A Simulation-Optimization Framework for Capacity-Constrained Fleet Maintenance

Companion code for the paper "When Does Prediction Accuracy Pay Off? A Simulation-Optimization Framework for Capacity-Constrained Fleet Maintenance." One simulation-optimization engine optimizes condition-based maintenance policies for any fleet that competes for shared maintenance capacity, and measures the value of remaining-useful-life prediction accuracy. A fleet enters through a problem specification; the engine is unchanged across fleets. Includes the AH-64 rotorcraft and organic-industrial-base depot instances, the heterogeneous-island genetic-algorithm search, the generalized decision report (resource utilization, binding bottleneck, and the marginal value of capacity versus accuracy), and scripts to reproduce the paper's results.