Review: Ship collision risk assessment using machine learning methods based on AIS data

The comparison table provides an overview of research dealing with ship collision risk assessment using various machine learning methods based on AIS data. A ship collision occurs when two or more vessels collide with each other, either while underway or at anchor. Ship collisions can result in significant damage to the vessels involved, as well as injuries or fatalities to crew members and passengers. They can also cause environmental damage, such as oil spills or other hazardous material releases, and disrupt shipping traffic in the affected area. Ship collision risk assessment is a process of evaluating the likelihood of a ship collision occurring in a particular area or waterway. It involves analyzing various factors that can contribute to a collision, such as vessel traffic density, navigational hazards, weather conditions, and human factors. The goal of ship collision risk assessment is to identify potential collision risks and develop strategies to mitigate them, such as implementing navigational aids, traffic separation schemes, or speed restrictions. Ship domain is a concept that plays an essential role in navigational safety of vessels. It refers to the surrounding effective waters that a navigator of a ship wants to keep clear of other ships or fixed objects. The classical definition of ship domain categorized the waters around a vessel into two zones: dangerous (inside ship domain) and safe (outside ship domain). The overlaps of ship domains indicate higher likelihood of ship collisions.