Virginia Tech institute tests automated vehicle technologies
- November 27, 2013
Sometime in the 2040s, your son or daughter is running late for a meeting across town. They pull up an app on their implanted smart device and find the nearest available driverless vehicle. The pod-like conveyance pulls up and waits for its passenger, like a theme park ride. The automated taxi drops the rider off at the appointed destination and then whisks away to retrieve its next client.
Except for the body cyber implant (which, really, is as good as anyone’s guess at this point) that may be a fairly accurate vision of the future of transportation, which is being pioneered at the Virginia Tech Transportation Institute (VTTI). The institute, which celebrated its 25th anniversary in November, is road-testing and researching a variety of new vehicular technologies that are years — and in some cases decades — ahead of what one can find in today’s auto showrooms.
“Long term, things are moving towards automation … but it’s an evolutionary process,” says Thomas A. Dingus, the institute’s executive director. “There are production vehicles out this year and last year that automatically brake. There are production cars out this year and over the next year that will actually do lane centering. That’s sort of the first step toward automated steering. … We’ve tested a lot of that technology here at VTTI over the last decade. It’s all moving toward full automation, although that’s quite a ways away. … Realistically you’re talking about 25 or 30 years, but it will go fast if the last 25 years is any indication.”
This year, Dingus was honored at the White House as one of the Obama administration’s 2013 Champions of Change for his work on 21st-century transportation solutions.
Since becoming VTTI director in 1996, Dingus developed it from a 15-employee center into the nation’s second-largest university transportation research institute, now employing a staff of more than 350. Its annual budget grew from $2 million to about $40 million in just a decade as Dingus built partnerships with the federal government and the automotive industry. (About 25 percent of VTTI’s annual research budget comes from private industry; the balance is from state and federal funding.) The institute conducts more than 200 research projects a year, including testing new technologies for automotive manufacturers as an independent research facility.
VTTI operates the Virginia Smart Road, a closed, test-bed facility in Blacksburg that includes a 2.2-mile test track and the 175-foot-high Smart Road Bridge, the tallest state-maintained bridge in the state. Researchers have logged more than 16,500 hours at the facility. This year, among many other projects, Tech researchers have road-tested the driverless Google Car at the Smart Road as part of a project in cooperation with the U.S. Department of Transportation, the National Highway Traffic Safety Administration, Google and General Motors. The project tests how drivers react to warnings to take control of semiautonomous vehicles.
There are different levels of automation in vehicles, explains Myra Blanco, who leads VTTI’s Automated Vehicle Systems research group, including features such as automated braking, lane changing and lane centering. Another technology has the car knowing its position relative to the car in front of it.
In the Google study, Blanco says, “We’re basically looking at the human system interaction, looking at how the driver will hand off or cede control to the vehicle, then if something happens, how they would regain control of the vehicle again.” For instance, if the car encounters an unexpected road condition such as road construction, it could signal the driver to take over. VTTI is studying how to make that transition seamless and safe.
The goal that transportation researchers are working toward is a fully automated car into which the user can input a destination and let the car safely do the rest. No company has gotten to that point yet, Blanco says, but the Google Car and projects such as CityMobil, a European pilot project testing automated mass transit, are steps toward the future.
Conceivably, automating vehicles could virtually eliminate car crashes, alleviate traffic congestion and reduce gasoline usage, Dingus says, but first there will have to be redundant, reliable systems that are beyond foolproof. “We still have 32,000 or 33,000 fatalities a year. People say once vehicles are automated, that number will drop substantially,” Dingus says, “but if you’re in an automated vehicle and it crashes, it’s going to be on the front page of every newspaper in the country.”
Toward that end, a major component of VTTI’s automated research is the ability to create a wireless network of smart vehicles sharing the road and trading data with one another.
VTTI is conducting connected vehicle tests on the Smart Road and in real-world driving conditions on its Virginia Connected Test Bed, a stretch of highway with 43 wireless networking devices located at intersections along the Interstate 66 corridor in Fairfax County near state Routes 29 and 50.
Using wireless technology, connected vehicles can build a virtual map of which cars are around them and where those cars are heading. Cars could relay information to each other such as warnings of slick roads so your vehicle will know to slow down, says Zachary Doerzaph, director of VTTI’s Center for Advanced Automotive Research. Connected vehicles also could collect information from road crews about construction hazards.
One of the studies VTTI is conducting examines how stop signs could become “virtual yield signs,” Doerzaph says, allowing you to drive through the intersection without stopping if the car knows no other vehicles are nearby. Another potential use would allow state transportation departments to improve traffic flow by directing traffic to different lanes depending on the vehicle’s destination.
VTTI is also looking at customizable software inside the car. Right now that means entertainment apps such as Pandora or Netflix, Doerzaph says, but one day the user could purchase additional software increasing the smart technology capabilities of the vehicle.
There are questions to be answered, including whether the network would compete with cellphone bandwidth or operate solely on short-range technology. Also, due to infrastructure costs, it’s more likely that fixed wireless broadcast nodes would be located at strategic locations instead of being ubiquitous on roadways, Doerzaph says. And for safety’s sake, systems must be more reliable than a cell phone connection.
Other challenges to using automated vehicles include regulatory questions and phasing out the existing fleet of non-automated vehicles. Cars with the newest automation and smart features likely will be more expensive, Dingus says, so it could take 25 to 30 years before everyone is driving a connected or smart car. VTTI also is looking into whether older cars could become retrofitted with adaptive technology, Doerzaph says, but those cars probably wouldn’t have as high a level of functionality as cars with built-in technology.
Other related research at VTTI examines “naturalistic driving” — monitoring drivers’ habits in the vehicle with technology including cameras and monitors. VTTI looks at factors such as texting (very bad), cell phone usage (surprisingly not so bad because the driver’s eyes are forward on the road when talking) and fatigue (a huge factor in crashes and near-crashes). Eventually VTTI research may be used to create more advanced, built-in monitoring applications for novice teen drivers that could send detailed report cards to parents about the teens’ driving habits, similar to current commercial products such as DriveCam and IntelliDrive.
“We want to make sure we get this right the first time,” Doerzaph says. “The worst possible outcome is for us to do all this work and when it gets deployed, it decreases safety instead of increasing safety.”