Wireless on the Road to Safety
by David Pescovitz
 Printer-friendly
version
Raja Sengupta, a professor of Civil and Environmental Engineering, is a principal researcher with California PATH.
|
In the last few years, the decreasing cost of wireless networking technology has untethered us from our Internet connections at home and at the office. Now UC Berkeley professor Raja Sengupta of the Department of Civil and Environmental Engineering is developing a system to bring WiFi into our cars. It’s not for Web browsing en route but rather a collision warning system that helps drivers keep their wits about them.
"Around 90 percent of crashes are caused by driver decision error or inattention," says Sengupta, who is affiliated with the Center for Information Technology Research in the Interest of Society (CITRIS). "Drivers fail to notice what's happening around them or they make a mistake reacting."
Sengupta's system, devised in collaboration with James Misener of the UC Berkeley-based California Partner for Advanced Transit and Highways (PATH), is based on an in-vehicle display that tracks your vehicle's location and those around you. For example, if you're tailgating someone, the screen might warn you to back off. Or an alert may sound if you're starting to change lanes and will likely hit another vehicle that's in your blind spot. The National Highway Traffic Safety Administration has been exploring such "active safety systems" for a decade or more. The problem though is that the ones that work are far too expensive to make it into the showrooms.
A screenshot from the collision warning system display (with a photo of the vehicles in a test run). The icon of the car on the screen becomes larger as the vehicle in front gets closer. Eventually, it turns red to indicate that a collision is imminent. (courtesy the researchers)
|
"These systems require 360 degrees of coverage," Sengupta explains. "And the sensor-based systems with that capability use radar or cameras costing hundreds or thousands of dollars."
The new approach--developed with support from General Motors --employs technology that has already plummeted in price due to economies of scale. Each vehicle would be equipped with a GPS receiver and WiFi radio.
"Commercial WiFi radios sold for home and office networking can be easily modified to work on a frequency band that the FCC has allocated for short range safety communications use," Sengupta says. "The radios are already in the tens of dollars range and might even tumble to cost a single dollar."
The GPS enables a vehicle to monitor its own direction, location and velocity. Every 100 milliseconds, that data is transmitted via WiFi to all neighboring cars. Since the system doesn't use range-finding sensors like radar, it can even gather information about vehicles that may not be in direct line of sight. Essentially, the automobiles are constantly forming ad hoc wireless networks as they pass one another, but instead of accessing Web sites or email, they exchange information about their physical place on the road. Sengupta calls it a "Cooperative Collision Warning System."
Sengupta and his colleagues have already successfully demonstrated the technology with a fleet of five vehicles. Still, the researchers have a long road ahead before the cooperative collision warning system is ready for commercialization. For example, data transmission protocols and error correction algorithms must be improved so that occasional missed bits of data, a given due to the speed and volume of cars on a freeway, don't result in hazardous system errors.
Meanwhile, the researchers are exploring concerns around privacy and data authentication. How can you ensure, Sengupta asks, that the data received is accurate as opposed to a faulty system or malicious attack on the network?
Once the technical challenges are met though, the "chicken-and-egg" problem must be faced.
"The safety system can't deliver much value at low market penetration," Sengupta says. "Your radio is basically useless unless the cars around you have one."
The researchers believe that government subsidies could help bring such a system to critical mass, the tipping point where the marginal cost to include it in a vehicle is less than the marginal value. And once the public becomes used to backseat drivers in their dashboards, the sci-fi notion of automated vehicles could become a bit more real. Self-driving vehicles raise a host of liability issues, Sengupta says, but there may be valuable baby steps on the way to automation.
"Perhaps if you're getting too close to the car in front of you, the vehicle could gently brake or exert a little back pressure on the accelerator," he says. "Of course, you could easily override the system. You need to be responsible for your own vehicle."
Raja Sengupta's home page
California PATH
Center for Information Technology Research in the Interest of Society (CITRIS)
Lab Notes is
published online by the Marketing and Communications Office of the UC Berkeley
College of Engineering. The Lab Notes mission is to illuminate groundbreaking
research underway today at the College of Engineering that will
dramatically change our lives tomorrow.
Media contact: Teresa
Moore, Lab Notes editor, Director of Marketing and Communications
Writer, Researcher: David
Pescovitz
Web Manager: Michele
Foley
Subscribe or send comments to the Engineering Marketing and Communications Office: lab-notes@coe.berkeley.edu.
© 2006 UC Regents.
Updated 3/1/06.
|
