Answers are grounded in this paper and cite their sources.
Winter navigation in the Baltic Sea relies on complex icebreaker assistance in ice-covered waters. However, the classification of various assistance operations from ship behavior remains limited. This paper proposes a method for detecting and classifying icebreaker assistance modes, focusing on escort , convoy , and cutting loose operations. The method initially identifies spatio-temporal interactions between the icebreaker and the assisted vessel to filter candidate assistance events. It then introduces a following concept and establishes a mathematical approach to capture shared characteristics of escort and convoy behaviors, followed by a timeline segmentation algorithm for mode classification. Furthermore, among the potential assistance candidates, cutting loose events are detected by modeling the interaction features between the icebreaker and the beset vessel. The work is applied to Automatic Identification System data from February 2024 in the Baltic Sea and validated against icebreaker operational logs from the IBNet system. Results demonstrate high precision and recall for escort detection, and reasonable performance for convoy detection across three Finnish icebreakers. Cutting loose detections are evaluated together with the corresponding assistance circumstances and associated sea-ice conditions, indicating consistency with operational practice. The proposed approach enables interpretable identification of icebreaker assistance modes from AIS data, supporting future operational assessment and risk analysis in winter navigation.
No takes yet. Share an insight, caveat, or question.
Xiang et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: