Application of artificial intelligence technologies for cost optimization in maritime shipping

Abstract

The study’s relevance stems from the sharp deterioration of sea transportation conditions, the rise in fuel costs, the dynamics of navigation and meteorological parameters, and the need for prompt decision-making in an environment of high uncertainty. Traditional methods of technical control, route planning, and forecasting of operational indicators no longer provide the required accuracy, increasing the risks of delays, cost overruns, and technical failures. The purpose of the article is to scientifically substantiate approaches to the integration of artificial intelligence technologies in the management of costs of maritime transportation through the improvement of forecasting, the improvement of the accuracy of operational planning and the minimization of technical and navigational risks. To achieve the goal, systematic and analytical approaches were used, enabling consideration of artificial intelligence technologies in relation to economic, technical, and navigational processes in sea transportation. A comparative analysis was conducted to generalize the results of modern scientific research and practical approaches to cost optimisation in shipping. The methods of theoretical generalization, structural-logical analysis, and the interpretation of scientific sources were also used to form conceptual conclusions regarding the possibilities and limitations of introducing artificial intelligence into maritime logistics.

It has been established that artificial intelligence models provide a significant increase in the accuracy of forecasting fuel consumption, travel route times, and the probability of operational delays. It was found that the use of machine learning models enables the formation of adaptive routes and optimized speed profiles that respond to changes in weather and sea conditions in a near-real-time mode. It has been proven that integrating intelligent algorithms into technical operations contributes to the early identification of degradation processes, the reduction of accidents, and the reduction of unplanned repairs. Implementation challenges related to data quality, computing resource limitations, digital platform compatibility, and algorithmic error risks are outlined.

It is summarized that the complex implementation of artificial intelligence technologies creates a sustainable model for fleet cost management, improves navigational accuracy, and increases the overall operational reliability of shipping companies. The expediency of transitioning to forecast-oriented management of ship technical condition and integrating analytical models into all key elements of the operational cycle is substantiated. Development of algorithms resistant to the incompleteness and noise of marine telemetry, expansion of the applications of digital doubles, standardization of data in global logistics networks, and creation of interpreted models suitable for practical verification by crews and engineering teams are considered promising.