Effective scheduling of ships remains essential for maritime operations optimization along with port congestion reduction and time reduction during port stays. In the paper, the topic of employing machine learning methodology to automate systems facilitating ship scheduling and the possibilities of enhancing them are addressed. By following the regression algorithms, a predictive model uses the operational characteristics extracted from AIS details and providing the ship types with waiting times and port entrance counts and the congestion measures to set the system for the prioritization of vessels. Reword the analysis by carrying out a feature engineering on raw data elements before calculating waiting time predictions as well as ship priority scores using RandomForestRegressor. Prediction system yields better accuracy in determining ship priority ranking that is validated using precision and recall and F1-score metrics. There was a proposed scheduling algorithm that would automatically schedule ship berthing by use of predicted priority scores, which were derived from the system. The practical benefits of the system are provided with visualization analysis and performance tests and real-time animation capabilities. By its intelligent system the port enjoys higher resource usage efficiency and operation speeds, which are applicable to different maritime conditions. Real-time scheduling modifications are possible by the system in such a way so that it can automatically adjust schedules upon arrival of new ships in the port. The system minimizes the wasteful periods and enhances the operational performance and impacts on the environment. The solution maintains scalability through its flexible system design which allows accurate deployment in various port scales and complexities. The findings suggest that machine learning technology contributes substantial benefits toward smarter decision-making within the port logistics sector so future systems can operate through data-based maritime infrastructure. Future research will prioritize the integration of adversarial training techniques as well as the implementation of diverse features with improved interpretation capabilities of the system.