While new technology can create automobiles that won't crash, rival systems and their backers clash

04/07/2002

By DOUG BEDELL / The Dallas Morning News

First came seat belts and air bags. Then scientists used supercomputers to design shock-absorbing car frames and chassis. Now new technologies have enabled researchers to create the ultimate lifesaver – cars that won't crash.

Experts say it's only a matter of time before car fleets are routinely equipped with systems that significantly reduce the collisions that claim a life every minute worldwide.

The precursors of such "smart car" systems have already pushed their way into some Mercedes-Benz, Lexus and Infiniti models as expensive options.

American big-rig truck fleets are much further along. More than 10,000 trucks on highways are outfitted with radar-based collision-warning systems that alert drivers to fast-approaching danger and induce braking. Data collected over millions of miles shows the systems have reduced accident rates by 70 percent or more.

"They do what the inattentive or distracted driver should be doing," said Skip Yeakel, Volvo Trucks North America principal engineer.

Gradually, these "adaptive cruise control" technologies are being expanded with high-tech features that allow cars, trucks and buses to communicate with each other, forming perfect, fast-moving, tail-gating platoons, said Steven Schladover, deputy director of California Partners for Advanced Transit and Highways, or PATH.

Experiments sponsored by PATH and conducted with the University of California at Berkeley portend a future in which multi-vehicle platoons can be steered using sensors embedded in the pavement, allowing passengers to snooze, read or watch TV. Several hundred test "drivers" have already tried auto-steered platoons, navigating a 71/2-mile section of Interstate 15 during demonstrations.

"It feels like you're in a train – a train of cars," Mr. Schladover said. "You don't see any separation between the vehicles, and, after a minute of feeling strange, most people relax and say, 'Oh, this is pretty nice!' "

But significant barriers must be overcome before automated car platoons hit the highway. Those involved in crash-avoidance research say chief among them are the high cost of key components and a fragmented vision of which technologies to use.

All of today's adaptive cruise control systems are built around sensors that detect the vehicle ahead either by radar or lidar (light-detecting and ranging, the laser-based equivalent of radar).

To perform properly, these systems must apply the brakes when they sense impending danger. Therefore, the only cars that can use adaptive cruise control systems are those already outfitted with electronic braking systems, or "e-brakes" – typically high-end luxury cars.

Lidar is less expensive to produce and easier to package, but its beams are befuddled by rain and snow. To lidar, a blizzard could just as well be an approaching vehicle, so car designers have had to create mechanisms to deactivate adaptive cruise control systems whenever windshield wipers are flipped on high or when skid control mechanisms are engaged.

Using lidar also requires that car designers create peepholes to the outside world through car bodies. Designers abhor that.

Currently, only one automaker, Lexus, uses a laser-based adaptive cruise control system, and its handling of the feature demonstrates the inherent problems, said Stuart Harris of Tier One, an advanced automobile electronics research firm in Mountain View, Calif.

'Driver aids'

Second, because carmakers are leery of lidar's limitations, they've made it expensive, Mr. Harris said. Those buying the Lexus LS430 luxury sedan can add an $800 lidar system only after agreeing to purchase a $6,000 add-on package consisting of four-wheel drive, memory seats and a sun roof. A lidar package for the upcoming Infiniti Q45 will require a $10,000 add-on package.

Radar-based systems, such as those on some Mercedes, are more expensive, but they can "see" several hundred feet ahead, even in fog or heavy rain. They don't require peepholes, and they aren't fooled by precipitation.

But there's a philosophical clash.

Engineers pushing lidar say any collision-warning device should never work beyond the sight of the driver. Doing so would only encourage people to drive faster under bad conditions, they argue.

Radar proponents, meanwhile, contend poor driving conditions are exactly when drivers need the most help.

The lidar-radar argument has stymied progress, but marketing concepts are also flawed, said Mr. Harris, who has studied the industry for 11 years.

