Supply chain planning using multi‐stage stochastic programming

Abstract

Purpose -The purpose of this paper is to explore the functionality of multistage programming approach on network supply chain structure. Design/methodology/approach -The general supply chain structure is considered and the supply chain planning model is developed using a two stage programming approach. The same model is extended to cover the applicability and advantages of a multi-stage programming approach. Findings -A multi-period supply chain model for new product launches under uncertain demand for supply chain network structure has been developed. The model allows simultaneous determination of optimum procurement quantity, production quantity across the different plants, transportation routes and the outsourcing cost in case of shortages. The proposed multi-stage model is compared with the standard two-stage model by examining the difference between the objective values of two solutions. The research clearly shows the importance of the multi-stage model as compared to the two-stage programming model. Research limitations/implications -The models developed here are limited to covering demand uncertainty, whereas real supply chain exhibits different uncertainties like capacity, processing time, etc. This can be the future direction for extending the work. Practical implications -The model is very useful in designing and planning the supply chain in an uncertain environment. The model allows the adjustment of the production plan as time progresses and uncertainties become resolved. Originality/value -The model uses a scenario approach to address the supply chain planning problem for a supply chain network structure under an uncertain environment and compares the two-solution approach for a set of problems. Generally supply chain costs are in millions of dollars and the saving using multi-stage programming can be significant.

Procurement risk management under uncertainty: a review

“…The third type is those seeking the optimal solution for multiple periods (e.g. Martel et al , 1995; Bonser and Wu, 2001; Yan et al , 2003; Nagar and Jain, 2008). Comparing with the single period model, more research can be done for multiple period model as long-term inventory planning is important, thus making it impossible to directly extend single period conclusion into multiple periods.…”

The design and planning of an integrated supply chain for perishable products under uncertainties

“…They considered both market demands and prices as uncertain parameters and performed the sensitivity analysis with 20 per cent deviation. Nagar and Jain (2008) considered both two-stage and multi-stage programming approach to study the supply chain network for new products under scenario-based demand uncertainty. The decision variables considered in their models were optimum procurement quantity, production quantity, transportation routes and outsourcing as well.…”

A review of discrete-time optimization models for tactical production planning

“…Uncertainty in the production environment is addressed by 39 references which employ different modeling approaches. Stochastic programming models are considered with distinct approaches, such as two-stage stochastic programming (Leung et al, 2006;Nagar and Jain, 2008;Schütz and Tomasgard, 2011;Wu, 2011;Zanjani et al, 2011) or multi-stage stochastic programming (Brandimarte, 2006;Denizel et al, 2010;Guan and Miller, 2008;Guan et al, 2009;Guan and Philpott, 2011;Koerpeoglu et al, 2011;Nagar and Jain, 2008;Zanjani et al, 2010b). Other stochastic programming proposals are presented by Karabuk (2008), Kim and Xirouchakis (2010), Sodhi and Tang (2009), Tempelmeier and Herpers (2011) and Tempelmeier (2007).…”