Leveraging Linear Programming-based T-spherical Fuzzy AROMAN Method for Vehicle Routing Software Selection in Last-mile Logistics

This paper introduces a novel approach, the linear programming-based T-spherical fuzzy alternative ranking order method accounting for two-step normalization (AROMAN), designed to assist last-mile delivery (LMD) companies in the advanced selection of vehicle routing software (VRS). Selecting an effective VRS is a critical yet formidable challenge for LMD organizations due to the overwhelming variety of available software options and the inherent complexity of urban logistics. These operations involve intricate resource allocation, schedules sensitive to real-time traffic, and the constant need to align delivery paths with rapidly evolving customer preferences. Furthermore, the LMD environment is characterized by significant uncertainties and multidimensional constraints that traditional selection methods often fail to capture. By incorporating T-spherical fuzzy logic and linear programming principles, the proposed method effectively models these complexities and uncertainties, providing a methodical framework for assessing VRS alternatives based on a diverse array of criteria. The integrated model is validated through a practical case study of an LMD company operating in England. A panel of experts established specific evaluation criteria to identify the most favorable VRS solution. This research contributes to the field of logistics and transportation management by presenting a robust, datadriven methodology that enables LMD companies to improve their competitiveness and operational efficiency in dynamic marketplaces.