Visual Attention Analysis for Critical Operations in Maritime Collision Avoidance

The research on how to provide effective onboard decision support for surface ships, especially in critical operational scenes such as collision avoidance, has been triggered as a goal in recent years. From the authors’ perspective, to achieve this goal, it is crucial to comprehensively understand the operational logic and mechanism of the human navigators. In this paper, we use wearable eye tracker glasses to collect visual attention data from the navigators. The scene is established as a collision avoidance task in a strait water channel on a maritime simulator. By using the concept of critical operations, the whole sailing is divided into cruising and maneuvering time windows. The visual attention is analyzed in terms of the transition time/frequency and area of interest. It is the first time to exclusively analyze the navigators’ visual attention in collision-avoidance tasks. The paper suggests a way to potentially predict the human-dominant onboard operations, which builds the basis for developing a better decision support system.