Saturday, November 7, 2009

58 > Defining The Problem

One of the hallmarks of PRT/PAT travel is that the vehicle takes you from your origin directly to your destination with minimal waiting time. There are no transfers, no stops for other passengers, etc. The track, being for small vehicles, can be so light and inexpensive that it can be built into very extensive networks very cheaply.

In practice, however, the initial track layout will be extremely limited and it will take many years to spread citywide. Therefore it probably won’t take you from origin to destination. Furthermore to really make an impact, PRT should address the commuter, not just travel around the central business district. What is the point of PRT if you need to drive 10 miles and pay for parking to get to it? Do the models that various PRT venders are promoting address the crowded freeways? Is any PRT model that addresses the freeways even close to the model that addresses the central business district? What about the suburbs?

If the need were only in the Central Business District (CBD), I think I would give the edge to the general design parameters envisioned by ITNS/Skyway Express models. This will surprise many, as I have a record of advocating a hanging system. The difference is, in the CBD, the main emphasis is footprint. Speed, and cost of track and stations are secondary. Having elevators in every station is actually a pretty good way to economize sidewalk space, and it takes less energy to lift a passenger than a whole vehicle. In an environment of all multi-story buildings, second floor stations can be accomplished with little more than a balcony.

The situation changes markedly as the system expands outward from downtown, however. Here, the expense of the track and stations becomes critical as individual station ridership begins to drop. In these areas, bus rides to town are short and convenient. Putting up elevator-equipped stations on every block is far less attractive than it was downtown. Now the advantage, in my opinion, shifts to hanging systems, because they are generally more versatile in terms of slopes and curves and multiple speeds, and can accommodate open-air stations that may double as bus stops. Because of population density, stations should still be within walking distance of each other, but many passengers will just be passing through.

Next comes the suburban sprawl. This mixed-use area goes on for miles and is punctuated by mini urban centers, residential neighborhoods and distribution parks. There is enough housing and employment that many of the residents never go downtown. Here distance, and therefore speed, starts to become a real issue. The potential riders currently utilize a variety of road types to get around. Often there is a freeway nearby. The challenge for the PRT planner is to provide a system that can match speed and convenience with the combinations of freeway and back street shortcuts that are utilized by the drivers living and working here.


Because of the sheer size of the suburban sprawl, it is unlikely that any form of PRT will blanket such an area for quite some time. This suggests a “low hanging fruit” strategy, where major “hub” areas are accessible but many low volume routes are not incorporated at first. City buses (horrendously inefficient for long trips because of the many stops) could, none-the-less, be a reasonable option for going a few blocks to get on the PRT grid. Here it would seem to make sense to have some kind of PRT express lanes that go quite fast to connect community hubs. This is an entirely different model than the CBD, both in terms of preferred track and stations and preferred vehicle. Speed, versatility and track cost would seem to be the main design factors.

Finally there are the outlying suburbs and satellite communities. Commuters typically travel at posted speeds until they get close to town and then traffic backs up. These people generally have large engine vehicles because aggressive driving, for them, is somewhat of a survival skill. Keeping these cars out of city limits would do a city a lot of good. Here 30 mph PRT would be useless. It must be much faster to compete with the freeway. People in outlying communities cannot expect to have PRT vehicles come by their homes. Dual mode vehicles would be a poor substitute for the pick-up trucks and SUVs that get them around now. To be perfectly honest, I have serious doubts about whether PRT is the right tool for this job. It certainly calls for fast vehicles on a fast track, but why individualized vehicles? A park-and-ride GRT (Group Rapid Transit) station on the out-of-town side and an ad hoc drop off scheme might be a more efficient. Here is some thinking on this.

The main drawbacks with group travel are waiting and having the station locations that are catered to the average passenger but are not anyone’s exact origin or destination. I believe, however, that with automation and an intelligent system these problems can be largely solved. For example, the problem of fixed scheduling and associated waiting is largely a communication problem, as is inconvenient transfer locations. (Passengers and transit have had to meet at a prearranged time and place, because it is presumed that they can’t talk to each other.) If the “system” knows the complete itinerary of every passenger, the right size vehicle can be sent at just the right time for the group. Is transferring a really the problem if the time waiting for the transfer vehicle is eliminated? After all, PRT vehicles can swarm incoming GRT vehicles moments before arrival, and the “system” can decide where this meeting would take place. GRT requires heavier track, but the other choice may be requiring all PRT vehicles to be more costly and robustly configured than would otherwise be the case. Building super fast PRTs to go slow is a waste. Building high-speed high-capacity track might be a better investment. After all, this might find dual use for freight. There is also the matter of spreading the weight. Perhaps a “fast lane” with greater headways between vehicles and greater spacing between bogies would not need to be that much more expensive.

