THREE ESSAYS ON VENDOR MANAGED INVENTORY IN SUPPLY CHAINS
The first essay considers a vendor V that manufactures a particular product at a unique location. That item is sold to a single retailer, the customer C. Three cases are treated in detail: Independent decision making (no agreement between the parties); VMI, whereby the supplier V initiates orders on behalf of C; and Central decision making (both Vendor and Customer are controlled by the same corporate entity).
Values of some cost parameters may vary between the three cases, and each case may cause a different actor to be responsible for particular expenses. Under a constant demand rate, optimal solutions are obtained analytically for the customer's order quantity, the vendor's production quantity, hence the parties' individual and total costs in the three cases. Inequalities are obtained to delineate those situations in which VMI is beneficial.
The problem setting in the second essay is the same with that of Essay 1, but the sourcing agreements investigated are now CI and C&VMI. In CI, as in the usual independent-sourcing approach, the customer has authority over the timing and quantity of replenishments. CI seems to favour the customer because, in addition, he pays for the goods only upon use. Under a C&VMI agreement, the vendor still owns the goods at the customer's premises, but at least can determine how much to store there.
The second essay thus contrasts the cases CI and C&VMI, and compares each of them to a no-agreement case. General conditions under which those cases create benefits for the vendor, the customer and the whole chain are determined.
Essay 3 investigates VMI and C&VMI separately for a vendor and multiple customers who face time-varying, but deterministic demand for a single product. In any of those agreements, the vendor seeks the best set of customers to achieve economies of scale. MIP models are developed to find that set of customers, and to determine the vendor's optimal production, transportation, and customer-replenishment quantities. The model for VMI is solved using a heuristic that produces two sub-models, and uses hierarchical solution approach for production, customer-replenishment and transportation decisions. C&VMI model is solved using Lagrangian relaxation. Various numerical examples are used to test the solution approaches used.
In the mean time, the customers can guarantee to be no worse off under VMI or C&VMI than the no-agreement case by setting the right levels of maximum inventory. A model to determine those levels and a solution algorithm are also proposed in Essay 3.
The first two essays can help a vendor or customer in a supply chain to determine the least costly sourcing option, which depends on the relative values of various cost parameters. A vendor with multiple customers can make use of the results in the third essay, which reveal the best possible economies of scale under VMI or C&VMI. Those customers can guarantee to be no worse of than traditional sourcing when they set the proposed levels of maximum inventory.
School:University of Waterloo
School Location:Canada - Ontario
Source Type:Master's Thesis
Keywords:management vendor managed inventory consignment supply chain mixed integer programming
Date of Publication:01/01/2006