A fuzzy event tree methodology modified to select and evaluate suppliers

• Published in: South African Journal of Industrial Engineering Vol. 31; no. 1; pp. 35 - 46
• Main Author: Ozfirat, P. Mizrak
• Format: Journal Article
• Language: English
• Published: Bedfordview South African Institute of Industrial Engineers (SAIIE) 01.05.2020; South African Institute for Industrial Engineering; The Southern African Institute for Industrial Engineering; Stellenbosch University
• Subjects: analytic hierarchy process; Component and supplier management; Decision making; Engineering, Industrial; Event tree analysis; fuzzy event tree analysis; Fuzzy logic; Fuzzy sets; Human error; Manufacturers; Manufacturing; Methodology; Methods; Probability; Ranking; Risk analysis; Studies; supplier evaluation; supplier selection; Suppliers
• ISSN: 1012-277X; 2224-7890; 2224-7890; 1012-277X
• DOI: 10.7166/31-1-2027

The supplier selection problem becomes more urgent as competition in the market increases. Quality, cost, and the timely delivery of a product mostly depends on the manufacturer’s suppliers and the materials supplied. Therefore manufacturers are very elaborate in selecting their suppliers and work hard to develop supplier selection strategies. In this study, event tree analysis (ETA) is used to solve a manufacturing firm’s supplier selection problem. ETA is a method that is traditionally used for risk analysis problems, combining the probabilities of risk occurrences subject to the necessary precautions. In this study, this structure is used to select and evaluate suppliers. An event tree is developed to analyse each possible supplier, with branching being used according to the supplier selection criteria. The probability of each branch is set as the performance value of the supplier according to the selection criteria. Finally, the supplier is evaluated by combining all performance values on an event tree basis. Fuzzy logic is also incorporated into the event tree methodology to decrease human error and the effect of uncertainty. Fuzzy triangular numbers are used to denote the performance values of suppliers, and fuzzy ranking is used to distinguish the suppliers into classes. The proposed methodology is applied to nine possible suppliers of a specific material. The results reveal that two of the suppliers dominate all the others in the fuzzy ranking.