The Design of Multi-Product Multi-Echelon Inventory Systems Using a Branch-and-Bound Algorithm,

Large-scale distribution systems require a hierarchy of retail stores and warehouses to satisfy the demand of their customers. Given the maximum number of installations and their possible locations, the problem is to determine which installations to include in the design of the system and which products to stock at these installations. Demand for the products is assumed known and may be deterministic or stochastic. The objective is to find a solution to this design problem which minimizes the total expected discounted cost for the lifetime of the system. This problem has been formulated as a combinatorial optimization problem and solved by a branch-and-bound algorithm. The subproblems of the algorithm are tractable integer linear programs. Applications of this model to the design of other multi-use, multi-facility systems are briefly described. Author