Factor analysis of icebreaker assistance operation for ice-going ships in the Baltic Sea

Merchant vessels navigating the Baltic Sea in winter often require assistance from icebreakers to create safe pathways and improve navigational efficiency. Given icebreaker resources are limited, assistance decision is important. The requirement for assistance depends on multiple factors, including ice conditions, weather, and ship characteristics. In this paper, we explore how data-driven techniques can enhance the current understanding of factors influencing the decision-making of icebreaker assistance. Firstly, the paper identifies multiple factors from previous winter navigation operations research. Then different data sources containing traffic data, environmental conditions, and ship characteristics are explored to find data about the identified factors. Finally, an integrated database containing these factors is established. Using a multi-step clustering method, data points in the database are classified as either assistance or independent navigation. Preliminary statistical analysis of the factors is performed to understand how they vary between independent navigation and assistance cases. Results show that weather and ship factors do not significantly vary compared to ice factors. Among the ice factors, ridge ice thickness and level ice concentration vary the most between independent navigation and assistance. These findings are aligned well with empirical knowledge and previous studies. The database and the empirical findings in this paper can provide insights for quantifying factor effects on the decision-making of icebreaker assistance and support the intelligent decision-support system for winter navigation.