Ant Colony Optimization for Efficient Emergency Ambulance Routing in Urban Environments

Abstract

Efficient ambulance routing plays an important role in emergency medical services. However, solving the ambulance routing problem remains challenging. This study investigates the performance of ant colony optimization to solve the ambulance routing problem, aiming to improve the quality of route planning under constraints such as traffic, patient urgency, and ambulance capacity. To simulate realistic emergency scenarios, 27 benchmark instances from the classical vehicle routing problem were adapted to the ambulance routing context by mapping depots to ambulance stations, customer nodes to emergency sites, and incorporating patient urgency and ambulance capacity. The performance of ant colony optimization was compared with the genetic algorithm and particle swarm optimization. Each algorithm was independently applied to all instances, and route quality was evaluated based on best route cost, average cost, and standard deviations. The experimental results show that ant colony optimization consistently outperformed both genetic algorithm and particle swarm optimization across most instances. Specifically, ant colony optimization achieved shorter total route distances, which are measured as the cumulative route distance of ambulances required to serve the emergency sites. These improvements were accompanied by greater consistency in solution quality across multiple runs. These findings suggest that ant colony optimization is a robust and effective tool for ambulance routing optimization. This study contributes to the growing body of work on intelligent emergency logistics by demonstrating the practical advantages of ant colony optimization in critical decision-making. The findings are valuable for optimization researchers in enhancing ambulance routing efficiency.