Visualization of capacity on PRT networks

[Chris Xithalis]

Regarding PRT capacity visualization:

A few months ago I received some comments/requests regarding the PRT simulator on how I could present a more general picture of the traffic on the network as a whole. The solution I came up with was drawing each vehicle as a small square and draw them against black background for bigger contrast. I found the result quite satifactory and so did the guy who made the comments. Some screenshots (black/normal map background):

Of course viewing that in motion is much more interesting. For anyone interested in doing so:

Download simulator:  http://students.ceid.upatras.gr/~xithalis/files/simulator.zip

unzip the files

double click run.bat to start the program (you must have java installed. If you don't you can get it at http://www.java.com/en/download/manual.jsp )

file->load network->larissa.her

action->design done

demand->demand program->larissa_program.txt

colors->min vehicle size 2X2

colors->do not show lines

city map->black background

simulation speed->20X

press 'S' to start/stop simulation.

move mouse with right mouse button pressed to move around the map

use the mouse wheel to zoom in and zoom out

Blue dots are empty vehicles, red dots are occupied. One can also zoom in at a station. All stations have 5 berths. Passengers are not shown , instead a number is shown at each station denoting number of passengers waiting in a queue. Other transit systems (e.g. an unloading LRT) are not shown but one can edit the rate of people arriving at a station to simulate a similar event.

Chris Xithalis

[Jerry Schneider]

At 01:58 PM 9/27/2010, you wrote:
>Regarding PRT capacity visualization:
>A few months ago I received some comments/requests regarding the PRT
>simulator on how I could present a more general picture of the
>traffic on the network as a whole. The solution I came up with was
>drawing each vehicle as a small square and draw them against black
>background for bigger contrast. I found the result quite satifactory
>and so did the guy who made the comments. Some screenshots
>(black/normal map background):
>

>[]

>
>
>
>Of course viewing that in motion is much more interesting. For
>anyone interested in doing so:
>Download
>simulator:

><http://students.ceid.upatras.gr/~xithalis/files/simulator.zip>http://students.ceid.upatras.gr/~xithalis/files/simulator.zip

>unzip the files
>double click run.bat to start the program (you must have java
>installed. If you don't you can get it at

><http://www.java.com/en/download/manual.jsp>http://www.java.com/en/download/manual.jsp

>)
>file->load network->larissa.her
>action->design done
>demand->demand program->larissa_program.txt
>colors->min vehicle size 2X2
>colors->do not show lines
>city map->black background
>simulation speed->20X
>press 'S' to start/stop simulation.
>move mouse with right mouse button pressed to move around the map
>use the mouse wheel to zoom in and zoom out
>
>Blue dots are empty vehicles, red dots are occupied. One can also
>zoom in at a station. All stations have 5 berths. Passengers are not
>shown , instead a number is shown at each station denoting number of
>passengers waiting in a queue. Other transit systems (e.g. an
>unloading LRT) are not shown but one can edit the rate of people
>arriving at a station to simulate a similar event.

I think this has some interesting possibilities. Ideally, a viewer
might be persuaded if a they first understood the operational
settings (assumptions) for the PRT system and then could view it's
operation over a set period of time for a given quantity of demand.
For example, starting from a vacant network, we will show you how we
can serve the trip requirements of X patrons, over a Y time period.
Not all trips would be "finished" so X would be some proportion of
total trip demand. This result would not be very different from the
stats calculated by Andreasson's simulation other than your would
have to watch all the little dots moving around for some period of
time. Would this be more persuasive? Worth the effort? Perhaps one
could show with a bar chart, the total trips generated at all
stations, over time, and those that are "completed" before the end of
the simulated time period. This would enable you to state that this
network can service K trips per M time period. Still the comparison
with a comparable LRT/Bus service would remain as a very difficult problem.

Andreasson's results for the Gothenburg PRT network: 27,191
passengers departed during the 30 minutes simulation period, using a
716 km network with 654 stations and 12,795 cabs. Lots more stats at:
http://faculty.washington.edu/jbs/itrans/ingsim.htm

What one needs is a similar calculation for an existing or planned
bus/LRT network for an spatially equivalent demand
using an appropriate number of vehicles, stations, and so on. Are
numerical comparisons sufficient or is visual confirmation required,
desired, useful or what?

