Case 1 Supply Chain Network Design

Stage 1 Determine the ability of the company to fulfill present and projected sales based on the current supply chain design. 

Stage 2 Evaluate how the company s ability would be improved, considering the potential expansions of plants and DCs already in the horizon. 

Stage 3 Optimize the global supply chain design that would deliver the best results for the entire 5 year time horizon. 

The MILP model had 7500 variables, of which 300 were binary and 7000 constraints. The mathematical model was coded in ILOG and solved using the CPLEX solver (www.ILOG.com). The solver reached optimality in 2 min. 

The stage 1 analysis showed a very close to full capacity utilization of the existing facilities under current demand levels. When considering future demand, the overall results showed that the existing supply chain could fulfill only 75% of the total projected demand, primarily restricted by production and distribution capacity. 

---

Constraints similar to Equation 5.1 arise in supply chain network design, where a company has the option to build warehouses of different capacities. In that case, x, is the quantity of product x stored at that location, xx is the square footage occupied by one unit of product x and by by b are the three potential warehouse capacities. Using binary variables, one for each b value, we can represent Equation 5.1 as a linear constraint. Define 5i, 82, 83 as the binary variables such that when 8j = 1, the RHS value is by Then Equation 5.1 can be written as. 

There are several published results of real-world applications using IP models for supply chain network design. We discuss briefly a few of the applications in practice. For interested readers, the cited references will provide more details on the case studies. 

Additional details on the MILP model are available in Chapter 8 as an illustrative case study for global supply chain network design. 

Porhllo, R. C., A. Ravindran, and R. A. Wysk. 2009 (September). Resilient global supply chain network design A case study. Working Paper. Department of Industrial and Manufacturing Engineering, The Permsylvania State University. 

The optimization models discussed in this chapter had a single objective— either to minimize supply chain costs or to maximize supply chain profitability, in case the product prices vary by location or customer. However, customer demand fulfillment and service are also important in designing a supply chain network. More recently, supply chain risk is emerging to be another important criterion (Supply chain risk is discussed in detail in Chapter 7). Hence, recent applications of optimization models have used multiple criteria optimization models for decision making. 

Data availability relates to information sharing. However, even if complete information sharing is in place, historical data might not be available. This problem is especially severe in the case of the design of a completely new supply chain network. Many statistical analysis methods used in the preselection stage depend... 

Tsiakis, P., Shah, N. and Pantelides, C.C. 2001 Design of Multi-echelon Supply Chain Networks under Demand Uncertainty. Ind. Eng. Chem. Res. 40,3585-3604. Simchi-Levi, D. Kamisky, P. Simchi-Levi, E., 2000 Designing and Managing the Supply Chain. Concepts, Strategies, and Case Studies. Irwin McGraw-Hill. 

The focus of the book is on the design and operation of the supply chain system, which involves connecting many production and distribution systems, often across wide geographic distances, in such a way that the businesses involved can ultimately satisfy consumer demand as efficiently as possible, resulting in maximum financial returns to those businesses connected to that supply chain system. The book includes several case studies on the design and operation of supply chain networks in manufacturing and healthcare. 

In some cases, companies want to design supply chain networks in which a market is supplied from only one factory, referred to as a single source. Companies may impose this constraint because it lowers the complexity of coordinating the network and requires less flexibility from each facility. The plant location model discussed earlier needs some modification to accommodate this constraint. The decision variables are redefined as follows ... 

The optimization model presented in this chapter was developed to design a supply chain distribution network for a consumer goods company located on a Caribbean island. This company is a major global competitor in its area, selling approximately 20 billion per year worldwide. The company has markets in more than 150 countries, and its brands are in first or second position in most of these markets. The country analyzed in this case study has sales of approximately 86 million per year. Its market includes 71 customers (66 retailers and 5 independent distributors), and it receives products from four manufacturing plants (all outside this country). To distribute the demand in this region, the company owns one distribution center located on this island. 

---