Prognostics-Based Two-Operator Competition in Proactive Replacement and Service Parts Procurement

Effective prognostics and timely maintenance of degrading components can improve the availability and economic efficiency of an engineering system. However, possible shortage of required service parts usually makes near-zero downtime difficult to achieve. To coordinate service parts availability with scheduled maintenance, it is necessary for the operator to decide when to order the service parts and how to compete with other operators in parts procurement. In this article, we consider a situation where two operators are to make prognostics-based replacement decisions and strategically procure the needed service parts. When a competition occurs, each of the operators has a bounded continuum of strategies. A one-shot sequential game (Stackelberg game) is formulated and a sequential, constrained maximin experimental design approach is proposed to facilitate searching for the equilibrium solution. This approach is quite useful in handling cases where the follower's best response to the leader's strategy, both chosen from continuums of strategy sets, is difficult to obtain analytically. Numerical studies on wind turbine operation are provided to demonstrate the use of the sequential decision-making method in solving such complex, yet realistic maintenance and service parts logistics problems.