University of Virginia researcher: U.S. steering toward virtual driving tests

Allie Johnson

In the driving test of the future, you might hop into a simulator to navigate virtual roads where a bicyclist pulls into traffic, an ambulance comes up behind you with sirens blaring or rain beats down on your windshield.

Researchers at the University of Virginia School of Medicine have installed simulators like that at two DMV branches in Virginia and hope to recruit 1,000 volunteers to try them out over the next year. The goal? To see whether simulators could be alternatives to old-fashioned driving tests. talked with researcher Daniel Cox, a professor of psychiatry and neurobehavioral sciences at the University of Virginia, to learn more.

Who will take your simulators for a spin this year?

These are DMV customers who are getting on-road tests. There will be two types of individuals: adults who are getting their license for the first time and adults who, because of some medical concern, need to have their driving ability evaluated. We will compare results from the simulator with that person’s performance during the on-road test, and whether they pass or fail.

What is it like to drive the simulator?

Inside the simulator, there’s a big, curved screen with three digital projectors that project a virtual world in front of you and to the sides of you. It’s as if you’re sitting in a car, and you have the steering wheel, gas and brake pedals, turn signal and dashboard in front of you.

How does the simulator test a driver’s skills?

There are two halves to the test. First, we test vision, motor control and cognitive abilities that are necessary for safe driving. For example, we test dynamic vision, reading text as it moves across a screen. We look at peripheral vision. We also test foot reaction, where you drive behind a lead car at a fixed distance and speed. Reaction time is how quickly a driver can withdraw their foot from the accelerator when the lead car’s brake lights come on. Motor speed is how quickly they can move their foot to the brake pedal and apply five pounds of pressure. And foot confusion is when they actually increase the pressure on the gas pedal instead of withdrawing it; they think they’re putting on the brakes.

What causes foot confusion?

It’s common both with novice drivers who haven’t learned that skill very well, and also with older drivers who have lost that cognitive ability.

Do you also test overall driving ability?

Yes. After the first part of the test, they drive a course on urban, rural and highway roads. We track how well they do this and how well they’re following the rules of the road. For example, if a police car comes up behind them with a siren, do they pull over like they’re supposed to? Do they make rolling stops at stop signs? When a car unexpectedly pulls out in front of them, do they hit the brakes appropriately? That’s what we call tactical driving – the ability to take basic skills and apply them in a real-world setting.

Why replace on-road tests with simulators?

On-road driving tests are expensive and dangerous. An on-road examiner was killed last summer because the driver had foot confusion and plowed into a tree. So, if they could use a simulator and get similar results, it would save money and it would be far less dangerous. Also, it would be objective. When a person fails a road test, they can complain, “Oh, the guy didn’t like me.” And it is, in fact, a subjective judgment whether or not the examiner thinks that driver passed.

Does a real-life test hold any advantages over a simulator?

From my perspective, no. In an on-road test, they can’t put you in harm’s way. They can’t have a pedestrian jump out in front of you. The simulator exposes people to real-world challenges that can’t be duplicated in the on-road test. Also, if you do an on-road test in a rural community, you may never see a stoplight. But if you do one in a city, you’ll never see open roads. And none of these on-road exams evaluate people on the highway.

How will data gathered during the study help further the use of simulators?

We want to get drivers to drive the simulator so we have a very robust database. We want to see, “What is the average reaction time? What is the average frequency of foot confusion?” Then when we have people who come in for a driving evaluation, we can say their performance is either within normal limits or outside of normal limits.

Will simulators eventually replace on-road tests at DMVs across the country?

Yes, if you can demonstrate that this is valid and inexpensive and useful. Then, people across the country will embrace it.





Free Insurance Quotes