Hostname: page-component-848d4c4894-75dct Total loading time: 0 Render date: 2024-05-19T08:47:10.727Z Has data issue: false hasContentIssue false

87 Virtual Driving Relates to Real-World Risky Driving

Published online by Cambridge University Press:  21 December 2023

Kathryn N Devlin*
Affiliation:
Drexel University, Philadelphia, PA, USA
Molly Split
Affiliation:
Drexel University, Philadelphia, PA, USA
Jocelyn Ang
Affiliation:
Drexel University, Philadelphia, PA, USA
Sophia Lopes
Affiliation:
Drexel University, Philadelphia, PA, USA
Aleksandar Gonevski
Affiliation:
Drexel University, Philadelphia, PA, USA
Oluwatoniloba Ogunkoya
Affiliation:
Drexel University, Philadelphia, PA, USA
Tasmia Hasan
Affiliation:
Drexel University, Philadelphia, PA, USA
Maria Schultheis
Affiliation:
Drexel University, Philadelphia, PA, USA
*
Correspondence: Kathryn N. Devlin, Drexel University, knd52@drexel.edu
Rights & Permissions [Opens in a new window]

Abstract

Core share and HTML view are not available for this content. However, as you have access to this content, a full PDF is available via the ‘Save PDF’ action button.
Objective:

Driving is a cognitively demanding activity commonly affected by brain injury and illness. Accurate driving assessment is essential for reducing risk, optimizing independence, and informing driving-related interventions. Virtual reality driving simulation (VRDS) enables safe, sensitive, objective, and standardized measurement of driving abilities. VRDS has been validated in relation to self-reports and driver records. However, self-reports are subjective, and driver records include only major events (collisions, violations). Video telematics platforms can measure naturalistic driving in a more objective and sensitive manner. The present study used video telematics to examine relationships between VRDS performance and directly observed naturalistic driving.

Participants and Methods:

20 healthy adult drivers (ages 23-61, mean age=36; 75% women) completed a VRDS assessment that included 1) driving on a straight road, 2) following a truck on a highway, and 3) reacting to a child running into a street to retrieve a ball. Primary VRDS measures were 1) speed and lane management on the straight road; 2) speed and following distance management in the truck-following task; and 3) reaction time, stopping, and distance from the child in the child-ball task. Participants also completed 28 days of naturalistic driving with a video telematics platform in their vehicle. Driving events were detected automatically using accelerometer, GPS, and video data, and driving behaviors were coded by driving risk analysts. The primary naturalistic measure was the number of unsafe driving behaviors per hour driven; specific driving behaviors served as exploratory variables. We examined correlations between VRDS and naturalistic driving variables. Given limited statistical power, we reported correlations that were small-to-medium or greater (r>.2) in primary analyses and medium-to-large or greater (r>.4) in exploratory analyses.

Results:

On average, drivers exhibited approximately one unsafe driving behavior per hour (M=0.9, SD=0.9, range=0.1-2.7). Common behaviors were failing to stop, unsafe following distance, speeding, and cell phone use. No collisions occurred. Average lane position in VRDS (specifically, leftward deviation from the center of the lane) was correlated with more real-world unsafe driving behaviors per hour (r=.35, p=.13), as were higher average straight road speed (r=.26, p=.27), greater straight road speed variability (r=.28, p=.24), and failing to stop for the child in the child-ball task (r=.22, p=.36). In exploratory analyses, failing to stop for the child was associated with real-world distracted driving (r=.45, p=.047), greater lane position variability in VRDS was associated with real-world unsafe following distance (r=.57, p=.009), and greater speed variability in VRDS was associated with real-world seat belt non-use/misuse (r=.49, p=.03).

Conclusions:

The present findings provide preliminary evidence that VRDS variables are related to directly observed naturalistic driving, supporting the potential utility of VRDS as a sensitive, ecologically valid driving evaluation tool. As the present study used a small sample of healthy drivers, further research will explore this topic in larger samples and in clinical populations, such as acquired brain injury. Future work will also investigate whether incorporating VRDS with conventional driving evaluation tools (e.g., neuropsychological tests, behind-the-wheel assessments) can enhance the ability of clinical driving evaluations to predict real-world risky driving.

Type
Poster Session 05: Neuroimaging | Neurophysiology | Neurostimulation | Technology | Cross Cultural | Multiculturalism | Career Development
Copyright
Copyright © INS. Published by Cambridge University Press, 2023