Wednesday, February 4, 2015

165> Tips for Designing PRT

Recently I have been corresponding with a new PRT developer, and I was struck by how similar his logic is to the many who have come before.  He initiated his communications by saying that the first step to a successful solution should be a fresh look at the problem.  He then stated his list of design criteria, tying back each decision to the problems that it would solve.  Before very long, he had narrowed his design to something very similar to the SMART platform, yet quite different in many important ways, and I found myself defending the specifics of my approach.  I thought I would share.

Design differences don’t just come from differing views on the practicality of one mechanical solution over another, but also from the original assessment of the problem itself.  PRT is the would-be solution to a host of different grievances.  To some, it is primarily an energy/climate saver.  To others the main emphasis is downtown mobility, while some would try to solve the problem of gridlock on our highways.  The various PRT solutions that are out there, it seems to me, are all pretty well crafted for their intended purposes.  What worries me is they stick to their respective missions a little too tightly. 

Current PRT offerings are not versatile enough to harvest sufficient passengers from the varied, real-world cityscapes to be economically viable until they have reached a critical mass of coverage.  The cost to achieve this break-even coverage is simply larger than what cities can afford or, at least, are willing to risk their careers over.  Here is a brief synopsis of the problem:

For PRT (or any other public transportation system) to be a serious contender, it must pass two tests. First, it must not lose money.  Second, it must perform its function better than alternative choices. When PRT can clearly meet these two criteria, it will be adopted, pure and simple.  Starting with the first test, let’s imagine the simplest, cheapest system, a simple loop between two stations.  PRT costs money, and that only gets returned through fares.  The separate costs of track, stations and vehicles must each be amortized though their respective lives and these numbers (along with operating costs) tell us how many passengers need to use those assets daily to make the system break even.  Here’s a hint.  It takes lots and lots of passengers to make the payments!  With two stations, the only people who will use the system are those who need to go to one of those two destinations. While you can buy fewer vehicles, and you can make smaller stations, you can’t reduce the track costs to where only the occasional pod will pay the bills, so that component (the track) bleeds money from the stations and vehicles, leading to its logical elimination.  Now we have another shuttle bus on the streets.

As for the second criteria, such a simple route lends itself better to a larger, scheduled vehicle anyway.  Since everyone is going to the same place, why not ride together?  Even a few stations between would not represent much added inconvenience.
Above is a “heat map” of 5 layouts, representing increasing numbers of loops and stations.  The warmer colors represent greater traffic.  These are not representations of city blocks nor would it ever be possible to have such evenly spaced stations.  Passengers are only where you find them, and then only some of the time.  What is shown, though, is how stations are interdependent for passenger traffic.  Each station is both source of new passengers and a destination for others who would not otherwise ride. In a full-sized system, (not shown) the vast majority would be hot yellows and oranges, and only the outer perimeter would be the unprofitable purple or blue.  PRT clearly can’t start too small.

Needing more closely associated network nodes to be effective is not unique to PRT.  The same heat map and lack of profitability at the outer edges can be seen with a bus system or parcel delivery, for example.  Yet networks are not necessarily all equal when it comes to the ability to grow “from seed.” The key is making each part of the network viable in its own right.  No doubt the builders of the very first stretch of railroad already had a pretty good idea that it would be profitable on its own, and each subsequent branch surely met this test as well.
When it comes to priorities for designing the perfect PRT system, this business of needing a whole network from the start is a whopper.  It is the elephant in the room.  Yet with proper design I believe a formula for growth can be found.  There’s more.  Next time you are somewhere that you think would benefit from a PRT system, try identifying the next closest station sites, and estimate how many passengers each such station would attract.  If your experience is anything like mine, you will soon see that good sites are not at all easy to come by, and those with lots of foot traffic are usually on private land or are squeezed by roads that have been widened as far as limited easements will allow. At best they are in patches, separated by areas that don’t need PRT at all.  I pity the poor city planner tasked with finding station sites every half-mile that have enough pedestrian traffic to warrant an elevator equipped station.