>Chris Xithalis
>
>--
>You received this message because you are subscribed to the Google
>Groups "transport-innovators" group.
>To post to this group, send email to transport-...@googlegroups.com.
>To unsubscribe from this group, send email to
>transport-innova...@googlegroups.com.
>For more options, visit this group at
>http://groups.google.com/group/transport-innovators?hl=en.

- Jerry Schneider -
Innovative Transportation Technologies
http://faculty.washington.edu/jbs/itrans

[eph]

Very cool. Is it possible to change the headway settings, passenger
per vehicle count, vehicle size etc... to simulate other systems?

Thanks,
F.

On Sep 27, 4:58 pm, Chris Xithalis <xitha...@gmail.com> wrote:
> Regarding PRT capacity visualization:
> A few months ago I received some comments/requests regarding the PRT
> simulator on how I could present a more general picture of the traffic on
> the network as a whole. The solution I came up with was drawing each vehicle
> as a small square and draw them against black background for bigger
> contrast. I found the result quite satifactory and so did the guy who made
> the comments. Some screenshots (black/normal map background):
>

[Robert Williams]

Perhaps the most used figure of merit when comparing existing transit
systems is "boardings per day", without regard to distance traveled or
direction. Light rail systems do well to get 50,000, with heavy rail in
the 200 to 300,000 range. The Andreasson study you cited had 27,000 per
30 min, which should produce over 500,000 per day. I should think
numbers like this would attract the public's attention.

It would be interesting to see what the PRT numbers are for simulations
of a given system when varying vehicle and system loadings. Can an
unstressed system still produce big numbers? When comparing with other
modes, perhaps looking at boardings/day/construction and/or operating
costs would be informative. I would guess the Gothenburg system would
cost $4-8 billion. Our 16 mile BART extension is presently priced at $5
billion, and LRT extensions are costing in the billions.

So just change "departures" to "boardings" and see what happens.

Bob Williams

[Chris Xithalis]

Some settings are editable.

The network topology is editable (lines,stations,depots).

Headway can be set between 0.3 and 2.0 seconds.

Speed can be set between 10 and 25 meters/second. Note however that slot size (= headway * speed) cannot be larger than 30 meters.

There is more information in the manual.

Different station sizes (other than 5 berths) could be possible but not implemented yet.Depots don't have a specific size.

Note that I tried to make simulator as versatile as possible in order to simulate as many systems as possible. However when some parameters get outside some certain ranges, problems that I can't cope with arise.

Example: The slot size is big e.g.100 meters  (that could be 4 seconds headway and 25m/sec speed). 2 vehicles running on lines A,B respectively approach a merging point simultaneously. The system must slow down one of them enough so that when the second one passes through the merging point the first one is 100 meters ahead. This is impossible to achieve on an arbitrary network since there is no guarantee that lines A and B are long enough to accomodate these maneuvres while keeping acceleration/deceleration to an acceptable level.

Chris Xithalis

[Chris Xithalis]

Actually this is the purpose of the simulator. To achieve this the user can create a "scenario":

-Design a network on a map

-Set the demand i.e. the number of passengers that want to use the system

-Simulate and see the results

Setting the demand involves creating a "demand program" which describes the volume of the passengers (passengers are counted in groups of passengers travelling together by choice, ride sharing is not supported), their origins and destinations. This can change over time. There are 3 different methods to do this. The manual has related instructions. I have used them to create "morning rush towards CBD and evening rush towards suburbs" scenarios.

Presenting the results (i.e. network capacity) in an easy to understand way is a tough task. The most accurate presentation would be a vector from each station to each other. Unfortunately in a network with 200 stations that would be about 40000 vectors, an absolute mess.

One could perhaps consider larger areas instead of individual stations to produce something like this:

However this graph misses the level of fine detail that sets PRT apart from mass transit, it actually suggests that a few mass transit lines, with adequate capacity could do the same job!

These are some examples of results the program provides:

http://students.ceid.upatras.gr/~xithalis/patras_en.html

http://students.ceid.upatras.gr/~xithalis/larissa_en.html

http://students.ceid.upatras.gr/~xithalis/tram_hermes_en.html

http://students.ceid.upatras.gr/~xithalis/attiki_sparse_en.html

http://students.ceid.upatras.gr/~xithalis/amsterdam_en.html

It also provides tables with arrivals/departures at stations, waiting times etc, which can be copy-pasted to an excel spreadsheet and produce additional graphs.

