Optimizing Indonesian Pioneering Maritime Network through Port Clustering

Abstract Indonesia’s unique geography, characterized by scattered islands, presents challenges for maritime logistics, particularly in remote and border areas. To address these challenges, this study proposes a K-Medoids clustering model to optimize the pioneering maritime network, which is crucial for logistics distribution and population mobility in Eastern Indonesia. Unlike previous deterministic models, this approach groups ports based on their spatial characteristics, aiming to enhance network efficiency and resilience. By analyzing the coordinates of 512 ports, the model identified 10 optimal clusters, validated through gap statistics, silhouette analysis, and the Davies-Bouldin Index. The results indicate significant improvements in route efficiency, reduced voyage durations, and lower fuel consumption, leading to cost savings and environmental benefits. The optimized clusters could also enhance port infrastructure utilization and connectivity, supporting regional economic development. However, there is a trade-off with potential reductions in ship call frequency at individual ports. Overall, the K-Medoids model provides a robust framework for optimizing maritime logistics, balancing efficiency and service needs, while promoting regional growth.