Developing an AI-enhanced maritime threat detection model: Predictive security framework for illegal fishing and piracy

Abstract

Background: The North Natuna Sea, as a strategic area of Indonesia, faces increasingly complex maritime security threats, particularly illegal fishing activities and piracy that threaten economic sovereignty and the stability of aquatic ecosystems. This research develops a maritime threat detection model based on Artificial Intelligence that is capable of predicting and preventing illegal activities through a preventive and proactive approach. Methods: Using a qualitative research method based on documentary studies, this research analyzes patterns of maritime threats, regional geopolitical dynamics, and the limitations of conventional surveillance systems that have historically relied on a reactive approach. The developed predictive model integrates three layers of technology: multi-source data integration (Sentinel-1 satellites, VIIRS, AIS, oceanographic data), predictive analytics utilizing Long Short-Term Memory (LSTM) and Random Forest algorithms, as well as an automated operational response system. Findings: Data analysis indicates that illegal fishing incurs economic losses of USD 25 billion annually on a national scale, with 112 vessels confiscated in the first half of 2024. Meanwhile, piracy incidents have increased from 10 incidents in 2022 to 18 incidents in 2023, affecting 126 crew members in 2024. The research results indicate that the predictive AI model is capable of increasing the detection rate of foreign vessels from 40% to 85% and reducing response time from 8 hours to 45 minutes. This system generates an updated maritime threat heatmap every 15 minutes, enabling the optimization of patrols and saving fuel consumption of up to 30%. Conclusion: This research contributes to the transformation of Indonesia's maritime security paradigm from a reactive approach to a predictive-preventive one, supporting the enforcement of economic sovereignty in Indonesia's Exclusive Economic Zone (EEZ), and providing high-tech solutions to address non-traditional security challenges in the contemporary era. This model can be adapted as a best practice for other island countries facing similar maritime threats, while also reinforcing collective maritime security mechanisms within the framework of regional and global cooperation. Novelty/Originality of this article: The preventive defense framework specifically developed for the conditions adopts a proactive approach that takes into account the unique geographical characteristics, regional maritime traffic patterns, and the geopolitical dynamics of the area.