Abstract
Aviation maintenance scheduling presents complex challenges due to dynamic task arrivals, stochastic service times, and the need to balance competing objectives. To address this, we introduce a novel framework that integrates these factors into a real-time, multi-objective optimization model. Our approach combines mathematical modeling with advanced meta-heuristic algorithms, supported by new theoretical performance guarantees. We evaluate nine algorithms across 810 experimental configurations, demonstrating that our proposed methods achieve statistically significant improvements over baseline scheduling approaches. Among single-objective metrics, Adaptive Tabu Search (ATS) achieves the lowest cost at $13,072 ± $4544, while multi-objective methods provide diverse Pareto fronts with a mean hypervolume of 0.0268 (normalized scale [0, 1]), dominating significantly more of the objective space than comparative methods. The framework demonstrates the potential for significant operational cost reductions and provides a robust theoretical foundation for developing next-generation maintenance scheduling systems.