Fuzzy multi-objective decision model for calibration supplier selection problem

Abstract

Quality products and competitively priced services are crucial in today's global markets. To provide quality at an acceptable price, companies seek out not only raw material and product suppliers but also calibration services suppliers. The calibration of measurement devices is one of the IS09001 standard requirements for quality. Although companies select and manage their calibration processes as economically as possible, cost is not their only selection criterion. Technical capability, documentation competence, performance history, warranties, and communications are also considered when selecting calibration suppliers. There are numerous criteria and methods in the current literature on supplier selection, but few studies have specifically examined the selection of calibration suppliers. This paper is the first study that specifically examines the selection process for calibration suppliers by utilizing selection criteria that were researched and presented prior to this study. Due to the varied linguistic expressions of criteria and the uncertain model parameters, this paper presents a fuzzy approach for selecting a calibration supplier. The model contains relevant calibration service quality parameters such as the weight of criteria, cost, calibration time, demand, technical capability, and number of experts. This study proposes a fuzzy multi-objective linear programming model that assigns the calibration supplier to the measurement device using two objectives: maximizing the fuzzy weight of the criteria and minimizing the fuzzy calibration cost through fuzzy calibration time, the number of certified experts, the technical capability of the company, and the measurement device demand. A two-phase approach is used to solve the fuzzy multi-objective decision model. Weight, cost, and calibration time are handled as fuzzy numbers for modelling the imprecise data. In an actual application, calibration suppliers are selected for 20 measurement device types and 161 measurement devices with the fuzzy multi objective linear programming model