Technology that would allow cars to talk to each other—to help prevent accidents and improve traffic flow—is about to get a real-world road test following new funding from the U.S. Department of Transportation.
Many high-end cars already come with sensors capable of spotting a vehicle in a driver's blind spot, or warning that the car is drifting out of lane. However, these technologies, which use radar, laser, or video sensors, have a limited view. Car-to-car communications could provide even more sophisticated earlier warnings—for example, when a car several vehicles ahead brakes suddenly.
Last month, the DOT awarded $14.9 million to the University of Michigan's Transportation Research Institute to test the technology, known as vehicle-to-vehicle and vehicle-to-infrastructure communication. The system to be tested relies on dedicated short-range radio communication to allow cars to signal one another and receive messages from traffic equipment.
The DOT estimates that 80 percent of serious crashes could be addressed by this technology. "This is the next major safety advancement, one that's comparable to seat belts, air bags, and electronic stability control," said Scott Belcher, president and CEO of the Intelligent Transportation Society of America, a nonprofit founded to promote advanced car technologies.
The technology will be tested in a variety of situations; it will alert the driver when it is unsafe to pass, and when someone is approaching an intersection at a speed that could cause a collision. Each car will be equipped with a radio that signals its speed and direction of travel, as determined by GPS, to other cars. It will also send this information to suitably equipped traffic equipment.
The University of Michigan is partnering with eight automakers, a number of which began working collaboratively to develop a uniform platform for implementing the technology in 1995. These carmakers will provide 64 cars equipped with the radios, while an additional group of ordinary cars will be fit with devices so they can transmit signals, making up a total of roughly 3,000 vehicles. Drivers will be recruited from among the 20,000 employees of the university's medical center.
Peter Sweatman, director of the Transportation Research Institute, says Ann Arbor is an ideal test bed, since it's a concentrated area with only three central thoroughfares out of the city, making it likely that the equipped cars will regularly encounter each other. The driving portion will run for a year, and data will be collected and may be used by the DOT's National Highway Safety Traffic Administration to decide, by 2013, if the technology has enough benefits to be approved. If approved, the technology would be rolled out over 10 years, Sweatman says.
"We believe this will happen in the near future," says Nady Boules, director of electric and controls integration research lab at General Motors.
Jim Keller, senior manager and engineer at Honda Research and Development, adds, "We see this technology as having huge potential in the future to affect safety."
The DOT's Research and Innovative Technology Administration, which is overseeing the program, released the following statement on the project: "This technology has the potential to be a game changer for safety. Research from NHTSA found that combined, vehicle-to-vehicle and vehicle -to-infrastructure technologies have the potential to address about 80 percent of all unimpaired car crash scenarios."
Joe Stinnett, a research engineer in active safety for Ford, is similarly enthusiastic. He says that, in addition to preventing common accidents, the technology could prevent traffic backups by keeping cars in step with one another. But he says one key area that needs to be addressed is security. "People could hack into the system, sitting on a bridge with their laptop transmitting false information," he warns. So a major challenge will be ensuring that the network is secure and that misbehaviors can be identified, he says.
Europe is on a similar track. In January 2011, the European Commission launched a three-year pan-European field test in seven sites across Europe to ensure the interoperability of the system. The effort includes 40 carmakers as well as suppliers, electronic manufacturers, and research institutes.
As vehicle-to-vehicle communication goes mainstream, it could even pave the way for fully autonomous driving. Google has been testing its own self-driving cars in California. So far those cars have logged 160,000 miles, but they rely on costly sensors. Vehicle-to-vehicle communication could allow for autonomous driving that's far less expensive, Belcher says. He expects that some autonomous driving features could appear in commercial fleets within five years. But he doubts that fully autonomous driving will take hold in the foreseeable future for one key reason: "Americans like to control their own cars," he says. (Technologyreview)