Reconfiguration of Quality Gates Based on Part Variation Modes in Multi-stage Manufacturing Systems

The growing production of high-tech, high-value products in strategic European sectors blended with the perception of zero-defect manufacturing (ZDM) set a premium on defect management strategies in multi-stage systems. The state-of-art end-of-line quality gate mechanisms face the challenge of variability and error accumulation from heterogenous sources, originated in up-stream process stages. On the contrary, the adoption of in-line monitoring and data-gathering systems allow defect detection and prevention with lower complexity and higher efficiency. However, the implementation of such stages, especially into mass-production environments, requires a comprehensive product- to system-level analysis. In this context, the integration of product-driven knowledge at system-level can be pursued by means of part variation modes (PVMs), which represent synthetic characterization of product deviation patterns. Therefore, the collected set of information can be further exploited to identify the quality fault root-causes and effects on the process sequence. This paper introduces a novel methodology for the evaluation of quality gates allocation in multi-stage high-volume production lines. The method integrates the quality information provided by PVMs with performance evaluation of production system based on the system dynamics. Preliminary results are provided for a real industrial case study from the automotive sector, demonstrating the significant advantages of proposed approach.