The track profile I have been working on is specifically designed to be adaptable for multiple weights and speeds. This brings up the question of track “permissions.” Obviously a heavy vehicle must not use light track, but light vehicles could use heavy track. Clearly slow vehicles should not hold up fast ones, but fast ones might want to use a slow track, on occasion. These questions are for a different day, but, to rap it up, I suspect automated transit is not a “one-size-fits-all” technology, and different parts of a city have differing transit needs, and therefore different optimal designs. I also wonder… What starting configuration gives the most bang for the buck?

Next week: The long awaited grand opening of the design collaboration site.



16 comments:

Andrew F said...

I think you raise a lot of good points here. I suppose this builds on your post about 'high-speed' PRT. There is a sweet spot between utility and cost where the PRT network+vehicle system competes well with the car for most trips while minimizing total system cost. I believe you are right that GRT is probably the way to go with trips beyond the 90th, or 95th percentile of distance, allowing higher speed-capable vehicles.

An additional thought here is that we could have two (or more) classes of vehicle as an alternative to separate GRT system. The faster vehicles would be more costly to produce, but hopefully not much heavier so as to be compatible with the standard guideway. Since most trips would not require these vehicles, they need only make up a small fraction of the fleet. This could be offered in lieu of a GRT system for long-distance trips, or as a premium service for those willing to pay for the convenience and privacy of an individual vehicle.

As far as the best place to start with a PRT system, I suspect that might be the inner suburbs. This avoids the major engineering challenges of building in the inner city, and avoids competing with high-speed, convenient forms of transit (I'm thinking subway). Perhaps the best idea would be to replace or augment a bus line that is reaching capacity with a simple PRT line along some or all of its length. Better a bus line that has some reasonable trip generators along its length, rather than just at the termini. Maybe a suburban centre and a higher-order transit line, like subway or commuter rail.

cmfseattle said...

US$200.00 a foot=
$1,056,000 a mile=
$580 a day, over 5 years=
5.8cents a mile, if 10,000 trips utilize that mile each day.

so, for an average pop. dens. of 3,000 per sq. mile, PRT/FRT would cost roughly 20cents a mile=
$4.00 a day for a 20-mile commute ($6,000 and 30,000 VMT over 5 years).

cmfseattle said...

check this out: http://shweeb.com/index.php?m=transport
http://shweeb.com/
their video shows a working prototype. i know from experience that an LiFePO4 battery and hub motor would add only 30 pounds, for a range of at least 10 miles (another 15lbs. would double that range). it recharges in about 3 hours. upfront capital is ~$700, incremental <5cents/mile.

since you don't really need a control system at first (to average ~10mph; at least now, there's a defined, affordable pathway toward control testing at higher speeds), this idea seems to overcome most of the objections. color me excited!

----

petroleum extraction yields a high Return On Investment, because you get both energy and materials. cheap petroleum is responsible for the choices we (in industrialized nations) have.

however, demand for petroleum is on the rise. i predict that eventually, AGT will affect pop. distro patterns because it can use resources more efficiently and therefore, can be more affordable than automobiles on roads.

congested roads/highways are the effect, cheap oil the cause.

p.s. my back-of-envelope calcs assume 1 guideway-mile per sq.mile, which would likely be 2 or 3 times that amount. and yea, i left out other costs, but i think commuter transit mode share could be at least doubled with "Tha Shweeb." i hate riding my bike in traffic/rain, (i do it because, even at-grade, it's the cheapest and fastest method for my 5-mile distance).

Dan said...

Dan the Blogger Responds-

Thanks for your thoughts. I like your idea of a “sweet-spot”. Sometimes a couple of words help frame a problem. About your idea of a “premium” PRT vehicle, I am pretty much on board with that idea, and this is why.

I have always winced at how many hats PRT venders are trying to wear – vehicle manufacturer, track builder, city-planner, software vender, etc. The whole idea of open source PRT was to get some standards established, so that various companies could work together based on their strengths. If there is a standard track that various vehicle manufacturers can build for, there can obviously be competing and/or special purpose vehicles. I like the idea of the vehicles evolving over time to go faster, or become more energy efficient or more comfortable. Design Darwinism.

