Can Youth with Autism Spectrum Disorder Use Virtual Reality Driving Simulation Training to Evaluate and Improve Driving Performance? An Exploratory Study

Abstract

Investigate how novice drivers with autism spectrum disorder (ASD) differ from experienced drivers and whether virtual reality driving simulation training (VRDST) improves ASD driving performance. 51 novice ASD drivers (mean age 17.96 years, 78% male) were randomized to routine training (RT) or one of three types of VRDST (8-12 sessions). All participants followed DMV behind-the-wheel training guidelines for earning a driver's license. Participants were assessed pre- and post-training for driving-specific executive function (EF) abilities and tactical driving skills. ASD drivers showed worse baseline EF and driving skills than experienced drivers. At post-assessment, VRDST significantly improved driving and EF performance over RT. This study demonstrated feasibility and potential efficacy of VRDST for novice ASD drivers.

A Pilot Study Assessing Performance and Visual Attention of Teenagers with ASD in a Novel Adaptive Driving Simulator

“…This result is in line with the reported outcomes of previous studies investigating driving in individuals with ASD (Classen et al 2013; Cox et al 2016; Daly et al 2014; Cox et al 2017). Notably, the majority of the observed trial errors occurred during driving tasks that involved turning the vehicle.…”

“…Cox et al (2017) conducted a rigorous study designed to assess the utility of simulated driving as a tool for driving training in individuals with ASD. This work also considered the impact of expert trainer feedback based on patterns of drivers’ visual attention by first collecting data from an eye tracker and then conducting post-training reviews to explore ways to improve attention.…”

“…This work also considered the impact of expert trainer feedback based on patterns of drivers’ visual attention by first collecting data from an eye tracker and then conducting post-training reviews to explore ways to improve attention. The researchers found that drivers with ASD performed more poorly than TD drivers on tasks related to general tactical driving and working memory (Cox et al 2017). The researchers also urge readers, rightly, not to generalize such negative results across all individuals with ASD because the condition’s heterogeneity, in some cases, may actually manifest itself in positive ways (e.g., higher-than-typical levels of caution and rule-following)—a point that has also been raised by other investigators (Huang et al 2012; Chee et al 2017).…”

“…The first step towards designing such a system would require an accurate, robust mechanism for monitoring eye gaze that can reliably detect the focus of a driver’s attention. The only other researchers to collect measures of gaze from individuals with ASD in a driving simulator reported only the proportions of gaze in different axes of the screen space (Reimer et al 2013) or used the collected data strictly offline (Cox et al 2017), and did not report fixation durations on specific objects or other regions of interest. The authors of this paper have developed such a system, and the studies presented here aim to demonstrate its sensitivity to different populations of users as well as its potential utility as an intelligent tool for driving intervention.…”

Motor learning and transfer between real and virtual environments in young people with autism spectrum disorder: A prospective randomized cross over controlled trial

“…Although, there are still some questions about whether the engagement in VR tasks is due to its novelty and motor interaction, studies highlight that motivation and attention are important features of VR systems (besides ease of use) which may contribute to increased interest and involvement in the tasks [Avola, Cinque, Foresti, & Marini, ; Osumi et al, ]. Moreover, VR may provide a step change to improve acquisition of a number of community skills and may enable learning in a safe controlled environment, improving adherence and enjoyment [Mineo, Ziegler, Gill, & Salkin, ] and allowing the possibility of enhanced transfer to real‐life behaviors [Cox et al, ; Patrick et al, ; Ross et al, ; Saiano et al, ; Saiano, Garbarino, Lumachi, Solari, & Sanguineti, ].…”