A robust fuzzy stochastic programming for sustainable procurement and logistics under hybrid uncertainty using big data

Today, in many organizations, the debate about the difference in core capabilities has become an important factor for market competition. Companies, based on the field of activity, decide to strengthen some of their capabilities, capacities, and expertise. Therefore, the focus of an organization on the strengths and efforts to develop its sustainability will lead to a competitive advantage in the marketplace. Due to changes in environmental factors, organizations have focused on carbon emissions in procurement and transportation that have the highest carbon footprint. This paper proposes a multiobjective, eco-sustainability model for a supply chain. The objectives are to minimize overall costs, maximize the efficiency of transportation vehicles and minimize information fraud in the process of information sharing within supply chain elements. Big data is considered in the amount of information exchanged between customers and other elements of the proposed supply chain; since there are frauds in information sharing then using big data 5Vs the model is adapted to control the cost of information loss leading to customer dissatisfaction. Since uncertainty is inevitable in the real environments, in this research hybrid uncertainty is considered. Because two sources of uncertainty are considered in most of the parameters, thus it is necessary to robustify the decision-making process. The model is a mixed integer nonlinear program including big data for an optimal sustainable procurement and transportation decision. A heuristic method is used to solve the big data problem that makes use of a robust fuzzy stochastic programming approach. The proposed model can prevent disturbances by using a scenario-based stochastic programming approach. An effective hybrid robust fuzzy stochastic method is also employed for controlling uncertainty in parameters and risk taking out of outbound decisions. To solve the multi-objective model, augmented epsilon-constraint method is utilized. The model performance is investigated in a comprehensive computational study. (C) 2020 Elsevier Ltd. All rights reserved.