Multi-objective multi-product sustainable newsvendor management in an emerging economy: Evidence and applications

Abstract

The idea of a sustainable newsvendor problem has been considered by some researchers recently; nonetheless, an inclusive multi-objective model including all sustainability pillars with a multi-product supply chain has not been designed yet. In this paper, the newsvendor model is developed considering all sustainability criteria to study a two-stage procurement system. The classic newsvendor model of optimizing supplier-retailer payoff is considered in this research to illustrate the economic perspective of sustainability. Moreover, other objectives are formulated to model the social and environmental pillars of sustainability. These objectives minimize the environmental effects of the standard newsvendor problem regarding the malicious environmental effects of unrecyclable unsold products. To solve the formulated model a compromise programming approach is hybridized with Lagrange relaxation and sub-gradient optimization for the proposed model. For more clarification, a real-world case study from the Fast-moving consumer goods (FMCG) industry with real data has been employed alongside sensitivity analysis. Moreover, to evaluate the efficiency of the proposed solving approach, the Genetic Algorithm (GA) and Simulated Annealing (SA) algorithm were also designed and their parameters were tuned using the Taguchi method. Comparing the obtained results with the classical newsvendor model, the single economic objective loses at least 0.76% and at most 14.1% of its achievements by a trade-off with sustainable criteria. Furthermore, results indicated that both in terms of optimality and solution time, the proposed method has an extraordinary performance compared to GA and SA.