As for the speed issue, about the only thing I can say is that it seems to me that the vehicles must go faster, or at least get you there faster, than cars do. That would mean that PRT routes should match or beat the speed of the streets they follow, taking lights stop signs and traffic into consideration.

One final observation – The weight/ cost considerations that are generally used for PRT are based on the premise of a downtown situation. The idea is to span fairly long distances without support. It seems to me that in a freeway situation there would be little objection to having supports spaced more closely. There could also be A frame supports, or other schemes that may enable greater weight without additional cost. This could enable GRT on a PRT track, just not on the track downtown with the long spans. From the point of view of trying to establish open standards, this would indicate a need for a high-speed GRT capable freeway track and a regular neighborhood track. I’m not sure about your 90-95% criteria, though. In some cities there is substantial traffic for a dozen miles or more from downtown. I can envision an 8 or 10 passenger vehicle arriving at a large park-&-ride every 1-3 minutes during rush hour, which would take passengers at above freeway speeds into the “grid” where there are waiting PRTs (that only go half as fast) to finish off the trip. True, there could be high-speed single passenger PRT vehicles on the track as well, but this scheme would reduce the general cost of the fleet.

Dan said...

Dan The Blogger Continues...

Hi cmfseattle- I really don’t know what to say about those numbers, because I don’t really understand where they came from. I am uncomfortable with numerical extractions that have an element of supposition at their root. The more you steps you put the figures through, the more they seem sound, and that can be dangerous. If the supposition is false then it's all false...
That Shweeb video and homepage is provocative. It reminds me of my first musings about PRT, which were very much about minimizing everything. (Like the J-Pods slogan, “It costs less to move less”) The whole Americans with Disabilities Act sort of shoots down that approach, though, and I’ve resigned myself to having the lightest track support, say, 50% ADA compliant vehicles.

I also must say there is a huge cultural and weather influence at play here. In Houston, we put in miles and miles of bike lanes that nobody uses. Of course we have to shower after any outdoor time most months of the year, and have frequent ozone alerts. In Tuscan, where I spent the better part of the last 2 years, there are few bike lanes but everybody is out riding. Up in New England there’s a lot of biking for a couple of months on nice days…I have not been in the State of Washington for over 30 years, but I understand you guys ride a lot.

I have yet to see a decent roof or bubble or something that can be bolted onto a bike for rain… What’s up with that? Anyway, in hot climates that Shweeb design looks like a horrible way to get heatstroke. Where’s the AC? As for motorizing it, well, it’s a terrible thing the way weight creeps in. Look how fast you can go on a little dirt bike. Yet weather proof it and add a couple of extra wheels and more comfortable seats, and voila. No performance even with 5 times the horsepower. Now you must carry around a bigger engine, which means a stiffer frame, which means bigger shocks, brakes, steering gear… That in turn means you need a still bigger engine, and on and on…

I’ll do the Shweeb people one better… ditch the capsule and seat and replace it with a Yucatan hammock. That weighs less than 2 lbs. Motorize that! (Isn’t that a picture… motorized hammocks….) Actually that reminds me of my hero, the late Buckminster Fuller, who pointed out the advantages of structural designs that separate structure into distinct compression and tension members, so that the tension members can be replaced by rope, string, cable, etc. (spoked wheels, suspension bridges) A hammock is separated this way. It’s all tension members, relying on the user to supply the parts that provide compression or shear strength.

I personally would like the Shweeb better (as an amusement ride) if I could hang facedown, hang-glider style, in the open air.

cmfseattle said...

how hot do auckland summers get?

have you asked dr. anderson about standardization? e-mail addy is on some of his papers. try the gmail one. i'd trust his numbers.

Dan said...

Thanks for the link cmf…
No, I haven’t contacted Dr. Anderson, nor has he contacted me. When I first started this blog I was particularly critical of the Raytheon phase of PRT design and sort of painted him, taxi 2000, and Raytheon with the same brush, until some alert readers showed me the error of my ways. I later found out that many aspects I was critical of, (especially the track size) had already been remedied. Later I wrote a post critical of his paper comparing suspended systems and supported systems, which I still regard as incomplete and biased. I none-the-less respect the man and his work (overall) greatly. I wouldn’t be surprised if he was a bit miffed at me, however.

I don’t know what he thinks about standards but it seems to me that he and Taxi 2000 should put the past behind them and work toward that end. For supported, urban systems, their designs, are, in my opinion, pretty hard to improve upon. I guess I just regard the downtown market as a bit of a niche. At some point I imagine people will start working on such standards right here though...I hope they do.

