October 17, 2003
When newspapers publish my op eds about light rail, I often get responses from people asking about personal-rapid transit (PRT). PRT is a dream of combining the convenience of the automobile with the mass production of transit.
As suggested by one proponent, automobile-sized cars would run on fixed guideways elevated above the ground that would blanket an urban area on a grid. You go to a station that is no more than a couple of blocks from your home, get on the next empty car, punch in your destination, and the computer-controlled car will whisk you there at, perhaps, 70 miles per hour.
While your automobile just sits during the twenty hours a day that you aren't driving it, PRT vehicles can be used all day long. So PRT advocates claim that it can be built and run at a profit and won't need the subsidies required for other rail transit.
My response is always: It sounds attractive, but it's not going to happen. We already have a personal-rapid transit system. It is called the automobile. The cost of building a parallel system will be huge, and until that system covers an entire urban area, few people will use it. Even after it is built, many if not most people will still want to drive, which will reduce the revenues to the PRT company. So no financier will want to risk building PRT.
Nor am I convinced that PRT is superior to autos even in theory, as autos don't require a walk to the station. Moreover, being privately owned vs. commonly shared PRT cars, autos provide the same sort of security provided by private property vs. the common areas in New Urban developments that lead to more crime (see the review of New Urbanism by an English police officer).
Even if PRT were superior to autos, there is the classic problem in technological change: Once one pattern has been established, it is very hard to change to a new pattern even if it is superior. That is why I am typing on a QWERTY keyboard instead of the allegedly superior Dvorak keyboard (though Dvorak's superiority is probably just an urban legend).
In any case, the economics of PRT won't pencil out unless you can convince the vast majority of people to switch. You won't be able to do that without first giving them a PRT system that serves every place they might want to go. That will be very expensive, and who will risk the billions of dollars needed to make that investment if there is no guarantee that it will lead most people to switch?
While the dream of anonymous PRT vehicles running on overhead tracks seems unlikely to ever happen, there is a way to turn our automobiles into a PRT system that combines the advantages of private ownership with the safety features of computer control. It's called Intelligent Highways, and it involves putting sensors in the roads that will control the steering, acceleration, and braking of autos driving on that road. Auto drivers would keep hands off as long as they were on the intelligent road, and would only take control of their vehicles when they left the road.
Because cars won't need to be spaced several car-lengths apart, Intelligent Highway lanes will easily have double or triple the capacity of ordinary highway lanes. Once a network of Intelligent Highways are running, computers can also automatically route people on the least-congested route. This can make congestion a thing of the past without having to build a lot more roads.
The problem with Intelligent Highways is the same problem with PRT: Who will build intelligent roads if no one has a car that can run on them? And who will buy an intelligent car if there are no roads to run on?
Fortunately, the solution to this conundrum is in sight. State highway agencies in Florida and California are already wiring new roads so that they will be ready to provide Intelligent Highway services when the cars are ready. Meanwhile, new cars are being built with most of the hardware needed to make Intelligent Highways work.
To work on an Intelligent Highway, cars need to be able to accelerate, decelerate, and steer themselves. Acceleration is done through cruise control, and probably half the cars being built today have cruise control.
Deceleration is provided through adaptive cruise control, a system found on high-end cars today that can sense if a car is in front of them and will slow down or brake to maintain a safe distance. Like cruise control, this will soon be available on mid-range cars.
That leaves steering. The first car that can steer itself is the 2004 Toyota Prius. Besides being the most popular hybrid-powered vehicle, Toyota built some models of the Prius with a self-parking option. Simply tell the car you want to park and it will park itself, using sensors to avoid hitting other cars and steering itself into place.
Given a self-accelerating, self-decelerating, self-steering car, all you need is the software to interpret signals from an outside source and you can run on an Intelligent Highway. For Macintosh users, at least, software upgrades are easy.
Unfortunately, the self-parking Prius is only available in Japan -- Toyota was worried about liability problems in the U.S. But some other company may make the same feature available in this country.
The problem with PRT is it is a revolutionary system that requires people to completely change their lifestyles and travel habits. When automobiles were first introduced a little more than a century ago, they offered huge advantages over horses and carriages but did not initially require major infrastructure improvements for them to work. In the same way, Intelligent Highways are more of an evolutionary change, with autos and highways co-evolving together.
For the immediate future, American cities will need to remove bottlenecks and make other improvements to deal with congestion. In the long run, however, the twenty-first century solution to congestion will be Intelligent Highways, not a nineteenth-century technology based on rails or other fixed guideways.