Um modelo para segmentação de fornecedores resilientes baseado em hesitant fuzzy linguistic term sets combinados com QFD e VIKOR

Abstract

Strategic supply chain management ensures the continuity of activities in organizations. The Covid-19 pandemic led supply chain managers to seek alternatives to the challenges faced, building resilient chains to deal with risks and supply interruptions. In this context, supplier segmentation proves to be very useful, as it defines the type of appropriate relationship between the purchasing company and some specific groups of suppliers. However, through a systematic review of the literature, only two previous supplier segmentation models were identified that take into account criteria related to supply chain resilience, which present limitations related to supporting decision-making under uncertainty and hesitation. To meet some research opportunities identified in the literature, this work aims to propose a resilient supplier segmentation model, based on the Hesitant Fuzzy Linguistic Term Sets method combined with the QFD (HFLTS-QFD) and VIKOR (HFLTS-VIKOR) methods. . HFLTS-QFD was used to define and weight supplier evaluation criteria, while HFLTS-VIKOR with possibility distributions was used to group suppliers within the segmentation matrix. A computational tool based on the aforementioned methods was implemented in Microsoft Excel to support decision making in supplier segmentation. The application of the proposed model based on a data collection instrument validated by the pilot company in the automotive sector and the performance of sensitivity analysis tests showed that the results produced are consistent. Three groups of suppliers were found, evaluated based on the segmentation matrix used. Therefore, to assist the company participating in the research, it was possible to indicate specific actions to improve the overall performance of each group of suppliers. Unlike previous studies, the model is suitable for group decisions under uncertainty and hesitation, allowing the use of linguistic expressions in judgments. It also allows the attribution of weights to experts and has no limit on criteria or evaluation alternatives. Therefore, the proposed model appears to be a viable option for segmenting suppliers, taking into account factors that intensify and reduce resilience. The participation of a real pilot company and the proposition of a decision-maker tool to assist in organizational development strengthen the adherence of the development of this work within the Master’s program in Administration.