Incentive, subsidy, penalty mechanisms and pooled, unpooled allocation of production capacity in service parts management systems

Atak, Erman
In this thesis, two systems are analyzed in order to gain insight to the following issues: (i) Effect of incentive, subsidy and penalty designs on decentralized system, (ii) effect of using production facility as pooled capacity (pooled system) and dedicated capacity (unpooled system) on capacity utilization and system profit. Regarding the first issue, three models are defined; decentralized model, centralized model and decentralized model with incentive, subsidy, penalty designs. In all models, there are two dealers and one item is under consideration and lateral transshipments are allowed. Dealers operate with four inventory level decision (strategies) that consists of base stock level, rationing level, transshipment request level and customer rejection level. Under the decentralized system, a dealer sets its operating strategy according to the strategy of the other dealer and maximizes its own infinite horizon discounted expected profit. In the centralized system, a central authority (say manufacturer) exists, which considers the system-wide infinite horizon discounted expected profit, and makes all decisions. Under decentralized system with incentive, subsidy, penalty designs, manufacturer tries different designs on decentralized system namely revenue sharing, holding cost subsidy, request rejection penalty, transportation cost subsidy and commission subsidy in order to v align decentralized system with centralized system. According to the results obtained, this alignment works best with nearly 40% revenue sharing percentage, low rejection penalty, high transportation cost subsidy under low transportation cost and commission subsidy under very low or very high commissions. Holding cost subsidy, on the other hand, is not a good strategy since it declines decentralized system profit. Considering the second issue, two systems are examined; pooled system and unpooled system. Both systems are centrally managed. In the pooled system, all capacity is dynamically allocated to either dealer considering maximization of system profit. In the unpooled system, capacity is shared among dealers and dealers are always allocated same percentage of the capacity. Infinite horizon average expected profit is maximized in both systems. The dealer having lower holding cost is allocated higher capacity in both pooled and unpooled system; however, exceptions exist in the unpooled system under low arrival rate. Highrevenue dealer is always allocated higher capacity in both pooled and unpooled system. Arrival rate affects both systems such that total capacity utilization increases with increasing arrival rate. From the profit point of view, pooled system has great advantage under low demand rate in general.