0ne of the dominant trends in technological innovation has been engineers' relentless success in stuffing ever more potatoes into the proverbial ten-pound sack.
Bundles of copper cable the size of salamis have given way to pencil-thick fiber-optic strands that carry hundreds of times more phone and data calls. Higher-speed mail sorters help move twice as much mail with the same number of people the Postal Service employed 25 years ago. Digital-compression techniques bring 400 channels to cable boxes and satellite dishes.
Now, the next holy grail in this endless quest to squeeze ever more through the same congested pipes is the great American highway.
The technology being developed as part of a three-year-old, $200 million federal research program goes by different names, such as smart roads, automated highways and intelligent-vehicle systems. The basic concept is to find ways to build enhanced roads where specially equipped vehicles can drive themselves.
In August, millions of drivers who endure soul-sapping, stall-and-crawl drives to work each day got a glimpse of what could be the new era. After months of sneak previews, officials of the National Automated Highway System Consortium formally unveiled an intelligent-vehicle system on a 7.6-mile stretch of the high-occupancy lanes of Interstate 15 just north of San Diego.
For four days, officials showed that it is possible for cars to be driven safely with a combination of computers in the trunk, magnetic nails buried in the pavement, radar-reflective road striping and dashboard-mounted video cameras.
In basic terms, the vehicles-including a bus driven by a Houston transit driver-were turned over to a very advanced form of cruise control. With magnetic sensors reading the embedded nails to identify the vehicle’s position on the road 250 times a minute and radar scanning the position of the vehicle ahead, computers handled the accelerator, brakes and steering.
The great promise of the technology is that it could squeeze far more cars into an existing roadway while safely increasing speeds from a choppy 15 or 20 m.p.h. to 65 m.p.h . or better.
“It’s very difficult and very expensive to build highways or add new lanes to existing highways. Therefore, we have to begin to use the highways we have more efficiently,” says consortium project manager James Rillings.
The widely flouted rule from driver education classes is that, to avoid rear-end collisions, you should always be able to count three seconds between the time the car in front of you passes a given spot and when you reach it.
In practical terms, at 60 m.p.h. that means a separation of264 feet between cars. With intelligent vehicles and automated highways, that distance could safely be crunched down to 20 or 30 feet.
How much it would cost to outfit a car with the necessary high-tech hardware is unclear. But automated highway planners are acutely aware of Detroit’s rule of thumb that car buyers will balk at spending more than about the cost of air conditioning-$800 to $1, 000-on any option, no matter how wonderful, so a key goal is perfecting mass-producable intelligent-vehicle systems in that price range.
Realistically, automated highway champions aren’t expecting to do more than double or triple the capacity of roads. But at rush hour, even 100 percent more capacity could go a huge way towards easing rampant congestion.
From 1970 to 1990, vehicle miles travelled in the United States grew more than 65 percent, and they are expected to double by 2020-but few imagine that highway capacity will double again. By some estimates, Americans collectively waste 5 1/2 million hours a day in traffic, and as suburban sprawl continues to overtake the countryside, those lost hours will continue to climb.
Ultra-vigilant computerized braking could sharply reduce accidents, given that driver error and inattention cause something like 90 percent of the 10.7 million crashes in the U.S. each year. Also, vehicle fuel efficiency could climb by 30 percent because, like gutsy Indianapolis 500 drivers who tailgate the race car ahead to take advantage of its slipstream, cars operating close together would encounter much less wind drag.
“It’s really exciting for about the first 15 seconds, then it gets really dull. It’s like driving with a chauffeur. You just let your mind wander,” Rillings says. The consortium, which began operations in 1994, has nine “core participants” including General Motors, Lockheed Martin, the California Transportation Department and major roadbuilders Bechtel Corp. and Parsons Brinckerhoff, and more than 120 associate participants.
More than 170 companies have joined the Intelligent Transportation Society of America. Allison Simmons of TRW Inc.'s Systems Integration Group, based in Fairfax, Va., estimates there is “potentially an $80 billion market to put this infrastructure in place.”
While the consortium is pushing ahead, it does not expect to have a working prototype for public use for at least five years. Some officials think it will be at least 15 years before automated highways are commonplace.
Consider just a few of the bugs that have to be worked out. How will this sit with insurers? What kind of innovation- chilling bonanza would trial lawyers reap from the first big pileup?
If, as is likely, automated lanes are built by peeling them off from existing jammed highways, what social-equity issues are raised by inflicting even worse congestion on working stiffs in pickups so that yuppies in souped-up Lexuses can read their Wall Street Journal all the way to work?
And will Americans trust the same government whose alleged incompetence they continually rail against to drive their cars?
“Before they spend money on it, they should show that a huge segment of the market will support it,” says Hank Dittmar, Executive Director of the Surface Transportation Policy Project, a coalition of 200 environmental, pro-transit and anti-sprawl groups largely funded by nonprofit foundations. “I don’t think the public will be that excited about letting government or big business drive their car.”
In a front-page USA Today story in April, Mike Doble, technology manager with GM’s Buick division, acknowledged that “the only thing we can’t do yet is get people to comfortably trust the system. It’s not a technology issue. Would you drive, closely spaced, at high speed, through San Diego?”
The consortium’s San Diego project is only one of many efforts underway to develop an automated highway. More than a dozen U.S. research universities have major intelligent-vehicle research programs.
In May, the Virginia Transportation Board awarded a $17.3 million contract to build the first two miles of what will be a six-mile automated travel lane connecting Interstate 81 to Virginia Tech University in Blacksburg. It is expected to be open for tests in 2000, and some- day after that to the public.
And in the Nevada desert near Carson City, the Federal Highway Administration has built a 1.8-mile track where, for more than a year, driverless trucks reading signals from a wire embedded in the pavement have been circling at 40 m.p.h. for 20 hours a day.
But this $12 million project has little to do with smart highways: The trucks are actually being used to test the durability of 20 different types of macadam that have been used on the WesTrack facility.
Even if automated highways never become a reality in our lifetimes, the research going into their development is already spinning off potentially marketable innovations.
One, a form of which is already being rolled out in Japan, is “adaptive cruise control,” which uses radar readings from a car’s front grille to determine whether the vehicle ahead is slowing down and automatically applies the brakes.
Other systems include collision avoidance warning devices that use ra- dar-a few thousand have been sold to truckers-and “lane-keeper” monitors using markers in the road and on-board sensors that warn a driver if the vehicle is veering out of its lane.
So even if truly intelligent highways may be a generation or more away, in the meantime, research will probably produce technology to help drivers be a little less stupid.