Machine Learning Of Port Terminal: Predicting Vessel Eta For Seamless Port Scheduling Operation: A Case Study Of Rotterdam Berth

Abstract

This thesis focuses on the issue of optimizing vessel scheduling at Port Rotterdam, a significant hub for global marine logistics. Due to the increasing volume of traffic and complexity of operations, effective management solutions are needed to ensure efficient resource usage and prevent congestion. This study forecasts the vessel's Estimated Time of Arrival (ETA) using machine learning, specifically neural networks, to enhance berth allocation. Through the combination of simulation-based optimization and predictive analytics, the study provides a robust scheduling system. This framework aims to improve the operational sustainability and efficiency by decreasing waiting times, berth location fluctuations, and associated costs. The findings indicate that because the Priority (PRI) rule balances cost and service effectiveness, it is a feasible substitute for actual implementation.