Andrew F said...

Dan, I think the key is that People Hate Transfers. People are willing to have their trip take a few extra minutes if it means they don't have the gather themselves up, exit a vehicle, wait in potentially unpleasant weather, board another vehicle, and resume what they were doing. Also factor in the strike that people like privacy, and GRT is a tough sell, except for longish trips.

The challenge of having multiple vehicle classes is that you make it increasingly unlikely that the vehicle you want is anywhere nearby, much less waiting for you at the station. That's why even going from one to two is a big negative, but one that might be sufficiently offset if it meant there would be a supply of high-speed capable vehicles mixed in to the fleet. But much more than that is asking for troubles, or transfers.

Dan said...

Hi afransen-
Believe me. I understand that this is a compromise.

However, I think people hate taking longer than the driving time even more; I would stay with my guideline that the system must represent an overall time savings to attract customers. Keep in mind the part I said about the PRT transfer vehicles swarming an incoming GRT. One PRT for one fare. No waiting. You seem to assume this isn’t possible. I am assuming that it is. If you are referring to transfers to neighborhood buses or light rail, well, that’s another matter.

Privacy is of course preferred, yet experience shows that people will use a system that saves them time and money. The case that GRT is not the preferred alternative over PRT is unquestionable, assuming sufficient track space, energy/speed, vehicles, stations, etc. But I am thinking mostly about trips to/from “park&rides” (public parking lots close to the freeway, well outside of town) on track that is designed for high speed. The GRTs might have speeds 3 times faster than PRTs. The alternative would be to funnel a million pokey little pods onto this long straight run where they would only beat freeway speeds during rush hour. I am thinking GRT exclusively for freeway median use, not suburbia.

There are a couple of possible “work-arounds” that come to mind. One is the possibility of having all high-speed be accomplished through “helper” or “engine” bogies. Because the PRT motors would run faster based on frequency, not voltage, (until they overheat from being overworked) helper bogies become possible. This is the idea I brought up in #56. Another alternative is to just build all of the PRT vehicles with oversized engines. This is not as crazy as it sounds, because unlike gasoline engines, electrical motors tend to draw only the power that they can put to work. Sure, it’s a waste to carry around the bigger motor if its power is not being used, but the electric bill won’t be much higher on account of it. It does jack up fleet costs though, and it does put a lower top limit on speed.

I guess the main problem is that there are just fundamental design differences, in both track and vehicle, between long range, high-speed travel to a few remote lots, and what is needed in town. Some compromise seems unavoidable.

Andrew F said...

Hi Dan. I took some time to think about this. As an aside, I believe blogger has a feature that shows 'Recent Comments' in the sidebar, making it easier to notice comments on older threads, that might be handy for both you and your readers.

Forgive me for quoting you in my response. I find it helps to clarify.

"However, I think people hate taking longer than the driving time even more; I would stay with my guideline that the system must represent an overall time savings to attract customers."

My point is that once you get people out of their cars, people are quite willing to trade travel time for fewer transfers. Imagine you take a trip where you spend 5 minutes in a PRT to a GRT line, exit the vehicle, move to the (waiting) GRT vehicle, spend 20 minutes travelling along the GRT route until are close to your ultimate destination on the other side of town. Once there, you debark again, board one of the waiting PRT vehicles that has swarmed this GRT vehicle and continue on to your destination, 5 minutes away. Your overall trip is 30 minutes travelling, and perhaps a minute for each transfer for a total of 32 (assuming no waiting for vehicles).

Imagine you could take the same trip, at somewhat lower speed on a PRT vehicle with zero transfers, taking 45 minutes. I think you might be strongly inclined to take the one seat trip, even though it takes an additional 13 minutes, since transferring is a hassle, you might have to transfer in unpleasant weather (I think of this because I live in Canada), and you will have less privacy on the GRT portion. Travel time on transit is not directly comparable to travel time while driving, since transit allows one to do productive work or leisure while in transit, which is difficult/illegal while driving.

"One PRT for one fare. No waiting. You seem to assume this isn’t possible."

That was not my intention. Even if there is no waiting, there is an aversion to transferring (it's not 'frictionless').

"If you are referring to transfers to neighborhood buses or light rail, well, that’s another matter."

In this case, GRT might get rid of waiting, but there is still the hassle of physically moving from one vehicle to another.

"Privacy is of course preferred, yet experience shows that people will use a system that saves them time and money."

