Development of an optimization model to determine sampling levels

Elizabeth A. Cudney,Ruwen Qin,Zlatan Hamzic
DOI: https://doi.org/10.1108/ijqrm-10-2014-0159
2016-04-04
Abstract:Purpose – As the complexity of the multi-component products increases the quality of these products becomes increasingly difficult to control throughout the supply chain. The first step to manufacturing a quality product is to ensure that the product components from suppliers meet specifications. Product quality can be controlled through sampling inspection of the components. The paper aims to discuss these issues. Design/methodology/approach – The model presented in this paper was developed to determine the optimal sampling levels for incoming lots containing parts for production and assembly of multi-component systems. The main objective of the model is to minimize the expected cost that is associated with a nonconforming item reaching assembly. Findings – In this research, the results showed that even with limited time available for inspection, performing sampling inspection significantly reduced the expected cost of a nonconforming item reaching assembly. The model, solved by the evolutionary algorithm, was able to provide a meaningful, near optimal solution to the problem. Originality/value – In this model the time available for inspection is limited, the distribution of defects is assumed to follow the binomial distribution, and the distribution of accepting the lot with defects follows the hypergeometric distribution. In addition, the inspection is considered to be accurate and, if a nonconforming item is found in the inspected sample, the entire lot is rejected. An example is given with real world data and the results are discussed as they relate to supply chain management and quality.
What problem does this paper attempt to address?