Let's Agree to Head Toward a Dual-mode Solution

by Joe Palen

Maybe I'm missing something here, but there seems to be more agreement going on in this debate than disagreement.

It appears to me that a major point of Vukan Vuchic is that low density suburbs tend to have both O/D patterns which are temporally and spatially dispersed, and existing guideway infrastructure (e.g. roads) which are not used to capacity. This is the reason that individual traveler operated vehicles have evolved to provide wide-area transportation service to these regions, and the influences behind this trend, will probably continue. I don't think anyone qualitatively believes that a new replacement system could be deployed to serve all the needs for these low density areas; its just a quantitative question of the optimal density vs. cost which would substantiate a new system deployment.

I think everyone agrees that CBDs and capacity constrained corridors represent the highest priority for replacement of traveler-operated vehicle-generated congestion. Everyone agrees that a dedicated guideway which minimizes conflicts with the existing system can provide higher throughput, convenience, and time savings, but will require substantial initial deployment costs. Everyone agrees that an automated system decreases labor costs and increases traveler convenience, but will require a separate guideway. So if we are going to build a new system with a new guideway, it might as well make maximum use of automation.

The primary focus of controversy appears to be in regard to the size of the vehicles for the new deployment, the degree to which they should be automated and how travelers will access the new system. Vukan Vuchic appears concerned that a lot of automated small vehicles running around will be more difficult logistically to handle than a smaller number of large vehicles, will require small (and therefore potentially unsafe) headways to get good throughput, and will have a 500(+) Kg weight-to-traveler ratio which will increase emissions and/or energy consumption.

Edward Anderson sites various models which indicate that the logistics of even CBD surge loading can be handled with adequate system design for a small vehicle automated vehicle system. He indicates that such a system can inherently service a wider spatial and temporal range of O/Ds, and therefore has great utility and versatility.

Martin Bernard pointed out that a problem with the deployment of any new transportation system is that travelers will tend to use the existing (SOV based) system to access it, and therefore either large parking lots have to be constructed, or publicly available station cars may be used. He mentions that many modal options may be applicable (e.g. a true multi-modal system), but seems to indicate that a dual-mode PRT type system might be the most appropriate for corridors of more then a few miles that don't have existing rail tracks.

Dennis Manning points out that a large fleet of PRTs would have the convenience of a fleet of privately chauffeured taxis without the labor costs. Dennis also states that such a system might do well to borrow some operational concepts from AHS.

Palle Jensen expands upon the benefits of a dual-mode PRT in that it can do everything that a captive PRT can do, but also increases accessibility to those low density O/Ds that Prof. Vuchic is so concerned about. As we all know, Palle has some rather detailed suggestions on how this might come about (see the RUF concept description).

So it would appear that a dual-mode PRT with small-to-medium size vehicles may provide an ideal mix of speed, throughput, convenience, and accessibility. The vehicles would only have to accelerate once between origin and destination and could wind draft at close headways, which would reduce both emissions and energy consumption. However, I'm sure Prof.. Vuchic (as well as everyone else) would be concerned with the safety and technical feasibility of close headway spacing for a stream of vehicles capable of both automated and traveler-based control. From the recent PRT conference , we know that Michel Parent, Mark Buehrer, Jay Andress, and (of course) Palle Jensen all support some aspect of this dual-mode idea. We also know from Jerry Schneider's paper (and others) that cost is an extremely critical factor for real world implementations.

So how could a dual-mode PRT be designed with 1) low headways, 2) high safety, 3) low guideway costs, and 4) low vehicle costs? These issues are worth examining.

In regard to motorized vehicles, you can't beat the purchase price of autos. A number of manufacturers are already tooled up and competing to produce them at $10K - $20K a pop, and it is unreasonable to expect a comparably sized vehicle to be produced for less then this. Many, if not most, new vehicles come off the line with computer controlled throttles, computer controlled (ABS) brakes, and power steering. The marginal costs of making these things completely computer-controlled is not much - certainly much less then designing a whole custom new vehicle. Also, a fully loaded passenger van weighs in at considerably under 500 kg per seat, which I'm sure would warm the heart of Prof.. Vuchic.

Most PRT designs assume an elevated guideway in order to assure the automated vehicles are not impeded by unanticipated outside influences. However, there are already existing facilities closed off from the outside world that have contiguous right-of-way (ROW) directly along the corridors with the heaviest demand- they're known as freeways. Being as the economic, political, and institutional costs of installing a guideway through multiple political boundaries could be enormous, a narrow freeway medium (for a narrow PRT guideway) may not be a bad choice. Few things could provide more incentive to get SOVs into a PRT than to have the PRT vehicles whiz by them every morning as they are stuck in traffic.

The guideway structural cross section need only substantiate the weight of small PRT (and possibly freight) vehicles, which is a few orders of magnitude less than the design standards for freeways (which must support overloaded trucks). This minimizes the need for extensive geotechnical foundations, substantially reducing the construction time and cost. Because a PRT is narrow enough that it does not require widening of over-crossings (as do conventional freeway lane expansions) and because it would not interface with the conventional roadway system at capacity constrained interchanges (which cause huge SOV traffic control and construction phasing problems), the guideway construction could be quite fast, simple, and (most importantly) cheap. An extended elevated guideway could be expanded to city centers, airports, etc. as necessary; but the majority of the guideway needs to be inexpensively built (right-of-way, traffic control, construction, and political costs) to make an initial PRT implementation FEASIBLE.