Privacy is a big appeal of car transportation. It's also one of the main selling points of PRT (no chance of getting stuck with a smelly drunk or crazy person).

Andrew F said...

"The case that GRT is not the preferred alternative over PRT is unquestionable, assuming sufficient track space, energy/speed, vehicles, stations, etc. But I am thinking mostly about trips to/from “park&rides” (public parking lots close to the freeway, well outside of town) on track that is designed for high speed. The GRTs might have speeds 3 times faster than PRTs. The alternative would be to funnel a million pokey little pods onto this long straight run where they would only beat freeway speeds during rush hour. I am thinking GRT exclusively for freeway median use, not suburbia."

That's fair, but then you still have the issue of a loss of privacy and one to two more transfers. For some people, GRT's speed might overcome these downfalls, but there will certainly be those who are willing to spend more time (and even money) to take the PRT for a transferless ride, provided the system gives them this option.

"There are a couple of possible “work-arounds” that come to mind. One is the possibility of having all high-speed be accomplished through “helper” or “engine” bogies. Because the PRT motors would run faster based on frequency, not voltage, (until they overheat from being overworked) helper bogies become possible. This is the idea I brought up in #56. Another alternative is to just build all of the PRT vehicles with oversized engines. This is not as crazy as it sounds, because unlike gasoline engines, electrical motors tend to draw only the power that they can put to work. Sure, it’s a waste to carry around the bigger motor if its power is not being used, but the electric bill won’t be much higher on account of it. It does jack up fleet costs though, and it does put a lower top limit on speed."

I agree. There are other alternatives, and some of them might be able to coexist.

"I guess the main problem is that there are just fundamental design differences, in both track and vehicle, between long range, high-speed travel to a few remote lots, and what is needed in town. Some compromise seems unavoidable."

And that is where my point about the sweet-spot (or optimal) trade-off between speed and cost/aesthetics comes in. I think that the ideal should be to limit the need (that is, make PRT no-worse or not-much-worse than GRT) for GRT to the longest 5% or so of trips. If every, or too many trips require 2 transfers (PRT-GRT, then GRT-PRT) plus walking to and from stations, this will hold back adoption and many people who might have sold their cars will opt to keep them. They key to adoption is that people will sell their cars. They have such high fixed cost, that once it because worthwhile to buy the car and insure it, the variable cost of fuel and maintenance is quite low and hard for PRT or any transit to compete with.

I apologize for the long, pedantic post.

cmfseattle said...

so ULTra updated their site like whoa http://www.atsltd.co.uk/prt/implementation/costs/
sorta answers a few of our questions re: starter network size.
lots of good info.

Dan said...

Dan the blogger, after a long absence, responds to afransen, cmfseattle.
I guess my principle case, afransen, is that every city is different. Houston, for example, has 20 lane (if you include feeders and HOV lanes) freeways extending over twenty miles from town. There are already “park and ride” lots in place along this corridor, but the buses from them don’t get people close enough to their destinations. There is already a very strong case for 6 to 12 passenger GRT here, because the highly commercial area, (where the work is) is pretty big, at least 50 square miles. Therefore getting within a couple of miles is half the battle. The problem is that most people don’t move around this 50 square mile area. They come, they work, they go home to the suburbs. It’s really a tough call to figure out where to start a PRT network, because the city developed along its freeways. It lends itself to (longer) linear, not (shorter) grid solutions. If I had the money to build, say, 100 miles of track I would probably split it 50-50 between GRT along freeways and PRT serving arterial boulevards. That would be, for instance, a 25 mile “X” and in turn, a hundred miles of PRT track. That proportion, I believe, would get the most people closest to their destinations fastest. It would also take the most miles off of the city’s total automobile travel. Of course in Houston 100 miles of track wouldn’t scratch the surface. Our tallest building is 6 miles from downtown…

So anyway my point is that sprawling flat city with no zoning is totally different than one, for example, that is surrounded on 3 sides by water. Additionally there is the question of what is needed most and first. PRT is not nearly as advantageous with very limited routing. All systems start with very limited routing. Therefore PRT might be less advantageous as a starter system, GRT less advantageous as a mature system. The sweet spot is a moving target.

Hi cmf- Interesting link.. So the top speed is 25 mph. And for every pound of capacity there are nearly two of vehicle weight. Hmmm. I was interested in the relatively long 10km range. It would seem it would be better to figure out a way to recharge more frequently, lose some battery weight and get a bit more performance. Who wants to go that far at that speed? The battery weight also means elevated track has to be more massive, expensive. Oh well, Rome was not built in a day.

