March 2001

From University of Michigan

Reducing driver distractions in the age of overload

Detroit---Twenty years ago, when driver distraction meant fiddling with a tape deck or focusing on passenger conversation, in-vehicle disturbances accounted for 9 percent of accidents. Today, experts estimate that at least 25 percent and perhaps as many as 52 percent of crashes occur because motorists are paying attention to the myriad communication, information and entertainment systems in their vehicles.

Unless vehicles---and the devices in them---are engineered to help prevent such accidents, the situation will only get worse, says Barry Kantowitz, director of the University of Michigan Transportation Research Institute. Kantowitz, who has studied driver distraction for 15 years, discussed possible solutions during a panel discussion at the Society of Automotive Engineers World Congress March 5.

"Cars are becoming more like airplanes," Kantowitz observes. "But airplane pilots are very highly trained and closely regulated and licensed. Once you get your car driver's license---and that's a fairly simple test to pass---you don't have to go back and be retested every couple of years. You don't have to be certified to use any amount of vehicle information."

Not only are motorists untrained in dealing with information overload, but in this age of multitasking, they also tend to misjudge the dangers.

"People overestimate their ability to do a lot of things at the same time, and they underestimate the probability of rare events like traffic accidents," Kantowitz observes. "I have friends who know better than to use a cell phone while they're driving, but they say, 'I'm very careful; I only use a cell phone when it's safe.'" The problem is that on the road, situations can change from safe to unsafe in the blink of an eye.

In research he conducted before coming to U-M in 1999, Kantowitz studied drivers' tendencies to drift across the line between their lane and the next, a measure called lane exceedance. When a driver is fully focused on driving, lane exceedance is low, the research showed. "But any other task---tuning the radio, reading a tachometer, any other task at all---pushes those lane exceedances up," Kantowitz says. In his study, drivers were safe in a driving simulator, but on the highway, wandering over lane lines can cause accidents.

Unlike other car components, such as brakes and power steering, the devices known as telematics---cell phones, navigation systems and the like---involve human thought processes. And that's where the danger lies. Most driver distraction stems from two types of mental operation: attention switching and attention "starvation," says Kantowitz.

In switching, "your mind is on one task, and it takes a while to change over to another task," says Kantowitz. "So if you're thinking about some contract negotiation, you may notice something out the window, but by the time you switch your attention to it, you've lost a fair amount of the time you need to avoid a potentially dangerous situation."

In starvation, "you might be attending to the right thing, but because you're doing other things at the same time, the total attentional capability that you have is inadequate." Like switching, attention starvation also delays a driver's reaction to highway hazards.

What can be done to minimize in-vehicle distractions? Kantowitz ticks off the possibilities from least desirable to most appealing:

--Government regulation. Effective, but not sensitive enough; could result in simple bans rather than true solutions.

--Higher insurance rates for cars equipped with telematics devices. Consumers and manufacturers will probably object.

--Warning labels. Drivers may ignore, and companies may still be liable if accidents result.

--Industry standards. Good idea, but so far, none exist.

--Human factors guidelines. Have been developed for individual devices, but do not take into account integration of many devices and the resulting information overload.

--Agreements among manufacturers. Have worked in Japan, where car companies agreed to limit driver's ability to enter information while vehicle is in motion; haven't been tried in the United States.

--Good engineering design. Involves designing devices---and integrated systems of devices---such that they can't be used improperly.

The last option, "plain old, good engineering design," is "the right thing to do" from a technical standpoint, says Kantowitz. For example, systems could be designed to allow drivers to perform certain functions, such as entering a destination in a navigation system, only when the car is parked. Other functions---zooming in on an electronic map, for example---might be permitted when the car is stopped at a traffic light, but not when it's moving.

Still, "someone has to say, here are the functions and here are the limitations, and the car companies have to agree to do it," Kantowitz points out. Otherwise, in response to competition, automakers "will pile on feature after feature and wind up with very complex systems that are not always safe to use."

This article comes from Science Blog. Copyright 2004

Archives 2001 E