Expert versus Novice mariners, a study using eye-tracking in an unfamiliar ship handling simulation scenario

Abstract

Eye-tracking technology is an area that has a lot of potential in discovering differences and behaviors of people in many industries. One of these is the maritime industry. This study will try to find differences in gaze activity between 4th-year Master Mariner students and maritime pilots. To justify the incorporation of new technology in, for example, education, it is first important to see if there are differences to be found that can help develop new ways of transferring knowledge. The work presented was done at Chalmers University of Technology, where ship handling simulations were executed by two sample groups that were equipped with eye-tracking glasses to find differences. Additionally, a questionnaire was filled out by the participants to see the differences between what they perceived they focused on versus what they actually focused on. Tracks from the simulation runs were saved and presented to see a difference in the approach of navigation between the groups. The result of the study shows that experts divided their gaze activity more between instruments and visual observations than novices. Novices were mostly focusing on their conning screens and straight ahead of the ship. When comparing the tracks, a large difference can be seen in the approaches. Results found show that multiple differences between the groups' gaze activity could be interesting to explore further such as the use of the navigational equipment. This can also be seen in the questionnaire responses, as both groups had similar answers for what they perceived to have focused on but chose to answer with different wording, suggesting they may have used the equipment for different purposes. Tracks show a distinct difference in approach execution and experts were a lot more consistent in how they performed each run. Although the results show us big differences between the groups, delimitations of this study mean that these results cannot be generalized as the number of participants was low and only included specific demographics.