Trajectory Clustering Model Using Density-Based Spatial Clustering of Applications with Noise (DBSCAN) in Maritime Routes

Abstract

This study develops a trajectory clustering model using the Density-Based Spatial Clustering of Applications with Noise (DBSCAN) method to enhance maritime surveillance in Indonesia. The objective is to group vessel trajectories based on movement patterns to detect anomalies and illegal activities such as transshipment. Data were collected from the Automatic Identification System (AIS), including vessel position, speed, and heading within Indonesian waters over a specific period. The research stages included AIS data collection, preprocessing to remove noise, applying DBSCAN for trajectory clustering, and evaluating model performance. The results demonstrate that the model effectively identifies vessel trajectory clusters with high accuracy, despite challenges posed by large and heterogeneous data. The novelty lies in applying DBSCAN to detect vessel trajectory anomalies in the context of Indonesian waters, a method rarely explored before. The impact of this research is the creation of a data-driven analytical tool that can enhance maritime security, detect illegal activities, and support decision-making by relevant authorities. This model is expected to provide a practical and effective solution for maritime traffic management in Indonesia.