Perhaps I should add some extra functionality e.g. to graphically present the volume of passengers each station received during simulation using variable sized circles (like the above picture). I'll see into that for a possible future version.

Chris Xithalis

[Richard Gronning]

I've always had an idea of a scale model in mind to sell the idea of PRT in, let's say, the Mall of America, or some such major gathering spot.

Now, after seeing Chris' Simulator, I have a similar idea, but I don't know if it would work. Here's the idea;

• Project the simulator on a large screen. (MOA has such a screen.)
• Have "stations," a simulated station that people could "board;" more specifically program a vehicle.
• The simulated boarder could pick some sort of way to identify his/her vehicle at a particular birth-station.
  
• There would be 5 simulation points to access the system.
• Then the "passenger-programmer" would take a simulated ride.

How much would it take to do this?
Dick

[WALTER BREWER]

Great idea!

 

I assume the "simulated boarder" is an actual person experiencing how the system works.

In a specific city popular activity centers, perhaps even local homes, could be used to help realism.

Also have person enter a mockup pod as if he/she were starting a trip.

 

I'm near the bottom of the list of knowing how simulators being discussed work. But isn't the process exactly what a real system would need to route pods to destinations?

 

Walt Brewer

[Richard Gronning]

Comments inserted below....

On 9/28/2010 8:44 AM, WALTER BREWER wrote:

Great idea!

 

I assume the "simulated boarder" is an actual person experiencing how the system works.

Exactly! He/she boards at a specific spot, selects a destination, and watches "his/her" vehicle merge and travel to the selected destination.

In a specific city popular activity centers, perhaps even local homes, could be used to help realism.

I hadn't had that idea, but terrific! Once the system "visits" a specific city, then the map of the metropolitan area would have been developed. Then the "riders" could pull the system up at home and show their friends. Now that's Public Relations at its best.

Also have person enter a mockup pod as if he/she were starting a trip.

Hmmm... Might be a bit expensive...

[Jerry Schneider]

At 06:48 PM 9/27/2010, you wrote:
> Perhaps the most used figure of merit when comparing existing
> transit systems is "boardings per day", without regard to distance
> traveled or direction. Light rail systems do well to get 50,000,
> with heavy rail in the 200 to 300,000 range. The Andreasson study
> you cited had 27,000 per 30 min, which should produce over 500,000
> per day. I should think numbers like this would attract the
> public's attention.

The Gothenburg simulation time period was the am peak - so assuming
that the peak lasts all day is not reasonable.
And, it took over 12,000 vehicles to do it. One has to wonder what an
investor would make of that. Of course, maybe the same system "could
have served" a demand twice as large without suffering untenable wait
times at the stations. So,
one cannot say that 27,000 is the "capacity" of the system with any
assurance. But the other wicked problem is structuring a rail/bus
system that is "comparable" or designing a PRT network that is
comparable to an existing rail/bus system. Comparable in capital and
operating cost? Or using multiple criteria?

>It would be interesting to see what the PRT numbers are for
>simulations of a given system when varying vehicle and system
>loadings. Can an unstressed system still produce big numbers? When
>comparing with other modes, perhaps looking at
>boardings/day/construction and/or operating costs would be
>informative. I would guess the Gothenburg system would cost $4-8
>billion. Our 16 mile BART extension is presently priced at $5
>billion, and LRT extensions are costing in the billions.

One could keep increasing the loading of the PRT system until the
wait times at the stations exceed some arbitrary number of minutes
and call that the "capacity" of the system. Of course, the wait times
would be exceeded at some stations first, others later, so one has
to deal with an "area-wide average wait time" which is not too
satisfactory. And, there are a number of other operating variables
that one can fiddle with that will impact the results. And, then
there are the physical system design changes that will become obvious
that, if made, will change the results. But the basic problem is
still finding a way to make a credible comparison between a network
and a corridor-oriented transport service. It gets so messy that few people
can deal with it and so they revert to their basic belief that
"bigger is better".

>So just change "departures" to "boardings" and see what happens.

