Location of global logistic hubs within Africa based on a fuzzy multi-criteria approach. (June 2019)

• Record Type: Journal Article
• Authors: Essaadi, Imane; Grabot, Bernard; Féniès, Pierre
• Is Part Of: Computers & industrial engineering. Volume 132(2019)
• Page Start: 1
• Page End: 22
• Publication Date: 2019-06
• Subjects: Global logistic hub -- Facility location -- Multi-criteria decision method -- Fuzzy set theory -- Fuzzy TOPSIS -- AHP; Engineering -- Data processing -- Periodicals; Industrial engineering -- Periodicals; 620.00285
• DOI: 10.1016/j.cie.2019.03.046
• Languages: English
• ISSNs: 0360-8352

Abstract

Highlights: Locating a logistic hub is considered as a multi-criteria decision making problem. A modified version of Fuzzy TOPSIS is suggested. A new fuzzy distance measures the distance between two solutions. A confidence level is associated to each criteria assessment. The interest of the method is shown on the real case of a 3PL. Abstract: Choosing the location of a global logistic hub for reaching a new market can be considered as a "facility location problem" addressed through optimization based on quantitative criteria, or as a multi-criteria decision making problem using both quantitative and qualitative criteria. The second approach has been chosen for considering the real case of a logistic provider in Africa. The first originality of this article is to suggest a global framework positioning the different types of criteria that may be used, based on an analysis of the relationships between the concept defining the decision making context. For assessing the various decision criteria, an important requirement of the company was to reuse when possible assessments coming from trustable external sources, mainly international organization indexes (World Bank, World Economic Forum, etc.), and to complete them with knowledge coming from internal experts. This knowledge being often imprecise and uncertain, Fuzzy TOPSIS, often used for Multi Criteria Decision Making, is chosen as a global methodology. The standard method has been modified on two important points: (1) a new fuzzy distance is suggested to measure the difference between two solutions, allowing to postpone the defuzzification process until the end of the reasoning, (ii) a measure of confidence has been added to each elementary assessment, allowing the experts to clearly distinguish the fuzziness of an evaluation ("around 10") and the possibility that this evaluation is wrong. We show in an illustrative example how taking into account this new distance and the suggested confidence level may yield richer results than the standard method.