Operation optimization module for Terminal Operating system

One of the most significant challenges in managing a marine container terminal is ensuring optimal utilization of warehouse space while maintaining high throughput of containers processed at different cargo fronts. The usual management process is as follows: a terminal specialist makes decisions on where to place each container based on a complex system of rules influenced by over 50 different factors. After making a decision, the specialist records the decision in the terminal operating system, providing further instructions for terminal equipment operators. Analysis of data from the Terminal Operating System has shown that up to 40% of movements are unproductive when a person is responsible for distributing containers in the work area. As an example of such movements - if it is necessary to move a container from its original warehouse location to the railway or inspection area, other containers may need to be moved, blocking access. The module for terminal operation optimization driven by AI-models was designed to predict decisions about placing a container in a terminal work area in order to reduce the number of unnecessary movements and, thereby, increase cargo turnover at the terminal. The project was initiated by Commercial Port of Vladivostok, the largest stevedoring company in Russia, and is being developed and implemented jointly by the IT department of the FESCO Transport Group (the owner of the port), Sber, and the Far Eastern AI Center of Far Eastern Federal University.