Because drivers are unfamiliar with the concept, carmakers are embracing the devices not as safety enhancements but as hopped-up cruise controls.

"I mean, sure, snow looks like a car to a laser, but what are you doing using cruise control when it's raining or snowing anyway?" Mr. Harris asks.

With such uncertainty, he says, many automakers have refused to change designs to accommodate adaptive cruise control, forcing engineers to build radar systems small enough to be mounted inside a car's front grille. That restriction has led to problems finding compact antennas that can properly focus radar beams on objects ahead.

Mercedes has pioneered a simple radar system that switches rapidly among three forward-looking beams strung across a single, short antenna wire, creating a scanning effect.

The system functions much like those being installed in truck fleets, gently applying the e-brakes automatically when slower-moving objects are detected ahead.

Fred Heiler, spokesman for Mercedes-Benz USA, said 10 percent to 20 percent of buyers are ordering the company's $2,875 adaptive cruise control package. In general, he says, Mercedes customers quickly adopt new features.

After all, Mercedes led the way in introducing anti-lock braking systems, skid control and other electronic features. The typical C Class model already contains about 150 microprocessors, leading engineers to joke that Mercedes puts wheels on a computer to keep it from dragging on the highway.

Mr. Heiler said customers tell him, "This system is so fabulous, it even helps me driving in the fog." Yet he is quick to point out that the radar system shouldn't be considered a safety device.

"It's a driver aid," he says. "It makes driving more comfortable, more relaxing."

At this point, adaptive cruise control systems are only effective when switched on during highway travel. The next step is making the systems work in heavy urban traffic conditions, the venue where most accidents occur.

'Stop and go'

Mr. Harris says reports from the Mercedes test-drivers are impressive.

"It doesn't let you hit the guy ahead of you; it just won't," he said. "In traffic, you can try and run into somebody, but it won't let you. They've had all these hot-shot drivers trying to defeat the system, and they can't do it. It's beautiful."

But with basic radar units now priced at $2,000 or more, it is doubtful similar systems will be available soon in anything but the highest-price passenger vehicles.

"You can't put an overly expensive system in a cheaper car because it just won't sell," said Craig Tieman, safety product manager for Delphi Delco, a leading developer of car radar components.

Still, there are strong incentives for expanding collision avoidance systems for passenger cars. Car and truck traffic accidents have killed about 37,000 people each year since 1994, according to the National Highway Traffic Safety Administration. Accidents will injure at least 10 million people this year, 2 million or 3 million of them seriously.

Hospital bills and property damage now equal 3 percent of the world's gross domestic product, says the Paris-based Organization for Economic Cooperation and Development. For the United States alone, that amount is about $200 billion annually.

For those and other reasons, the U.S. National Transportation Safety Board asked the U.S. Department of Transportation last June to develop performance standards for collision warning systems on large trucks and to make them mandatory on all new commercial vehicles.

The goal, said Mr. Yeakel of Volvo Trucks, is to cut commercial vehicle deaths in half by 2010.

"Some of these technologies create driver overload, but this is the opposite of that," Mr. Yeakel said. "It eases driver workload, and it just plain works."

Meanwhile, technology marches on. Last month, Philips announced that it will begin production this year of a microprocessor chip especially developed for advanced adaptive cruise control systems based on cheap, tiny digital cameras.

Available for 2005 models, this Philips-MobilEye unit will be built with enough computing power – 300 million instructions per second – to decipher complex images, detect passing and crossing vehicles, estimate a vehicle's yaw and pitch rates, predict traffic patterns and help send telemetry data to other vehicles traveling nearby.

And, before you know it, cars may very well be platooning inches apart, incapable of colliding – either on highways or side streets.

"The highway industry designs systems to last for decades; with the vehicle industry, it's years; with electronics, it's months," said Mr. Schladover of PATH and the platooning, auto-steering cars of California.

"Trying to get them all working together is a major, major challenge."