Automatic Identification System (AIS)-Based Ship Trajectory Modellinfor Indonesian Sea Transportation Monitoring

The aim of this research is to build a ship trajectory model and display ship predictions for the next period, using 10 minutes of preliminary data on the trajectory of ships sailing in the Java Sea, and then predict the trajectory of the next ship in the next 20 minutes and the next 40 minutes. This study uses data mining steps in AIS data mining which is then processed using the LSTM learning algorithm with sequence prediction. By using some static data on AIS, that is SOG, Latitude, Longitude and MMSI from the ship during November 2018 in the Java Sea, Indonesia. Results of the study indicated that the RMSE resulted in a value of 0.13. This research is original in its approach to utilizing AIS data with the LSTM algorithm for ship trajectory prediction in the Java Sea. The integration of sequence prediction with specific ship data parameters provides a novel methodology for maritime navigation and safety management. This study contributes valuable insights into the predictive modelling of ship movements, which can enhance the efficiency and safety of maritime operations. The impact of this research is significant for maritime navigation and safety. By accurately predicting ship trajectories, it can help in avoiding potential collisions and improving route planning. This can lead to better fuel efficiency and reduced operational costs for shipping companies. Additionally, the methodology can be adapted to other regions and types of vessels, providing a broader application for global maritime safety and logistics.