Efficient and timely delivery of goods is considered the most critical operations for a port, and their optimization is crucial for the success of businesses and economies. In the last years, The machine learning algorithms have demonstrated that are advantageous in improving port operations, and this by the prediction and optimization of various aspects of the supply chain. In this paper, we present a case study of the implementation of machine learning algorithms for the optimization of hydrocarbon operations at Tangier Med Port. We selected four popular machine learning algorithms and tested their effectiveness in predicting cargo volumes and optimizing hydrocarbon liquid storage. Our results showed that the random forest algorithm outperformed the other algorithms with an accuracy of over 90% in predicting cargo volumes and a significant improvement in hydrocarbon liquid storage efficiency. The implementation of the algorithm in the production environment conduced to a considerable reduction in turnaround times, productivity improvement, then it has increased customer satisfaction. Our study is to demonstrate the advantages of the machine learning algorithms to improve the port operations and to provide the valuable insights into their implementation in real time scenarios. As a future work, we can study on exploring the use of other machine learning algorithms and their integration with other technologies such as the use of the Internet of Things (IoT) to promise further port operation optimization.

smart management, maritime transport, hydrocarbons, artificial intelligence, prediction, Tangier Med Port.