Network Design Methods

Network Design Methods

As the very name suggests, these methods determine the location of production, stocking, and sourcing facilities, and paths the product(s) take through them. Such methods tend to be large scale, and used generally at the inception of the supply chain. The earliest work in this area, although the term "supply chain" was not in vogue, was by Geoffrion and Graves [1974]. They introduce a multicommodity logistics network design model for optimizing annualized finished product flows from plants to the DC's to the final customers. Geoffrion and Powers [1993] later give a review of the evolution of distribution strategies over the past twenty years, describing how the descendants of the above model can accommodate more echelons and cross commodity detail.

Breitman and Lucas [1987] attempt to provide a framework for a comprehensive model of a production-distribution system, "PLANETS", that is used to decide what products to produce, where and how to produce it, which markets to pursue and what resources to use. Parts of this ambitious project were successfully implemented at General Motors.

Cohen and Lee [1985] develop a conceptual framework for manufacturing strategy analysis, where they describe a series of stochastic sub- models, that considers annualized product flows from raw material vendors via intermediate plants and distribution echelons to the final customers. They use heuristic methods to link and optimize these sub- models. They later give an integrated and readable exposition of their models and methods in Cohen and Lee [1988].

Cohen and Lee [1989] present a normative model for resource deployment in a global manufacturing and distribution network. Global after-tax profit (profit-local taxes) is maximized through the design of facility network and control of material flows within the network. The cost structure consists of variable and fixed costs for material procurement, production, distribution and transportation. They validate the model by applying it to analyze the global manufacturing strategies of a personal computer manufacturer.

Finally, Arntzen, Brown, Harrison, and Trafton [1995] provide the most comprehensive deterministic model for supply chain management. The objective function minimizes a combination of cost and time elements. Examples of cost elements include purchasing, manufacturing, pipeline inventory, transportation