Bruce Attah said...

I see the logic in having two different vehicle types, one for high speed and one for low, but but I wonder if it is compelling.

One possible problem I see is that it might involve a false economy: you get "right-sized" motors in your vehicles, but you might have to operate a larger total number of vehicles in order to achieve this.

Another issue is that I'm really not sure if the longer, faster routes need to be GRT, rather than PRT, even if they need a different type of vehicle. What's the alleged advantage of GRT? Is it line capacity? cost-per-seat? something else? How big is this advantage, and how is this advantage obtained? It's not quantified.

If it's a cost advantage, is it really big enough to outweigh potential disadvantages to the user in privacy and to the operator in guideway costs? Is it big enough that most/all passengers, given the choice, would take the GRT rather than pay more to travel by PRT? Will the advantage be lost during off-peak hours?

Quick change of subject: ULTra's 25 mph operating speed. It sounds pretty leisurely, but it's quicker than the norm for urban public transport, and quicker than a lot of urban road traffic. 12 mph is the average speed of car traffic in London, and at rush hour this falls to 7 mph. On a 10 mile PRT network, a typical trip would probably not be longer than 5 miles, which amounts to a 12 minutes at 25 mph. To achieve that kind of trip time in inner/central London in safe, stress-free, uncrowded comfort would be a dream come true.

Dan said...

I like to bring these issues up because I am still trying to achieve an open source design standard for track. I agree with what you say, or at least somewhat share your concerns. I think it is my role to be a bit of a “devil’s advocate” in this blog. This is true not just to provoke discussion but also as a designer’s philosophy. One has to push the limits of what is practical to tell what those limits really are. I don’t want to design a track that can’t adapt to potentially good business models.
In Houston, A very fast GRT would be a big hit along the freeways leading into town with average commutes being about 15 to 25 miles, if the GRT had access to an in-town PRT network, so most passengers could be dropped at or near their work. The key is the remote parking lots way out of town that aggregate passengers and pair them with destinations. Another factor is that the GRT would be traveling empty when outbound, a different dynamic than an “in town” PRT network. There are questions about how available PRTs would be, that far out, or how much they would congest the system as they work their way toward the track out of town. Where I live we already have a fairly successful “Park and Ride” bus system with designated lanes for high occupancy vehicles. The problem is that they don’t have dynamic routing based on the particular passengers. In a kiosk based system of smaller GRT vehicles, the “sign on the bus” would reflect actual requests in real time, and the small size would ensure faster boarding and departure. True, this is not real “point-to-point”, but, with luck, it there might be 20% tranfers and only 30 % would have to wait through a drop off. The rest would get there directly. At any rate I guess every city and even every route would be different, so, again, it’s about flexibility.
London sounds like Boston. Very challanging to drive. I guess a key question is this. How many of the people caught in gridlock have both their origin and destination within that gridlock? Will a “close-in” PRT system really help? Or should the system intercept cars before they get close?

Ryan Baker said...

Definitely right that there are different levels of service required for every area. But you know, the space needed for a person doesn't vary based upon where they are coming from.

Now what will also vary is individuals requirements. Disabled are the most obvious (and legally protected) example, but it does go a little deeper than that.

Thinking of those two factors, the logical conclusion is to see what can be done to allow them to vary independently. Looking at most PRT designs, they are most of the way there already. Flashback to:

http://openprtspecs.blogspot.com/2009_10_01_archive.html

Maybe there is some real gold here. Any reason this can't be the general purpose design? That is, any reason why the pods can't be held to the track by a motorless connection and pushed by an engine?

That idea actually makes the Schweeb idea a bit more practical. You made the assumption that motorizing the Schweeb would require a stiffer frame, etc, but it doesn't does it? The motor is suspended from the track, not from the pods. If someone's cycling without the engine then no weight, if they need/want some help then it's still the same pod and all you add is the engine.

I'll also add, I do think that transfers of the type you suggest are an option and I think GRT via transfer has sensibilities that go beyond speed. In the right situations it's more efficient. But those situations aren't totally consistent so any design should make sure that GRT track remains compatible with PRT vehicles. If you can separate the engines and thus allow GRT and PRT to share the track at high speed then afransen can get what he'd want, and those looking to save a dollar or two can take GRT during rush periods. Outside of rush periods I'd think GRT should stop running, because if it has to run near empty, you're better off discounting the PRTs and encouraging their usage.