It seems to me that Andreasson's simulation is dealing with boardings
as the demand numbers are patrons which occupy the vehicles which
depart with an average occupancy on board.

[Roy Reynolds]

Two links of interest.  As an Orange County resident of 20+ years, it isn't like this has never been suggested to Disney.  In fact, one of our proposals was rejected by their legal department and returned. 

PRT would make enormous sense to Disney.  As I recall they have approximately 17 parking lots which must connect to the Maingate (entry to the two existing Parks, Disneyland and California Adventure), the one-third mile long shopping complex that separates the Parks, Disney's three existing hotels, about other major and minor 100 hotels in the surrounding area, the Anaheim Metrolink/Amtrac train station, two hospitals and dozens of off-site shopping venues and, of course, the third park when (not if) it's built (currently this extremely valuable property is used for their cast parking, connected with a fleet of shuttles to the Parks about 1/4-mile away).

Also of interest is the idea that a PRT-like animation was used in three separate scenes in "The Incredibles" -- an excellent, Oscar-winning film for which a sequel is suggested fairly often in the entertainment press.

http://thedisneyblog.com/2010/09/27/disneyland-to-bring-back-peoplemover/
http://ocresort.ocregister.com/2010/09/27/is-disneyland-bringing-back-people-mover/56864/

Roy Reynolds
PRT Strategies

[Ian Ford]

As shown by Starbucks and FedEx, markets don't work in the simplistic
manner of undercutting the price for the same product. Creating a PRT
system that is the "equivalent" service of a bus system, then asking
"does it have enough capacity" is backwards (as I think everyone here
understands). The main question should be, what is the cost or profit to
get X% of total travel. Because, that's a policy question. The question
of capacity is an engineering question. Cost-versus-% is a single
measure that can apply to any transit system. So the answer to the
capacity question is that there is no limit, but that the cost per
boarding goes down as the capacity goes up.

On 2010-09-28 10:32, Jerry Schneider wrote:
>
> One could keep increasing the loading of the PRT system until the wait
> times at the stations exceed some arbitrary number of minutes and call
> that the "capacity" of the system. Of course, the wait times would be
> exceeded at some stations first, others later, so one has to deal with
> an "area-wide average wait time" which is not too satisfactory. And,
> there are a number of other operating variables that one can fiddle with
> that will impact the results. And, then there are the physical system
> design changes that will become obvious that, if made, will change the
> results. But the basic problem is still finding a way to make a credible
> comparison between a network and a corridor-oriented transport service.
> It gets so messy that few people
> can deal with it and so they revert to their basic belief that "bigger
> is better".

--
Ian Ford i...@ianford.com 505.750.IAN8

[Roy Reynolds]

[Kirston Henderson]

>From: Ian Ford <i...@ianford.com>
>To: transport-...@googlegroups.com
>Subject: Re: [t-i] Visualization of capacity on PRT networks
>Date: 28, Sep, 2010, 12:22 PM

>
> As shown by Starbucks and FedEx, markets don't work in the simplistic
> manner of undercutting the price for the same product. Creating a PRT
> system that is the "equivalent" service of a bus system, then asking
> "does it have enough capacity" is backwards (as I think everyone here
> understands). The main question should be, what is the cost or profit to
> get X% of total travel. Because, that's a policy question. The question
> of capacity is an engineering question. Cost-versus-% is a single
> measure that can apply to any transit system. So the answer to the
> capacity question is that there is no limit, but that the cost per
> boarding goes down as the capacity goes up.

Ian.

I believe that you just posed the question that when answered provides
the reason that not many are seriously considering PRT systems. Of course,
if you asked the same question regarding LRT, the answer would be the same.
If if were not for Federal Government pushing and funding for LRT, there
would not be much of that either.

Kirston Henderson
MegaRail® Transportation Systems

[Kirston Henderson]

From: Roy Reynolds <roy.re...@prtstrategies.com>
To: Transport-Innovators <transport-...@googlegroups.com>
Subject: [t-i] Is Disneyland bringing back People Mover?
Date: 28, Sep, 2010, 12:32 PM

Two links of interest.  As an Orange County resident of 20+ years, it isn't like this has never been suggested to Disney.  In fact, one of our proposals was rejected by their legal department and returned.  

