Forward and Reverse Logistics Network Design with Sustainability for New and Refurbished Products in E-commerce


  • Yash Daultani1 (Indian Institute of Management Lucknow, India)
  • Naoufel Cheikhrouhou1 (University of Applied Sciences Western Switzerland, Switzerland)
  • Saurabh Pratap1 (Indian Institute of Technology (BHU) Varanasi, India)
  • Dhirendra Prajapati1 (Indian Institute of Information Technology Jabalpur, India)

There has been an enormous growth in the availability of refurbished goods in the online marketplace. These days, consumers can buy either the new products or refurbished products based on their budget and individual preferences. As a result, e-commerce firms need to redesign their existing forward and reverse logistics networks while focusing on supply chain sustainability. This paper proposes a novel forward and reverses logistics network design (FRLND) along with a consumer pickup and demand facility within the promised time window while addressing the complexities related to e-commerce platforms, suppliers, manufacturers, third-party logistics providers, retailers, and customer tiers. A mixed-integer non-linear programming (MINLP) model is developed to minimize the overall anticipated cost that consists of costs related to procurement, production, inventory holding, shortages, material for return units, recycling, repairing, disposal, and transportation cost, across the entire supply chain network. The problem under consideration is NP-hard in nature. The special challenges of the problem in consideration are to consider all pickup and distribution nodes of retailers/customers within the range of promising time horizons. For solution purposes, the Block-based Genetic Algorithm, Fruit-fly Algorithm, and CPLEX are used. Computational experiments show the comparative charts and trends that are put on to an extensive range of practical scenarios. The experiments reveal that the Genetic Algorithm performs well than the Fruit Fly algorithm in terms of rate of convergence and solution quality in all cases of interest. CPLEX solution provides the minimum optimal value.

Download full PDF Get metrics Rate article