But how do you get a conventional auto vehicle design to run on an automated PRT guideway at low headways with high safety? Well, some in the AHS world would say that you run vehicles (modified for complete automated control) along something analogous to a conventional freeway lane. Unfortunately, these lanes are too wide (e.g. suck up too much expensive ROW). Because the vehicles are not laterally constrained, a vehicle component failure or unexpected piece of debris could cause the closely-spaced following vehicles to careen into each other, crashing or rolling laterally. Also, these vehicles inherently would have to have some type of command and control (C&C) functions coming via RF through the air waves, making them susceptible to the safety problems of hacker terrorists and disgruntled nerds.

These problems are mitigated if we assume the lane is only slightly wider then the vehicle. Even if the vehicle is totally out of control, the worst it could do is toboggan to a stop. In fact, if we are going to have a physical barrier on both sides of the lane, we might as well use that as a continuous control surface to guide the vehicle (so that it normally never physically contacts anything). This close proximity to the guide rail means that a low power "leaky cable" continuous antenna can be used (without a FCC license) to provide virtually unlimited RF communication bandwidth to the vehicles for physically hack-proof C&C functions, as well as entertainment functions. {Travelers are going to want the capability to inexpensively watch cable TV, surf the net, video conference, play games, etc. in the otherwise non-productive time as they are being whisked to/from work}

Because this PRT concept has vehicles that are laterally constrained, guideway structural support need only be supplied directly under the wheels, making both ground-based and elevated guideway construction much simpler and cheaper. This physical configuration of load bearing and guiderail cross-section forms the guideway into a natural truss, providing both vertical and lateral stability for elevated sections (which is kind of important to us out here on the earthquake prone Pacific Rim) using the minimal amount of steel. The low weight design not only maximizes the span length, but allows pre-fabrication off the construction site, using "off-the-shelf" cross-sections, thereby decreasing construction costs.

What about headway, safety, and vehicle breakdowns? Well the platoon headway spacing should be pretty close to zero because 1) this would reduce energy consumption, and 2) if in the unlikely case a vehicle had a catastrophic mechanical failure where the drive train suddenly locked up, there would be minimal impact speed between vehicles. After the initial contact, the vehicles following in the platoon would simply slow down and push the broken down vehicle to the nearest egress station. However, this might scrape the paint off the disabled vehicle, so an additional safety feature would be helpful (see below).

Conventional rail systems, including monorail and Palle Jensen's RUF system, achieve both lateral control and weight bearing by physical contact with the rail(s). This causes mechanical wear, noise, and lubrication problems. Also, minor lateral imperfections in the rail construction can cause the vehicle to sway as it traverses the track, the momentum of which then induces greater lateral strain on the rail, eventually leading to (traveler disturbing) vehicle vibration and oscillations. This situation is avoided if we use a rubber tired vehicle with electronic lateral control (via a cheap IR distance transducer off the side guide rails) and exactly match the vehicle velocity to the super-elevation. {Smooth ridability is a desirable feature to travelers.} However, to assure absolute safety in case of a catastrophic control system failure, a mechanical fail-safe system would be useful.

It would be useful to have a small bracket extend from the vehicle to wrap around the guiderail (sorry, I can't provide a picture here) upon check-in procedure to the PRT. Although no part of this bracket would normally contact the guiderail for normal operation, it would provide an absolute fail-safe safety feature, as well as providing a convenient mount for the lateral offset control transducer. A right bracket would be extended to the right guiderail for vehicles preparing to access or egress the PRT. For mainline travel, the right bracket is retracted and the left bracket is used for primary control, thereby bypassing off-ramps. A physically contacting version of these brackets could also be used as an electrical pick-up off the guiderail to charge up EVs (or Zero Emission Vehicles-ZEVs), for a truly environmentally benign form of transportation. Indeed, some type of inter-city guideway electrical pick up may be necessary to make ZEVs into a viable consumer market and allow them to achieve the market share directed by some legislation.

The wrap around brackets could also be used for emergency breaking, greatly reducing the stopping distance over conventional tire-only breaking. Also, in the event of a control system or a mechanical breakdown, these brake pad surfaces would not abrade the guiderail. These small extendible vehicle brackets would be the only physical modification needed to conventional production vehicles (other than power steering actuator control). I think the additional feeling of safety provided the travelers (and liability protection provided the operators) would be well worth this small marginal cost.

Even if someone shot out all the tires of a moving platoon of vehicles on a turning elevated guideway during a freezing rain blizzard when there was a complete power and computer system failure during an earthquake - no traveler could be injured. This may be no small selling point.

It is possible to go on and on, further cultivating this dual-mode concept, but that's not really the point here. {These ideas have been much further developed then might be implied here. Left out are a number of important details, such as an inexpensive and reliable mechanism to exactly longitudinally sync the vehicles within a platoon, find a non-communicating lone stalled vehicle, find any other impediment that might intrude upon the guideway; construction details; geometric design considerations; etc. These issues may lie beyond the immediate interest level of the reader. Further details can be supplied as appropriate.}

The point is that I believe perhaps the most robust new transportation option is not just a viable combination of what Prof.. Vuchic, Mr. Bernard, and others have said here, but should include some of the more pragmatic ideas from other domains as well. A distillation of the most useful of these ideas, along with some of the more practical aspects of AHS thinking could, as Mr. Manning indicates, produce a more viable option then either the PRT or AHS communities alone have come up with so far.

Joe Palen is a Professional Engineer who has been involved in transportation system development for the past 20 years.

Return to the Innovative Transportation Technologies home page.

Last modified: January 21, 1997