PRT would make enormous sense to Disney.  As I recall they have approximately 17 parking lots which must connect to the Maingate (entry to the two existing Parks, Disneyland and California Adventure), the one-third mile long shopping complex that separates the Parks, Disney's three existing hotels, about other major and minor 100 hotels in the surrounding area, the Anaheim Metrolink/Amtrac train station, two hospitals and dozens of off-site shopping venues and, of course, the third park when (not if) it's built (currently this extremely valuable property is used for their cast parking, connected with a fleet of shuttles to the Parks about 1/4-mile away).

    Back to the question just posed by Ian Ford, I believe that you will find that they consider that current fleet of shuttles the most cost/effective solution.

Kirston Henderson
MegaRail® Transportation Systems

----------

[Jerry Schneider]

At 10:22 AM 9/28/2010, you wrote:
>As shown by Starbucks and FedEx, markets don't work in the
>simplistic manner of undercutting the price for the same product.
>Creating a PRT system that is the "equivalent" service of a bus
>system, then asking "does it have enough capacity" is backwards (as
>I think everyone here understands). The main question should be,
>what is the cost or profit to get X% of total travel. Because,
>that's a policy question. The question of capacity is an engineering
>question. Cost-versus-% is a single measure that can apply to any
>transit system. So the answer to the capacity question is that there
>is no limit, but that the cost per boarding goes down as the capacity goes up.

I think that designing a PRT network that can be used to generate a
forecast of a desired market share amount and that provides the owner
with a desired ROI (perhaps with a subsidy) is essentially a design
engineering, not a policy, task. Cost (capital plus operating) versus
market share (two forecasts) are required to produce the numbers to
use. If one needs to compare PRT project A with a bus system project
B, then one could compare forecasts of market share and subsidy or
profit required and perhaps that would be sufficient for investors,
public or private, to make an "rational" decision. Isn't this just
the usual cost/benefit ratio approach?

Clearly, the more extensive and costly the PRT network is, the larger
market share it should be able to capture - so long as
the network design is well-matched to urban form and the trip making
desires of the spatially distributed demand. At some point the law
of diminishing returns might result in cost outstripping market share
increases, especially in lower density areas where automobility is the norm.

The above forecasts (PRT and Bus) would need to be capable of
indicating what modes would lose market share, unless one assumes
that the total trip-making market is not limited and that the demand
for transport service is essentially limitless.

>On 2010-09-28 10:32, Jerry Schneider wrote:
>>
>>One could keep increasing the loading of the PRT system until the wait
>>times at the stations exceed some arbitrary number of minutes and call
>>that the "capacity" of the system. Of course, the wait times would be
>>exceeded at some stations first, others later, so one has to deal with
>>an "area-wide average wait time" which is not too satisfactory. And,
>>there are a number of other operating variables that one can fiddle with
>>that will impact the results. And, then there are the physical system
>>design changes that will become obvious that, if made, will change the
>>results. But the basic problem is still finding a way to make a credible
>>comparison between a network and a corridor-oriented transport service.
>>It gets so messy that few people
>>can deal with it and so they revert to their basic belief that "bigger
>>is better".
>
>--
>Ian Ford i...@ianford.com 505.750.IAN8
>

[Jerry Schneider]

It's hard to argue that the Federal Government (FTA in the US DOT) is "pushing"
LRT. The facts are that they are overwhelmed with LRT project
proposals from cities
which represent far more money than they have available. Their main task is
evaluating these proposals an selecting a few that they consider
"worthy" or which
have strong political support that they dare not thwart (e.g. the
extension of the METRO
line to Dulles Airport). If they are pushing anything these days, it
is (and has been) BRT and
Streetcars. The FTA's job is to distribute the money that Congress
has appropriated for
transit projects. The demand for federal matching funds is much
greater than the supply.

I do agree that there would be far less demand for costly LRT
projects if the federal cost-share funds were not available.
However, given that the "public" and APTA keep saying "we want more
public transit" the Congress is unlikely
to change the game much, if at all. If PRT were to become a candidate
technology, it would be interesting to
see how competitive it would be - against BRT and Streetcars, primarily.

[WALTER BREWER]

So here I am again guys!

The USA military has ways to make decisions like this quantatatively.
(Includes sometimes vidualizations, so this is not a new thread though its
already off that track.)

The responsibility rests with the UnderSecretary for R &D.

