ROBUST OPTIMIZATION APPLIED TO THE DAILY SCHEDULING OF ELECTIVE SURGERIES WITH RESOURCE PARALLELISM

This paper proposes an exact robust optimization method for elective surgery scheduling in hospitals, considering uncertainty in procedure durations. The formulation adopts the knapsack uncertainty set and is solved through a cutting-plane algorithm for managing resource parallelism, together with a scenario generation algorithm based on dynamic programming to handle delays in operating room closing times caused by deviations in surgery durations. Computational experiments were conducted using real-world data from two hospitals, covering 331 operating days. The results indicate that the robust solutions maintain quality similar to deterministic ones regarding operating room utilization and allocated surgeries, while reducing overlaps within the same operating room and excessive resource usage, as verified through Monte Carlo simulations.