Comparing eye-tracking system effectiveness in field and driving simulator studies

Background: Several studies have been developed by combining the benefits of eye-tracking systems with driving simulators to simultaneously investigate driving behavior and the potential source of distraction. However, little effort has been spent in terms of eye-tracking validation in the driving simulator environment. Objective: The overall aim of this study is to validate an eye-tracking system within the context of a driving simulation environment by considering a specific urban context application. Methods: Both a field survey and a driving simulation experiment have been developed for a case study located in Rome, Italy. The selected real road sections and events have been reproduced on the driving simulator system and an eye-tracking has been used to record the eye movements both on board of a real vehicle and on the simulator. The eye movements of 14 participants in the field survey and 18 participants in the driving simulation tests, as well as their driving performances, have been collected while approaching an urban intersection and in relation to two specific road events: i) the presence of a speed limit sign and ii) the presence of a crossing pedestrian. Results: Eye tracker parameters and driving performances were compared between the real driving tests and driving simulator experiments in order to validate the eye-tracking system. It has been validated for both the events in terms of duration and distance of the eye fixation. Conclusion: The results demonstrate that the eye-tracking system stands as an effective tool for studies and applications in a virtual reality environment.