A Hybrid Simulation–Optimization Framework for Ground–Aerial Vehicle Routing in Disaster Logistics Systems

Background: Natural disasters severely disrupt humanitarian logistics by damaging road networks, creating inaccessible demand locations, and introducing substantial uncertainty in demand, travel conditions, and vehicle availability. These conditions make conventional deterministic vehicle-routing approaches insufficient for reliable relief distribution. This study proposes a hybrid simulation–optimization framework for coordinated ground–aerial vehicle routing in disaster logistics systems. Methods: A Mixed-Integer Linear Programming (MILP) model is first used to determine heterogeneous truck routes, drone assignments, and launch/recovery decisions for accessible and inaccessible nodes. The resulting solutions are then evaluated in a hybrid Agent-Based and Discrete-Event Simulation environment under stochastic demand, travel-time variability, road-accessibility disruptions, and drone failures. Results: Computational experiments based on eight urban districts in Tehran show that the hybrid truck–drone system can maintain service coverage in damaged networks and reveal important operational vulnerabilities. The results also indicate that greater use of agile light-duty trucks can improve adaptability in disrupted urban networks. Conclusions: The framework links optimization with simulation-based robustness assessment, providing a scenario-based analytical approach for evaluating relief-routing plans under disrupted operating conditions.