Vessel Trajectory
Vessel trajectory analysis focuses on understanding and predicting the movement patterns of ships, primarily using Automatic Identification System (AIS) data, to improve maritime safety and efficiency. Current research emphasizes developing advanced machine learning models, such as transformers, recurrent neural networks (RNNs, including LSTMs and GRUs), and convolutional neural networks (CNNs), often combined for enhanced performance, to predict vessel trajectories and cluster similar movement patterns. These advancements are crucial for applications ranging from anomaly detection and collision avoidance to optimizing vessel routing and managing the risk of invasive species spread via ballast water. Furthermore, incorporating additional data sources, like meteorological information and navigational context, is improving prediction accuracy and reliability.
Papers
Improved context-sensitive transformer model for inland vessel trajectory prediction
Kathrin Donandt, Karim Böttger, Dirk Söffker
Short-term Inland Vessel Trajectory Prediction with Encoder-Decoder Models
Kathrin Donandt, Karim Böttger, Dirk Söffker
Incorporating Navigation Context into Inland Vessel Trajectory Prediction: A Gaussian Mixture Model and Transformer Approach
Kathrin Donandt, Dirk Söffker