Data Science: Strengthening automotive safety with in-lab technologies

By Samantha McReavy

A high-tech driving simulator at the University of Guelph is using data to improve driver safety by identifying factors and behaviours that increase the risk of collisions.  

The University of Guelph DRiVE lab (Driving Research in Virtual Environments) is led by Drs. Lana Trick, Department of Psychology; Andrew Hamilton-Wright, School of Computer Science; and Michele Oliver, School of Engineering. The lab houses a driving simulator—a full car body surrounded by viewing screens that immerse drivers in a 300-degree, wraparound virtual driving environment.  

“It feels like driving,” says Trick. 

The simulator collects data on every aspect of driving, including lane deviations and steering, braking time to hazards, speed and aggressive manoeuvres such as tailgating and gunning the motor. 

Automotive fatalities are a leading cause of injury and death, especially in young adults. 

In 2020, motor vehicle accidents killed 1,745 people and seriously injured 7,868 more. These numbers were down one per cent and 12 per cent from 2019, respectively.  

The DRiVE lab researchers aim to reduce the number of collisions.  

“Even when collisions aren’t fatal, they can result in serious and sometimes lifelong injuries,” says Trick. “If we can reduce the number of collisions, we can prevent overall loss and suffering.”  

The simulator’s eye-tracking monitor determines where drivers are looking, and a 12-camera motion capture system monitors how drivers use controls (many drivers have difficulties with touch screens).

This equipment can measure the effects of distraction (cellphones, texting), fatigue, cannabis and alcohol. It can also gauge the effects of age and driving experience and the driver’s emotional state.  

Researchers may use the simulator to develop and test new in-vehicle technologies. 

Trick has also studied mind-wandering among drivers. “Many have the experience of finding themselves at home with no idea how they got there,” she says.  

Drivers distracted by thoughts become less aware of their environment. Mind-wandering is especially common among commuters who travel the same path day after day.  

Interestingly, listening to audiobooks may help prevent mind-wandering. Trick and her colleagues found that listening to an audiobook of Harry Potter on the road improved driving in some environments.  

But it’s not all about driver behaviour. Road design can also play a major role. 

Researchers can determine what changes may improve safety by creating simulations of existing and future roads. Trick says signage on roundabouts and the design of five-way intersections are among the options to reduce high collision rates. 

In a follow-up study with graduate students Heather Walker and John Bettinger, the team used the driving performance data produced by the simulator to determine when drivers were mind-wandering and to predict hazard response time.  

Using drivers’ self-reports of when their minds wandered and data from the simulator, the researchers were able to show that the drivers’ attention drifted before they even noticed it.  
Besides looking further at mind-wandering, Trick plans to study autonomous vehicles, including the transfer of control between driver and vehicle.  

This research was funded by the Canada Foundation for Innovation, the Natural Sciences and Engineering Research Council, the Ontario Centres of Excellence, the Ontario Ministry of Transportation and Kodsi Forensic Engineering.