A data-driven hybrid framework for analyzing vessel detentions in Australia

Port State Control (PSC) plays a critical role in safeguarding maritime safety and environmental protection by identifying and detaining substandard vessels. Given the complexity of detention decisions and the diversity of vessel profiles, there is a growing need for interpretable, data-driven approaches that reflect regional enforcement contexts. This study proposes a hybrid analytical framework that integrates Entropy-Weighted Grey Relational Analysis (EW-GRA) and Association Rule Mining (ARM) to evaluate vessel detentions in Australia. EW-GRA ranks deficiency categories by their influence on detentions, while ARM reveals frequent co-occurrence patterns, uncovering relational structures among deficiencies. Using ten years of detention records from the Australian Maritime Safety Authority (AMSA), the analysis identifies International Safety Management (IntSafeMan), Emergency Systems (EmSys), and Water/Weathertight Conditions (WatWeath) as the most influential deficiencies. Stratified analysis by vessel type and age reveals increased risks among bulk carriers and older vessels, supporting targeted inspection strategies. The ARM results show that IntSafeMan-related deficiencies frequently co-occur with crew welfare and operational deficiencies, indicating compound risk factors. These findings align with AMSA’s compliance priorities, particularly in relations to planned maintenance, emergency preparedness, and working conditions under the IntSafeMan Code and the Maritime Labour Convention (MLC). The proposed framework enhances interpretability and visualisation, offering practical utility for inspection planning and regulatory oversight, thereby contributing to more effective, risk-informed PSC enforcement in the Australian maritime domain.