Designing profitable and responsive supply chains under uncertainty

In this paper, we develop a multi-objective two-stage stochastic programming model, which takes into account the selection of warehouse and retailer sites and the decision about production levels, inventory levels, and shipping quantities among the entities of the supply chain network. The first objective function is to maximise the chain's total profit over multiple periods, and the second objective function is to minimise the total travel times for unsatisfied customers, whose demands must be met by retailers which have been established in other markets, to maximise the chain's responsiveness. Demands, selling prices and productions times at manufacturing sites are all considered as uncertain parameters. The two objective functions are in conflict with each other, and we use epsilon-constraint method to generate a set of Pareto optimal solutions for the proposed multi-objective problem. We then generalise the case and assume the uncertain parameters are continuously distributed random variables and use a simulation approach called sample average approximation (SAA) scheme to compute near optimal solutions to the stochastic model with potentially infinite number of scenarios. A computational study involving hypothetical networks of different sizes and a real supply chain network are presented to highlight the efficiency of the proposed solution methodology.

Automated summary

In this study, a multi-objective two-stage stochastic programming model is developed to address various decision-making aspects within a supply chain network. By utilizing the epsilon-constraint method to generate a set of Pareto optimal solutions, treating uncertain parameters as continuous random variables, and employing the SAA scheme for near optimal solutions, the efficiency of the proposed solution methodology is demonstrated through computational studies involving hypothetical and real supply chain networks of different sizes.