This thesis deals with developing and solving tactical inventory planning models for a supply chain that will enable the individual companies to determine their ordering policies efficiently under different conditions. The supply chain is modeled as a single warehouse supplying a product to several retailers which in turn satisfies the end consumer demand. The supply chain operates under a decentralized control, i.e., each location is managed by an independent decision maker (DM).

In the first model, using the conventional single cost objective framework we propose a new coordination scheme that enables the warehouse to better manage its inventory and at the same time meet the retailers' demands without deviating too much from their requirements. To avoid the problem of estimating marginal cost information and to incorporate the DM's preference information, a more realistic multiple criteria model is then developed. To account for discounts in shipping, actual freight rate functions are used to model transportation costs between the stages. The conflicting criteria considered are: (1) capital invested in Inventory; (2) annual number of orders; (3) annual transportation costs. While the first two models deal with deterministic demand and constant lead time, the third model deals with stochastic demand and random lead time. In addition to the above three criteria, fill rate is used as a fourth criterion to measure customer satisfaction. The multiple criteria models are solved to generate several efficient solutions. The value path method, a visual tool, is used to display tradeoffs associated with the efficient solutions to the DM of each location in the supply chain.

The models are tested with real world data obtained from a Fortune 500 consumer products company. Additional problems faced while extending the theoretical models to the real world data are addressed and solved. The decision making process is simulated by using an executive from the company to be the DM for the warehouse. The preference information is obtained using standard multi-criteria techniques to generate the set of efficient solutions. The DM adjudged the multi-criteria methodology to be a more effective decision making tool since he had to evaluate tradeoffs and use his judgment to choose the most preferred solution.