Simulated Driving Assessment Fuels Teen Driver Research

Oct 15 2014

Simulated Driving Assessment Fuels Teen Driver Research


Drivers who turn the key of the driving simulator at The Children’s Hospital of Philadelphia Research Institute’s Center for Injury Research and Prevention (CIRP) quickly become immersed. Small details like active steering and brakes, an actual vehicle instrument panel and seat, the change in the cup holder, and a working radio with knobs create an authentic feel of being on the road. Pushing the gas pedal sets dynamic traffic situations in motion on the three high-fidelity screens. What a driving experience!

Helen Loeb, PhD, a biomechanical research engineer at CIRP, joined the research team one year ago and has the led the effort refining the programming that displays dynamic, realistic traffic scenarios and collects research and clinical grade data. The simulator offers the CIRP@CHOP team an opportunity to safely and reproducibly place drivers in situations that are the most challenging for teen drivers to negotiate — a major advance over relying solely on road tests. Weather, time of day, road friction, and traffic density are also programmed into the simulations, providing a wide variety of driving conditions to efficiently collect a broad representation of a teen’s real-world driving behaviors during a single drive. simulated driving

“If you’re on the road showing your kid how to drive, hazards appear just by chance,” Dr. Loeb said. “Here, we can create difficult traffic situations and expose teens to them in a safe environment.”

A CIRP team, led by Drs. Flaura Winston and Catherine McDonald, dedicated to teen driver safety research  is utilizing the driving simulator in a unique way to create a clinical-grade assessment of teens’ driving deficits with respect to skill and safety. The process of designing the Simulated Driving Assessment (SDA) tool began four years ago with funding from a Pennsylvania Department of Health grant. Two years ago, research statistician Venk Kandadai, MPH, became the project manager.

“When you think of an assessment for asthma, you have spirometry,” Kandadai said. “For cardiovascular disease, clinicians use electrocardiograms and stress tests. What we’ve built is a generalizable assessment that can essentially diagnose poor driving performance that is directly linked to the most serious crash scenarios for teens.”

Dr. Loeb has created a toolkit called DriveLab along with the SDA that reduces the immense amount of simulation and video data collected from participant drivers to produce easy-to-interpret results. Every action by the driver is recorded by the computer software in tandem with a portable eye-tracking system and a digital video that captures the driving scene, the participant’s hands on the wheel, and feet.

One of the studies conducted to validate the SDA as a research tool involved 21 teens ages 16 and 17 who had been licensed drivers for fewer than three months and 17 experienced adults 25 to 50 years of age who had been driving for at least five years with no history of police-reported crashes. They participated in 35- to 40-minute sessions of simulated driving. The study team compared the teens’ performance to experienced adult drivers who underwent the same simulations.

The SDA replicates the three most prevalent crash types among young drivers ages 16 to 19: crashing while turning left at an intersection, rear-ending a lead vehicle, and running off the right side of the road. Driving scenarios are set up in such a way that a collision will only occur when a driver makes a critical error.

“Typically, what we found is that teens tend to drive faster and follow the lead vehicle more closely than adults,” Dr. Loeb said. “I also found it fascinating that a number of teens didn’t brake as hard as adults in emergency situations. Some teens even missed the brake pedal or actually accelerated instead of braking. It seems like they panic for an instant and lose precious time. On the road, this is what makes the difference between a near miss and a bad crash. ”

Dr. Loeb, Kandadai, Dr. McDonald,  Dr. Winston and their colleagues Thomas Seacrist, MBE, Yi-Ching Lee, PhD (Simulator Program director), and Dana Bonfiglio (Simulator Program coordinator) performed statistical analyses and processed each driver’s data so that it all boiled down to a set of scores ranging from 0 to 117 errors. More errors were directly related to simulated crashes. A professional driving evaluator also independently reviewed videos of the driving simulator sessions, and his scores supported the SDA’s findings, which encouraged the research team.

CIRP’s SDA team is currently constructing an automated report card of a number of driving metrics to provide direct feedback to participants within a few minutes of completing the simulation. The researchers eventually will link video footage to the report so that drivers can replay their errors and view them from an outside perspective.

“They will see how things happen in a flash,” Dr. Loeb said. “There is no better way for them to understand than to actually experience it.”

Kandadai and Dr. Loeb anticipate that the SDA will be a widely applicable tool for scientists, clinicians, schools, and motor vehicle commissions. It also could be used to provide valuable information on other populations’ driving performance, such as adolescents with ADHD and the elderly.

To learn more, please check out part 1part 2  and part 4 of the Simulated Driving Assessment blog series.