Machine learning techniques in maritime environmental sustainability: A comprehensive review of the state of the art

Abstract

With the development of the global maritime industry and the intensification of environmental challenges, machine learning technology has emerged as an innovative solution to the environmental sustainability issues in the maritime industry. This study comprehensively reviews the applications of machine learning in the field, with a focus on two key sectors: ships and ports. It delves into important topics such as ship energy consumption prediction, ship emission prediction, ship emission monitoring, port emission prediction, port air quality prediction, and so on. This review provides an in-depth analysis of the current research status, challenges, and future directions. The review finds that in terms of applications, research related to ships is relatively mature, while research related to ports is limited. In terms of algorithms, Random Forest, Artificial Neural Networks, and Gradient Boosting Machines are the most widely used. As the industry continues to grow, future research may focus on the integration of multi-source heterogeneous data, improvement of the interpretability and generalizability of machine learning models, and utilization of more advanced models and algorithms, which are expected to improve the development in the field and contribute to maritime environmental sustainability.