Uncertainty evaluation in the prediction of defects and costs for quality inspection planning in low-volume productions
See our publication: M. Galetto , E. Verna, G. Genta, F. Franceschini, Uncertainty evaluation in the prediction of defects and costs for quality inspection planning in low-volume productions (579.83 kB). Published on: THE INTERNATIONAL JOURNAL OF ADVANCED MANUFACTURING TECHNOLOGY. DOI: 10.1007/s00170-020-05356-0
Quality-inspection strategies play a pivotal role in providing consumers with high-quality and defect-free products. In order to withstand the competition, organizations have an increasing interest in designing quality controls that are effective in detecting defects and economically viable. Recent studies have proposed a preliminary method to evaluate inspection effectiveness and cost in low-volume assembly processes, characterized by the production of single units or small-sized lots, even spread in long periods. Based on this method, the present paper aims to define a procedure to evaluate the robustness of defect and cost predictions in quality inspections of low-volume productions. The research questions which are specifically addressed concern how the uncertainty of models for defectiveness prediction can be assessed, and how this uncertainty may affect the selection of the most effective and affordable inspection strategy. The proposed approach allows to accurately analyze and compare different inspection strategies in terms of effectiveness and cost. First, the uncertainties of the statistical variables of the model for defectiveness prediction are evaluated by applying the law of combination of variances. Then, by combining the contributions of several inspection design parameters, the uncertainty is propagated to two indicators which quantify the overall effectiveness and cost of inspection strategies. In order to test the proposed methodology, a practical application concerning the assembly of mechanical components in an industrial manufacturing context is presented and discussed.