial applications, perhaps by developing more robust, uncertainty-aware, and transferable models.