Slashdot Mirror
New Thoughts in Public Transportation
Matthew Shaylor writes "The BBC has the following article about an ultramodern public transport system to be tested in Cardif. Unlike conventional public transport, this consists of small cars that running on tracks can automatically take themselves to the correct destination. This allows there to be a mesh of tracks and stations thoughout a city, as opposed to traditional transport which tends to run along corridor routes to a city center. An interesting paper is available. Future versions may have dual control to allow people to drive the cars from the nearest station off the track to their homes. A true replacement for the car!"
Another good web site on the topic is Taxi2000.
Make sure you check out their FAQ.
The important topic that's always brought up is infrastructure. The beauty of the PRT design is that the infrastructure costs aren't all that appalling, since all the system needs to run is a narrow elevated track which can be built above existing roadways (so no right of way issues, etc). Yeah, it's more expensive than bus stations, but it's *way* cheaper than tunnels or elevated train track.
This is a concept commonly refered to as Personal Rapid Transit (PRT), a subset of Automated People Movers (APM) found at many airports. PRT has been around for a while and has somewhat fringe supporters (like me). Edward Anderson at the University of Minnesota has generated some of the most credible system designs and incorporated under Taxi2000. In fact, Raytheon developed a full test track of Anderson's concept outside of Boston; Bostonians can visit thier Marlborough, MA facility and see the future,!
The reason that PRT remains a fringe concept is related to economic challenges, not engineering ones. Although there are claims to the contrary, the general problem is that - like all public transit - PRT require a very high inital capital outlay. In dense urban areas, right-of-way costs are prohibitive. However, just as with information networks, public transit networks generate positive externalities: the larger the system, the more useful it is to everyone.
Furthermore there is little incentive to invest in expensive public works projects have prevented the testing of a fairly unproven technology. Public agencies would much rather invest in light rail systems that they have seen before than fancy driverless systems. Also, there is no conclusive proof that these decentralized systems can sustain the high corrider passenger/hour throughputs that make public transit so desirable for really dense urban areas.
Hopefully, projects like Cardiff will succeed and PRT will get recognition and legitimacy, but this is a technology that has been kicking around for a while and - as you can probably tell - is not insanely complex. As usual, economics and public policy get in the way of interesting engineering!
Urban Light Transport has more information on their web site, including some much higher-res images, FAQs and other info.
The most interesting (and not really mentioned) factor is that the automatic taxis don't travel on predetermined routes, they navigate their small network of paths to get to your destination.
This is called Personal Rapid Transit, and the first PRT system in use was a "demonstration project" in Morgantown, West Virginia, funded by the U.S. Dept of Transportation. (Morgantown is the home of West Virginia University, and the system linked the WVU campus and downtown Morgantown.) It was built in the early 1970s, but I believe it is no longer operating. Subsequent to the development of the Morgantown project a similar system was developed at the Dallas-Ft. Worth airport. All of the "ultramodern" features described for the system in Cardiff were used there: variable destinations, multiple route paths, standby cars to "flex" demand, etc.
The submitter of this article makes a slight mistake in his summary: PRT, including the Cardiff system, does not envision users being able to take vehicles off the tracks. There have been rail- and rubber tire-based PRT systems proposed, but even the rubber tire-based systems are designed for a dedicated, exclusive right of way. (Several mass transit systems, notably Toronto's, use rubber tires instead of rail.)
PRT suffers from a relatively simple problem: massive capital costs. I believe what finally killed the Morgantown project was a moment of clarity at the Urban Mass Transit Administration (UMTA, the U.S. D.O.T. agency that oversaw the project). A consultant pointed out that while the PRT system had been fun, it would have been substantially cheaper to simply buy every student and staff member of WVU a new car every two years. (My stepfather was the smart-aleck consultant.)
The Cardiff project? Three words: Big Government Boondoggle. The fundamental problem of PRT is the fundamental problem of Light Rail and Monorails too: they are dedicated right-of-way solutions that run along an extremely expensive path. (Even if the cost of construction is trivial, the cost of land acquisition is enormous. If the cost of land acquisition is NOT enormous then there isn't sufficient population density to support a fixed right-of-way system.) It is dramatically cheaper to buy buses. It is dramatically more efficient to run buses. Buses can change routes instantly--so buses that "prowl" the city center Monday through Friday can run on suburban loop routes among shopping malls on Saturday and Sunday. And a bus-based transit system only requires a marginal additional cost for right-of-way (bus stop marking, signs, shelters, etc.).
But buses don't have the sex appeal of big transit projects, so people still throw money at thirty-year-old concepts and call them "ultramodern technology."
How 'bout if we haul out the big networking technology of the time, and proclaim ARCNET as "ultramodern" networking?