Development of Prediction Methodology for the Enhancement of Maritime Pilot Occupational Safety

Abstract

Maritime transportation is a highly complex and hazardous industry where accidents can occur, even with stringent precautions. One particular concern is the occupational safety of maritime pilots during their transfers to and from vessels. These transfers typically involve using pilot boats and passing critical stages such as car transfers, walking on piers, pilot boat transfers, pilot ladder transfers, and climbing the ship's freeboard to reach the bridge. Each stage presents unique risks, including environmental factors, vessel conditions, and adherence to safety protocols. This study aims to enhance occupational safety for maritime pilots during their transfers to and from vessels through comprehensive analysis, risk assessment, and the development of preventive strategies. Chapter 2 explores the maritime pilot transfer operation, focusing on the involved parties, procedures, and international regulations. It highlights the roles of maritime pilots, pilot boat crews, ship owners, ship crews, and port authorities. The chapter discusses the stages of pilot transfer by boat and emphasizes compliance with safety protocols. It also addresses pilot ladders' construction and rigging requirements and associated equipment. The chapter underscores the importance of coordination and regulation adherence for safe and efficient pilot transfers. To gain a deeper understanding of the factors contributing to occupational accidents during maritime pilot transfers, Chapter 3 focuses on collecting and analyzing accident data records. Through a comprehensive peer review process, the study assesses the surrounding conditions of each accident and identifies the primary factors that contribute to their occurrence. A descriptive statistical analysis lists the contributing factors, possible consequences, and their frequencies. Researchers conducted a chi-square test using SPSS software to explore the correlation between surrounding conditions and the consequences of maritime pilot occupational accidents. The results of the statistical analysis and chi-square test indicate that there were no significant outcomes or conclusive findings. As a result, researchers propose an alternative methodology to develop a predictive risk model in Chapter 4. The paper utilizes historical accident records as a conventional approach to risk assessment, aiming to predict maritime pilot occupational accidents by developing a model using the RF (random forest) method in RStudio software. Utilizing a comprehensive dataset comprising 500 accident reports provided by maritime pilot organizations, the model's effectiveness is validated to assess its predictive risk capabilities. In Chapter 5, to recognize the need to address the specific risks associated with maritime pilot transfers, the paper proposes a detailed risk analysis using the FTA-ETA (Fault Tree Analysis-Event Tree Analysis) method within a fuzzy logic environment. The primary objective of this analysis is to identify and quantify the root causes of accidents, assigning importance levels and probabilities of occurrence to each factor by creating a Fault Tree (FT) diagram. The Event Tree (ET) analysis also considers accidents' potential outcomes and associated probabilities. By doing so, the paper aims to raise safety awareness among maritime pilots, pilot organizations, and ship crews involved in the transfer process. This study enhances safety in maritime pilot transfers by identifying and mitigating the root causes of accidents and understanding the potential consequences. By applying statistical analysis, predictive modeling, and the FTA-ETA method in conjunction with fuzzy logic, the study provides valuable insights for risk management and the development of effective prevention strategies. This comprehensive approach enables stakeholders to take proactive measures to minimize risks and ensure the safety of all parties involved in maritime pilot transfers. Chapter 6 presents a case study on analyzing maritime traffic data to assess the risk of pilot accidents in Busan Port, South Korea. Busan Port is a major port with high seaborne trade, resulting in significant pilotage work. The developed model is implemented using real-time AIS data collected during a seven-day maritime traffic survey in September 2019. The study focuses on the pilot boarding and disembarkation locations in Busan North Port, aiming to identify high-traffic areas and potential risk factors for pilot accidents. The analysis includes integrating accident data from the Korea Maritime Pilot Association (KMPA) into the random forest model. The results show that the developed model accurately predicts pilot accidents with a high level of accuracy. The study highlights the importance of considering the ‘caused by’ factor in accident analysis, as it has a more significant impact than environmental factors. The findings suggest that analyzing seven days of maritime traffic data can provide insights equivalent to observing traffic for a year, enabling port authorities to quantify the risk level and make informed policy decisions to enhance safety measures. By utilizing the developed model and analyzing traffic data, port authorities can allocate resources effectively and prioritize response efforts in areas with a higher risk of accidents, ultimately ensuring a safer environment for maritime pilots.