The mobility problems faced by the urban and suburban populations differ greatly, and often PRT designers are biased by their own experience.  There are differences due to geography (such as climate) and even city-to-city topographical differences, such as water frontage or hills.  Political differences are important because they effect funding and attitudes toward public works in general. There are even historical differences.  One that I am often reminded of is (then) President Eisenhower’s “cold war” strategy of initiating of the US interstate highway system, resulting in urban neighborhood segmentation and the suburbanization of an entire nation.  Now, 1 out of 4 miles driven in the US is on an interstate, often at maddeningly slow speeds! The point is that these are among the factors that individualize a city’s needs, even as the economics of starting a PRT company push toward a more “cookie cutter” approach, and therein lies the rub.

This is where “SMART” comes in.  “Suspended Multi-axis Automated Rail Transport” is not a PRT system per se, but a development platform.  Any system that can move vehicles in 3D can also work in 2D.  Any system that can travel at 90 mph can also travel at 30.  If it can turn in a 6’ radius, it can certainly manage any street corner.  The idea to avoid “baking in” needless limitations.  The SMART vehicles are very capable and full featured because someday they might be expected to be that way. After all, people will eventually want, from PRT, the same qualities they would expect from a manually driven taxi service, and investors, equally, will prefer a fast, efficient and comfortable fleet to maximize returns.  In the meantime, there are tough choices in store.  SMART, with all of its “bells and whistles” is, admittedly, an extremely (if not prohibitively) expensive way to start.  Yet its basic architecture, especially the track, addresses the above problems squarely by being uniquely versatile.   
With PRT requiring a whole network to operate profitably, and with the vast differences between different cityscapes, the problems that predicate a PRT designer’s job are immense and varied.  Some examples include how to put dirt-cheap stations everywhere and anywhere; how to take the shortest route between lucrative sites instead of being detoured by worried landowners; how to piggy-back other services, like carrying utilities or street lighting; How to give local planners more and better options; How to get property owners to allow stations on their parking lots and in their lease spaces; How to attract the most riders without cutting fares, how to encourage night traffic and freight, how to minimize operating costs, and so forth.  

Why PRT has not entered the mainstream is not simply a matter of suspended vs. supported design, linear vs. rotary motors, or third rail vs. battery power.  It is the big structural issues regarding network size, versatility, and the business model, and translating these overarching issues that into physical form should be “job one” for the PRT designer. 

Saturday, January 17, 2015

164> Papa Bear, Mama Bear, Baby Bear

As promised in my last post, I have modified the SMART PRT (Suspended Multi-axis Automated Rail Transport – Personal Rapid Transit) design yet again, this time to accommodate featherweight “pods” and enclose the bottom of the track, as well as fine-tuning certain proportions.  
There is also a secondary rationale for considering ultra-light PRT systems, which is the matter of parcel and light freight delivery. Companies such as Amazon are clearly putting heavy emphasis on speedy delivery these days, and the ability to integrate a steady stream of varying vehicles is, unlike days past, essentially a given. Couple this will the fact that such track would be many times cheaper, and a very good case can be made for exploring the matter further. 
In response to this challenge I submit the pictured multi-sized approach. In the top picture, the comparatively tiny profile of the “Baby Bear” track can be seen, which does not even reach the man’s knee. 
Before going any further, though – a disclaimer: The pictures are to illustrate component geometry only. The track is shown as if it can be simply extruded, which it can’t, and I just thickened the top of the larger sizes to show where ribs would most probably go.
Also, I’m sure that there are some newer readers who may have never considered the matter of switching suspended PRT, and who may be confused by the various end-views. To them, I offer this more complete, yet still simplified, end view and explanation.

To switch suspended PRT, the part of the vehicle that rides inside of the track (the bogie) must grip one side of the track or the other, so that when the track routing diverges, the vehicle’s drive unit (bogie) does not fall out of the track’s ever widening slot. In this end-view, the pair of steering-guide wheels on the right “clamp” into rolling engagement with the track, while the left ones spread apart to release that side. In other words the bogie, shown in blue, can be fully supported by either side of the track and so can follow either as they split apart into different routes. The weight forces of the hanging passenger cabin are shown by the three red arrows. 

