To see how decisions at one part of a supply chain effect the overall performance of a system, we ran a simulation called the beer game. The supply chain consists of a retailer who orders from a distributor who orders from a wholesaler who orders from a factory. At the beginning of each period, each stage of the chain orders upstream and receives the order shipped out to them two periods ago (the order they placed 4 periods ago) unless the next stage upstream is backlogged. All orders are eventually filled when inventory becomes available. The holding cost specified for each location are (in $/keg.period): factory: 0.25, distribution center: 0.50, warehouse: 0.75, and factory: 1.00. Additionally, the penalty cost for a shortage is zero for all stages except the retail stores where the penalty cost is estimated to be $10.00 per keg/period.After trying many different strategies, the best policy I was able to come up with had a total cost of $122.00. This was achieved using choice 4, the base-stock policy. This policy re-orders a specified amount, less inventory on hand and pipeline inventory. The player specifies the base stock quantity for the retailer, warehouse, distributor, and factory. When this policy was used at each point in the supply chain, the lowest cost strategy was achieved. LocationBase Stock AmountCostRetail300101.55Warehouse21010.21Distributor2107.70Factory1503.41Total122.87Because the retail store encounters such a high penalty for shortages, it is best to keep them well stocked. They also have the highest holding overagecost, but at $1.00 it is only 1/10 of the shortage underagecost. If the overage and underage costs were equal it would make sense to always order enough to anticipate having the mean (50) on hand. This policy is not optimal however, when it costs the retailer more for a shortage than for excess. By trial and error, I found a base stock quantity of 300 units to minimize the cost to the ret...
