DOI
https://doi.org/10.3390/jmse8030201

Summary
The freight transportation is analyzed in this article. The authors state that it is one of the main problems in transport organization because of many factors that must be considered. Firstly, the state-of-the-art research is presented with the summary of main models used for cargo transportation, that is, models for transnational cargo transportation, cargo transportation with a dedicated initial delivery port, freight transport with dedicated initial delivery ports and terminal port of cargo distribution, cargo transportation on a circular chain of ports. Secondly, the mathematical formulations for each model are provided. The system can automatically optimize routes based on the selected model and therefore reduce human impact and the workload required to plan the routes. It is concluded that similar approaches can be transferred to other means of transportations.

Relevance for Complex Systems Knowledge
The maritime transportations system should be modelled as a complex system due to the complexity of route maps, diversity of ships and various requirements and constraints, dynamics and internal network. Usually, a number of approximations and assumptions are applied in order to construct a model for route planning. However, these assumptions often do not reflect the real network as it is too complex to construct deterministic model with optimized resources. The machine learning methods enable to create a reasonable model for freight transportation under the given initial data (initial and / or terminal port, constraints) automatically, optimize resources and minimize human impact.

Point of Strength
The strength of the article is a provided machine learning model of the freight transportation in a large complex maritime transportation system. It is a useful example of artificial intelligence application in practice.