USING EYE TRACKING TECHNOLOGY TO COMPARE HAZARD DETECTION IN VEHICLE DRIVING VERSUS INTERACTIVE DRIVING SIMULATOR ACROSS TWO AGE GROUPS

ECU Author/Contributor (non-ECU co-authors, if there are any, appear on document)

Mary-Grace K McDonald (Creator)

Institution

East Carolina University (ECU )

Web Site: http://www.ecu.edu/lib/

Abstract: Rationale: Although current research exists regarding the validity of using a driving simulator compared to on-road driving for evaluation , there is a lack of evidence regarding hazard detection between the two driving conditions. Additionally , there is no clear consensus on whether old and young adults detect hazards differently. Purpose: This study utilized eye- tracking technologies to assess the validity of using a driving simulator compared to on-road driving for evaluation of hazard detection. The study also explored age differences in hazard detection across these two driving conditions. Design: This study was a 2x2 factorial design with repeated measures to determine if hazard detection is similar during on-road driving and simulated driving across two age groups. In addition , secondary analyses were completed to determine if hazard detection is similar between (1) on-road and simulated drives for younger drivers , (2) on-road and simulated drives for older drivers , (3) older adults and younger adults for on-road driving and (4) older and younger adults for simulated driving. Participants: Participants were 18 younger adults (ages 22-39) and 16 older adults (ages 65+). Methods: Each participant completed a simulated drive and on-road drive in their personal vehicle while wearing the eye-tracking glasses. Standardized pedestrians served as potential hazards and were placed throughout each driving condition in similar locations (three-way stop sign , pedestrian by the car , four-way stop light). Eye tracking metric including total fixation duration , fixation count , total visit duration , and visit count were analyzed. These metrics were compared between driving conditions and age groups. Results: Repeated measures ANOVAs were completed to compare simulated and on-road performance for all participants. Results showed significant differences between driving conditions for fixation count (F(1 , 20) , p=0.045) at the four-way stoplight and for total fixation duration (F(1 , 20)=13.669 , p=0.001) , total visitation duration (F(1 , 20)=12.876 , p=0.002) , and visit count (F(1 , 20)=9.547 , p=0.006) at the three-way stop sign hazard. Repeated measures ANOVAs compared performance between older and younger adults for both driving conditions with no significant differences found. For the secondary analysis using paired samples t-test , results within the older adults indicated there was a statistically significant difference between driving conditions for fixation duration (t(10)=3.491 , p=0.006) and visit duration (t(10)=3.406 , p=0.007) at the pedestrian by the car and for fixation duration (t(9)=3.833 , p=0.004) , visit duration (t(9)=3.841 , p=0.004) , and visit count (t(9)=2.751 , p=0.022) at the three-way stop sign. No significant differences were found with younger adult performance. Two sample t-test compared older and younger adults' performance on the on-road driving condition with no significant differences found. Two sample t-test compared older and younger adults' performance on the simulated drive. Statistically significance difference between older and younger drivers included fixation duration (t(20)= 3.255 , p= 0.030) and visit duration (t(20)= 3.236 , p= 0.034) at the four-way stoplight. There were no significant differences between age groups for the simulated drive at the three-way stop sign and pedestrian by the car. Discussion: The results of this study indicated that hazard detection behavior was similar across the driving conditions , with a few exceptions. Therefore , supporting the use of driving simulators as a safe mechanism to observe driving habits , behaviors , and mistakes without risk to the driver , evaluator , or environment. Additionally , these findings indicate that , despite age-related visual decline , older adults detect hazards similarly to younger adults , especially when assessing on-road performance.

Additional Information

Publication

Thesis

Language: English

Date: 2019

Keywords

older adults, driving, eye tracking, driving simulation, on road, hazard detection

Subjects

Email this document to