Block assembly scheduling in shipbuilding involves complex multi-stage precedence relationships and strict due-date constraints, and its difficulty is further increased by processing-time variability. Deterministic scheduling approaches typically eval...
Block assembly scheduling in shipbuilding involves complex multi-stage precedence relationships and strict due-date constraints, and its difficulty is further increased by processing-time variability. Deterministic scheduling approaches typically evaluate schedules under fixed processing times, making it difficult to fully represent the impact of variability observed in practical operations. This study proposes a Simulation-based Adaptive Large Neighborhood Search (SIM-ALNS) framework. In each iteration, multiple candidate schedules are generated and evaluated through discrete-event simulation, and a relatively superior schedule is selected based on simulation-based performance comparison. Through this process, the effects of processing-time variability are directly incorporated into the search procedure. Computational experiments show that the proposed SIM-ALNS achieves a maximum additional improvement of 4.92%p and an average improvement of approximately 1.81%p compared with an existing simulation-based search strategy under identical conditions. These results indicate that comparing multiple candidate schedules through simulation can enhance scheduling performance under uncertainty. The proposed SIM-ALNS provides a practical scheduling approach for block assembly environments with uncertain processing times, enabling the selection of schedules with higher execution stability.