Multiple can-order level for can-order policy optimization: A case study

Inventory is the products which are stocked for to fulfill the demand in the future. Here the problem may occurrs if the demand of the item is intermittent, since the inventory policy is difficult to be determined due to the uncertain demand of the item that has high error rate in the prediction. This problem is occurred in the spare part needs of Company X. Therefore, a joint replenishment method, can-order policy that has multiple can-order level (si, cij, Sij), can be implemented to coordinate the order of some different items. A non-linear integer programming model, which can be categorized as NP-hard problem, is used to handle the problem. The contribution of this paper is on the use of the approach on the real case which never done before. A metaheuristic approach, simulated annealing, is applied to find a satisfied solution with a fast computation time. Global criterion method is also used to reach the multi-objective function, which are to minimize total inventory cost and number of carriers shipped by supplier. According to experiment done with 3 suppliers and 157 items, the result obtained is better than the existing condition in Company X with number of carrier reduction of 2.6% and cost saving of 14.43% or equal to $ 81,570.79.