A base period inventory policy for decentralized by aij20926


									        A base period inventory policy for decentralized
                         supply chains

                       Atul Rangarajan a, A. Ravi Ravindran a,
                  aDept. of Industrial & Manufacturing Engineering,
                             Pennsylvania State University
               310 Leonhard Building, University Park, PA 16802, U.S.A.

No part of this paper may be quoted or reproduced without the specific
permission of the authors.

This paper introduces an inventory policy called the Base Period policy for a
decentralized supply chain with a single warehouse and multiple retailers. The policy
requires that the retailers order once every base period or at integer multiples of the base
period. Such a policy is more general and flexible than the Power-of-two policy for the
retailers. A line search algorithm is used to determine the optimal order policies for the
retailers. We analyze the performance of the policy and demonstrate the key role played
by the choice of the base period length. We derive an upper bound on the error induced
by following the proposed policy and demonstrate that a prudent choice for the base
period length guarantees a solution within 94% of the optimal solution for the retailer
problems. We prove that the proposed policy provides a better solution to the retailers
compared to the power-of-two policy solution. The resulting warehouse problem is a
non-stationary, non-nested, infinite horizon problem. We present a methodology that uses
the base period policy structure to convert this into a finite-horizon problem which is then
formulated and solved optimally as a Wagner-Whitin model. A numerical example is
used to illustrate the model and the solution procedures.

To top