EnvShip-Bench: A New Benchmark for Vessel Trajectory Prediction

Vessel trajectory prediction is crucial for intelligent shipping, maritime surveillance, and navigation safety. However, existing public maritime AIS (Automatic Identification System) resources often suffer from inconsistent forecasting protocols, variable data quality, and a lack of benchmark-ready contextual annotations, which impede fair comparisons and the development of context-aware models. To address these limitations, EnvShip-Bench has been introduced as a unified benchmark for short-term vessel trajectory prediction. It is constructed from extensive raw AIS data from the Danish Maritime Authority (DMA) and NOAA, processed through a common pipeline. EnvShip-Bench establishes a standardized forecasting protocol, including specific observation and prediction windows and sampling rates in vessel-centric local metric coordinates. Beyond its large-scale core, it offers a compact, quality-first subset for efficient experimentation and includes synchronized environmental and nearby-vessel context extensions. This comprehensive benchmark supports trajectory-only, environment-aware, and interaction-aware forecasting within a unified evaluation framework, providing a robust foundation for maritime trajectory research.