It is also noteworthy that certain parts of the track are eliminated where there is no switching. In this picture, the steering guide wheels are not quite contacting anything. Side-to-side guidance is maintained by the large, counter-rotating horizontal wheels; to do otherwise would simply add wear to the smaller wheels for no reason. The actual contact rails are only at switch points and taper into contact gradually for a smooth transition. Also, particularly observant, long-time readers will note that some track parts will need to be modified or removed for tight radius turns, both side-to-side and up-and-down.
Now, back to the top pictures. In these two illustrations, BTW, the steering guide wheels are engaged as if they were approaching a “Y” in the track.

The “baby” track is sized for speeds below 30 mph, (48 km/h) with weights below 500 lbs. including vehicle and passenger/cargo. It has 8” (215 mm) solid drive wheels. (It could handle more with steel wheels and very frequent supports but this would be inconsistent with most applications) The track width is 13.5”. (343 mm) 
To the person’s top-right are the “baby” wheels within a larger (mama-bear) track. Directly below are the wheels for the “mama-bear” vehicle within the same, matching track. This size (30” or 762mm wide) is consistent with the 3.5 person vehicles that have been shown in earlier posts, weighing in at about 1300 lbs., fully loaded. The performance would be generally auto-like, with quick enough acceleration to allow fairly short on-ramps, (an often overlooked design consideration!) and speeds approaching 50 mph. (80 km/h) It is not impossible that faster speeds would be practical…Only testing will tell. 

The second picture above  illustrates (left to right, clockwise) the “mama bear” bogie in the “mama bear” track,  the “mama-bear” bogie in the “Papa bear” track, and the “Papa bear” bogie in the “Papa bear” track. This larger wheel and track would be for still higher speeds and/or loads. Not much bigger than the “Mama Bear” combo, (35” or 890mm wide) the steering guide wheels get an extra two inches of diameter while drive wheels become auto-sized. The guide wheels of heavy vehicles can also be made of longer-lasting, harder polyurethane formulation or even steel, as sound is less of an issue because such routing would typically be along highways, more than fronting homes of businesses. Being designed for long distances and for larger (GRT) shuttle vehicles, this larger sized track would not universally assume full 3D routing and would be encased in a heavier structural elements. High speed sections would require some additional interior space to allow air to slip around the speeding bogie. Some very large cities might opt for the “Papa Bear” track throughout, to accommodate faster personal vehicles designed for typically long commutes. It is, after all, only 5 inches wider and height is more a function of span/vs load than wheel size.  I am somewhat dubious about the utility of making this largest track universally “baby bear” compatible, and one approach would be to design such track to be easily convertible for this purpose, rather than to include this feature initially, like the pictures shown above. This is primarily because of the speed differences, since any quiet and smooth running 8” wheel couldn’t tolerate continuous, heavily loaded, high speed use for very long, and so must be confined areas that are appropriate for low speeds anyway. For lighter parcel delivery, however, in, say, the 25 kg load range or less, it is a different story. Here small vehicles could almost fill the track spaces between large GRT vehicles, not slow them down at all, and still be nearly silent when they get into neighborhoods… without needing new wheels every thousand miles. 

What is regrettable is that the Mama Bear track isn’t smaller and cheaper – more of a mid-point between the larger and smaller sizes. Unfortunately once you can physically over-pack a vehicle with weight, (as is unpreventable with the interior cabin space required by disability laws) you can theoretically load each vehicle just as much, and as long as there is the remote possibility of a sudden stopping of traffic, where suddenly such loaded vehicles are bumper to bumper, applying all of that weight to a span, or even all applying emergency brakes at once on a curve, (thus pulling the track forward and sideways and down) … there is the need for some pretty beefy track, at least that is how it will be regulated. 

Ironically, the baby bear track can be small because it is small… If track and vehicles are cheap enough, it becomes easy to pay for, even without heavy ridership, which permits longer headways and slower speeds, so everything can be lighter-duty. Of course the “baby bear” can’t exist as a transit system alone, since it does not meet the criteria mentioned above. The Mama and Papa Bear systems need to really pack the tracks and run as fast as possible to pay their way, especially until the business model proves itself and costs begin to drop. This kind of traffic is best promoted by providing a certain degree of comfort…And comfort adds…you guessed it… weight. In PRT, there is no such thing as a free bowl of porridge!

Next step - Adapt this wheel geometry to actual stock steel, practical fabrication methods and truss design, as well as to finally address the long-neglected issue of “third rail” placement.