Handbook of Driving Simulation for Engineering, Medicine, and Psychology

Eye Behaviors: How Driving Simulators can Expand their Role in Science and Engineering

Authors
Donald L. Fisher, University of Massachusetts
Alexander Pollatsek, University of Massachusetts
William J. Horrey, Liberty Mutual Research Institute for Safety

Abstract
The Problem. It is almost a truism that what we do not see can threaten us as drivers. It is also becoming more and more apparent that even what we do see can threaten us if we are otherwise engaged. In either case, information about eye movements is essential in order to understand how differences in people (age, experience, disease), exposure to drugs and alcohol, the environment inside the vehicle (in-vehicle devices, texting, cellular phones), and the environment outside the vehicle (traffic, weather, night time, traffic control devices) impact driving. Much is known about what differences exist in the spatial and temporal characteristics of the eye behaviors among different groups of individuals. However, less is known about why these differences exist across the different groups of individuals. Nor is much known about how the spatial and temporal characteristics of eye behaviors change as a function of different environments inside and outside the cabin of the automobile, or for that matter why they change. Role of Driving Simulators. Eye movements are gathered in the field infrequently at best, even with today’s relatively non-intrusive and inexpensive eye trackers, often because one cannot safely stage the sorts of events about which eye movements are so informative. As an alternative to gathering eye movements in the field, driving simulators have three distinct advantages: drivers can be deliberately placed in dangerous scenarios, the scenarios to which drivers are exposed can be rigorously controlled, and the frequency with which drivers encounter scenarios in which meaningful eye behaviors can be extracted can be greatly increased. Key Results of Driving Simulator Studies. Recent studies of eye movement behaviors on driving simulators have begun to help put together a picture of why there are large differences in the temporal and spatial distribution of the eye movements as a function of differences in drivers and the environment. Additionally, recent studies have used what is known about the factors that influence eye behaviors to change drivers or the environment in ways which reduce the most risky eye behaviors. Scenarios and Dependent Variables. The types of scenarios used to capture information on eye movements depend critically on the questions that are being asked. It is made clear what the key characteristics of the scenarios are for each question. The types of dependent variables that are analyzed include characteristics of the spatial (e.g., horizontal and vertical standard deviation; range; looked versus did not look; first- and higher-order transition probabilities) and temporal (e.g., total time on task; average glance duration in a particular area or at a particular object; maximum glance duration away from the forward roadway) distribution. Platform Specificity and Equipment Limitations. Eye movement behaviors are almost always studied on simulators that have at least three screens. Gathering data on a single monitor presents special challenges. Equally important, there are limitations in all eye trackers that can make collecting and analyzing difficult. These more technical issues are not addressed in any detail in this chapter.

Keywords
Eye Tracking, Eye Glances, In-vehicle Distractions, External Distractions, Crashes, Novice Drivers, Older Drivers, Music Retrieval Systems, Cell Phones, Traffic Control Devices, Digital Billboards

Key Points
• Much is known about where individuals glance and for how long they glance as a function of individual differences, activities, and distractions inside the car, and events and stimuli outside the car.
• Where drivers glance is a very good indication of what they are processing; how long they glance is a good indication of the difficulty of what they are processing.
• Crashes can be traced to three factors in which knowledge of eye movements is critical: especially long glances away from the forward roadway, failures to anticipate hazards, and looking but not seeing.
• Novice drivers fail to anticipate hazards much more often than experienced drivers, both because they do not appreciate the need to make anticipatory glances and because they are overloaded with the demands of driving.
• Older drivers fail to anticipate hazards at T-intersections largely because of cognitive declines, not because of physical, psychomotor, or visual loss.
• In-vehicle systems such as iPods lead to frequent and especially long glances away from the forward roadway, which can be partly mitigated by music retrieval systems with a single entry voice interface.
• Cell phone users are likely to look at safety-relevant information without fully processing it; they also fail to adjust their pattern of glances to focus on the most safety-relevant stimuli in the environment.
• Traffic signs and pavement markings can increase anticipatory eye movements.
• Digital billboards do not reduce the time the drivers spend with their eyes on the forward roadway, but they do reduce the time that drivers spend scanning the sides of the roadway.

Web Resources
Web Video 18.1: Video recording of a driver who fails to take an anticipatory look to the right for pedestrians entering a marked midblock crosswalk who are potentially obscured by a vehicle stopped in the parking lane.

Web Video 18.2: Video recording of a driver who does take an anticipatory look to the right for pedestrians entering a marked midblock crosswalk who are potentially obscured by a vehicle stopped in the parking lane.

Web Figure 18.1: Truck Crosswalk Scenario Perspective View (color version of print figure 18.2).

Key Readings
Chapman, P. R., and Underwood, G. (1998). Visual search of driving situations: Danger and experience. Perception, 27, 951–964.

Horrey, W. J., and Wickens, C. D. (2007). In-vehicle glance duration distributions, tails, and model of crash risk. Transportation Research Record, 2018, 22–28.

Maltz, M., and Shinar, D. (1999). Eye movements of younger and older drivers. Human Factors, 41, 15–25.

Mourant, R. R., and Rockwell, T. H. (1972). Strategies of visual search by novice and experienced drivers. Human Factors, 14, 325–335.

Pradhan, A. K., Hammel, K. R., DeRamus, R., Pollatsek, A., Noyce, D. A., and Fisher, D. L. (2005). The use of eye movements to evaluate the effects of driver age on risk perception in an advanced driving simulator. Human Factors, 47, 840–852.

Strayer, D. L., Drews, F. A., and Johnston, W. A. (2003). Cell phone induced failures of visual attention during simulated driving. Journal of Experimental Psychology: Applied, 9, 23–52.