<![CDATA[As the world’s largest archipelagic country, Indonesia relies on maritime logistics for energy distribution, particularly in the eastern region, which is vulnerable to extreme weather. This study validates the feasibility of The Smart Archipelago framework, an artificial intelligence (AI)-based decision support system designed to enhance the resilience of energy logistics. The case study focuses on a disruption event in the Ambon-Sorong shipping route in May 2023, analyzing three main modules: integration of oceanographic, meteorological, and operational data; predictive analytics through the comparison of four machine learning models (Logistic Regression, SVM, Random Forest, LightGBM); and a prescriptive module based on economic feasibility analysis. Results show that the tuned SVM achieved the best performance on the test set (F1-score 0.63), while tuned Logistic Regression demonstrated the highest stability in cross-validation. The gross cost-benefit ratio reached 429% for a single idealized avoidance scenario; however, after adjusting for prediction uncertainty based on the model’s precision (0.62), the model-adjusted ROI is approximately 291%, which remains economically favorable. Sensitivity analysis across conservative to optimistic operational assumptions confirmed the robustness of this economic justification. These results are supported by a heuristic recommendation system based on key variables such as wave height and current speed. These findings confirm the potential of The Smart Archipelago to be implemented on multi-year historical data as a step toward predictive and adaptive maritime logistics systems.]]>
Copyrights © 2026
