WAY: Estimation of Vessel Destination in Worldwide AIS Trajectory
Jin Sob Kim, Hyun Joon Park, Wooseok Shin, Dongil Park, Sung Won Han
2025-12-18
Summary
This paper focuses on predicting where ships are going, using the data automatically broadcast by ships called AIS. It tackles the challenges of unreliable and incomplete AIS data to make more accurate long-term predictions.
What's the problem?
Currently, using AIS data to figure out a ship's destination is difficult because the data isn't always perfect – signals can be missed or inaccurate, and ships don't report their location constantly. Existing methods struggle with these gaps and biases in the data, especially when trying to predict destinations far into the future, like days or weeks ahead.
What's the solution?
The researchers developed a new deep learning model called WAY. They first reorganized the ship's travel path into a structured sequence using a grid system to reduce errors caused by inconsistent data. WAY then uses a special architecture with layers that analyze the ship’s movement and other characteristics to predict its destination. A key part of their solution is a new training technique called Gradient Dropout, which helps the model learn more effectively even when dealing with varying lengths of travel data. Essentially, it prevents the model from getting overly influenced by shorter trips.
Why it matters?
Accurately predicting ship destinations has a lot of real-world benefits. It can improve maritime safety, help with efficient port operations, and even assist in tracking illegal activities at sea. This research provides a more reliable way to forecast ship movements, which could lead to better decision-making in various maritime applications, and even estimate arrival times more accurately.
Abstract
The Automatic Identification System (AIS) enables data-driven maritime surveillance but suffers from reliability issues and irregular intervals. We address vessel destination estimation using global-scope AIS data by proposing a differentiated approach that recasts long port-to-port trajectories as a nested sequence structure. Using spatial grids, this method mitigates spatio-temporal bias while preserving detailed resolution. We introduce a novel deep learning architecture, WAY, designed to process these reformulated trajectories for long-term destination estimation days to weeks in advance. WAY comprises a trajectory representation layer and Channel-Aggregative Sequential Processing (CASP) blocks. The representation layer generates multi-channel vector sequences from kinematic and non-kinematic features. CASP blocks utilize multi-headed channel- and self-attention for aggregation and sequential information delivery. Additionally, we propose a task-specialized Gradient Dropout (GD) technique to enable many-to-many training on single labels, preventing biased feedback surges by stochastically blocking gradient flow based on sample length. Experiments on 5-year AIS data demonstrate WAY's superiority over conventional spatial grid-based approaches regardless of trajectory progression. Results further confirm that adopting GD leads to performance gains. Finally, we explore WAY's potential for real-world application through multitask learning for ETA estimation.