Inside Edward Anderson’s personal computer, people commute by a form of transportation the world has never seen: bubble-shaped vehicles on elevated tracks that hold three adults. They punch in a destination, and the automated cars disperse through a dense, gridlike rail network. The stations are situated on sidetracks, so a car can bypass intermediate stops and cruise straight to its destination.
When Anderson fires up his simulation, hundreds of vehicles move around fluidly without crashing or getting stuck in traffic jams. What’s more, the cars wait for riders, not vice versa; an empty vehicle is never more than three minutes away.
The system, called Personal Rapid Transit, or PRT, combines the convenience and personal service of a taxi with the efficiency of automated public transit and seems tantalizingly cheaper than conventional rail systems. Sounds too good to be true? Well, there is one drawback: so far it exists only in Anderson’s computer.
That could change soon, however, because PRT may finally break out of its silicon cocoon. Chicago’s Regional Transportation Authority has embraced PRT as the solution to suburban traffic jams and given Anderson’s company, Taxi 2000 Corporation, $1.5 million to explore the concept.
Anderson, also a professor of aerospace and mechanical engineering at Boston University, has devoted more than 20 years to PRT. The system would require extensive computer power. Now that ultrapowerful microprocessors are cheap, PRT finally has a chance of getting off the ground.
The instantaneous traffic management of hundreds of cars would be a chore for a single central computer, so Anderson divided the task among three kinds of computers. A computer on board each vehicle has the simple task of remembering the destination. Computers at each grid intersection interrogate approaching cars about their destinations and flip track switches accordingly. These wayside units are directed by a central computer, which orders empty vehicles to stations and relieves congestion by shunting cars through less-traveled parts of the grid.
This scheme has won Anderson high marks from his colleagues. Given the software capability we have in this country, it really seems like a trivial problem, notes Jerry Schneider, a professor of civil engineering at the University of Washington and a PRT enthusiast.
What has gotten transportation experts interested in PRT is the low price tag of the track, by far the costliest element of a new transportation network. Anderson’s computer simulations revealed that the cars required only a slim elevated guideway that, at three feet across, is barely wide enough to walk along.
The simulations looked at the civil engineer’s familiar demons: static and dynamic loads, which are the forces exerted against a surface by still and moving vehicles, respectively. As a rule, a vehicle’s dynamic load is far more punishing than its static load. A long, heavy railcar making occasional journeys over a rail brings a heavy dynamic load as it passes. The vibrations are such, says Anderson, that to make the rail stiff enough for a big roaring thing you really have to beef it up. So while it’s obvious that lighter vehicles mean lighter guideways, what wasn’t obvious was how much lighter the guideways could be. The simulations showed that small vehicles running frequently create far less stress because they trick the track into thinking it’s receiving a single, static load. We found, he says, we could cut the weight by a factor of 15 compared with heavy rail.
All that remains, it seems, is for Anderson’s simulations to meet reality. He’s got these vehicles running around in his PC, says Schneider, and it looks really good, but until you actually do it and prove it, there’s always some question. Officials in northern Illinois still have to drum up $30 million to construct a test track. But if the project gets the green light, rapid transit may become as personal as Edward Anderson’s computer.