A Decision Support System for Supplier Selection in Public Procurement: A Case of Banaras Locomotive Works, Varanasi

Abstract

This study presents a mixed-method approach for the problem of supplier selection in the context of public procurement. A recent tender invited by BLW, a unit of Indian Railways producing ‘diesel-electric’ and electric locomotives for Indian Railways and export market for the propulsion system (the costliest and most important item in the assembly of the locomotive), has been considered in this work. Initially, the criteria for evaluation of vendors are identified through the survey of experts from Indian Railways, who are engaged in procurement for the organization. These criteria are used to calculate the inter-se ranking of the offers received against the procurement tender. For the purpose of evaluation, the weights of the attributes are calculated using the Analytic Hierarchy Process (AHP) by taking the inputs for relative comparison of the attributes from experts through a questionnaire-based survey. Then, the ranking of the offers, as well as the relative importance of offers, is calculated using the TOPSIS method. Further, a mixed-integer programming (MIP) problem has been formulated to ensure the optimal order allocation in the multi-sourcing environment with an objective of maximizing the total value of the purchase. The model is populated with constraints that limit the order quantity, like selection of suppliers, the order quantity, minimum order quantity, and minimum number of suppliers to be selected. Results of the case are presented considering various scenarios. As a result of the study, it has revealed that the multi-sourcing of the suppliers for a specific item in given situation has an additional penalty on the organization. The effective and optimal use of the resources can be ascertained with the help of the proposed study. The biggest advantage of the study is engagement of potential suppliers, evaluators, and the end user. Continual system improvement is additional and inherent advantage reaped through proposed model. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023.