A new arXiv paper introduces TempTPI, a model that adapts the Informer architecture to forecast the future paths of ships over long time horizons. The approach learns from AIS tracking data, where conventional Transformers can struggle with long input sequences. The authors target improvements in navigational safety and logistics planning for the maritime sector.
A new arXiv paper introduces an autonomous GeoAI agent designed to plan maritime routes through the Arctic. The work responds to shifting sea-ice conditions that extend seasonal access but bring greater operational, environmental, and community risks. The agent aims to support eco-conscious navigation decisions as northern shipping corridors become more viable.
A new arXiv paper reports on a survey of maritime professionals regarding their perceptions of and trust in AI-supported decision-making tools and autonomous surface ships. The authors argue that how crews and operators view these systems will shape whether they can be integrated safely into real-world maritime operations. The work focuses on operator attitudes as a key factor in adoption rather than on technical performance alone.