For a system to stay operational, maintenance of its components is required and to maximize the operational readiness of a system, preventive maintenance planning is essential. There are two stakeholders—a system operator and a maintenance workshop—and a contract regulating their joint activities. Each contract leads to a bi-objective optimization problem. Components that require maintenance are taken out from operating systems and sent to the maintenance workshop, which should perform all maintenance activities on time in order to satisfy the contract. Upon being maintained, the components are sent back and available to be used in the operating systems. The ability of the workshop to fulfill the contract is highly dependent on its capacity. Our modeling of this system–of–systems includes problem structuring of the planning of preventive maintenance for the operating systems, the maintenance workshop scheduling as well as the stocks of damaged and repaired components.
The mixed-integer linear optimization (MILP) model we present is partly based on an optimization model of a preventive maintenance scheduling problem with interval costs over a finite and discretized time horizon, which we generalize and extend with a non-preemptive flow of components through the workshop and the stocks of (damaged and repaired) components. Our results measure the effect of the workshop capacity on the level of component availability as well as on the utilization rate. We also analyze the maximum possible reduction of the costs of two stakeholders for increased levels of the workshop capacity. The resulting modeling can be utilized as a decision aiding tool, and help improve the decision making processes for the stakeholders.