He/she has analytical resources, most through organizations familiar with
the decision process for military needs compared to status of new
technologies.

It is also impacted by the lobbies and industrial orgs who "have the best
answer" already.
And of couse the political impact.

Not easy. Not alwas exact optimum, but it muddles through pretty well.

USDOT, (repeat,repeat, repeat), does not have such..

As expreseed a week or so ago I had hoped the revisions in the $30 billion
transportation stimulus package, inc. the "Bank" etc would generate somethin
similar.

Later statements dash my hopes significantly.

Agreed the :customer" in the military case is different, than hundreds of
communities. But there are some commanalities helpful to making chices that
can be determined.

Walt Brewer

----- Original Message -----
From: "Jerry Schneider" <j...@peak.org>
To: <transport-...@googlegroups.com>

[Robert Williams]

I was actually proposing something much more simple than a one-on-one
full mode comparison. My local light rail system is 42 miles long and
has 37,000 boardings per day. Today the system would cost somewhere
north of $2 billion. Simulations of a new or existing PRT network
costing in the same ballpark (assume some cost/mile) could be done for
system loadings of, say, 25-50%. Calculate the boardings/day (this
information may already be lying around). I don't know how it would
compare with the light rail, but would like to find out.

No, this would not be a direct mode comparison. The two systems would
be serving entirely different purposes (although throwing in the county
bus system makes them similar). Also, it is comparing assumptions to
reality. But "boardings" is a metric transit agencies use and people
understand. If it can be shown that a realistic PRT system can match or
beat other modes in this important metric, then I think it would be
significant and at least a useful talking point.

Yes, departures and boardings are the same thing. I was just suggesting
using the industry term in case the industry didn't understand.

[Jerry Schneider]

At 01:44 AM 9/29/2010, you wrote:
> I was actually proposing something much more simple than a
> one-on-one full mode comparison. My local light rail system is 42
> miles long and has 37,000 boardings per day. Today the system
> would cost somewhere north of $2 billion. Simulations of a new or
> existing PRT network costing in the same ballpark (assume some
> cost/mile) could be done for system loadings of, say,
> 25-50%. Calculate the boardings/day (this information may already
> be lying around). I don't know how it would compare with the light
> rail, but would like to find out.

The most recent PRT network simulation, done commercially and
published, conducted by an unknown transportation consultant for an
undisclosed client and city is available for your review at my
prtquick.htm webpage. It's also a bit unclear as to what operating
assumptions were used or how the demand data were generated but ...
it is extensive and detailed.

There is also a page at another website that shows what kind of a PRT
network one could build for the same amount of money that would be
needed for LRT or monorail. See:
http://kinetic.seattle.wa.us/nxtlevel/prt/xcomp.html (especially Part II)

Boardings are used by the industry but are somewhat misleading, since
many transit riders board twice (coming and going) and so the number
of "different" people served is about half of what might be
understood by the public - good for transit agency PR, maybe not so
good for public understanding. Newspaper reporters never mention this
in their articles.

>No, this would not be a direct mode comparison. The two systems
>would be serving entirely different purposes (although throwing in
>the county bus system makes them similar). Also, it is comparing
>assumptions to reality. But "boardings" is a metric transit
>agencies use and people understand. If it can be shown that a
>realistic PRT system can match or beat other modes in this important
>metric, then I think it would be significant and at least a useful
>talking point.

OK - that is a different problem. Tshen the question that need to be
answered is who are the people (not just the number of them) who
would benefit from a rail investment versus the PRT? And, you are
still stuck with comparing reality with a forecast, with all of its
assumptions about the future, the travel behavior of individuals and
the attributes of competing modes. What is required to get people to
accept the forecast as being "credible"? I've seen PRT forecast
results easily shredded and destroyed by critics who simply challenge
some of the assumptions that they are based on as being "unreasonable".

>Yes, departures and boardings are the same thing. I was just
>suggesting using the industry term in case the industry didn't understand.

I understand - but someday the transit industry needs to become more
realistic and credible, as critics often refer to the "transit math"
that they often employ.

[Marsden Burger]

Jerry,

Do you have a more exact link to this study than:

my prtquick.htm

Thanks,

Marsden

-----Original Message-----
From: transport-...@googlegroups.com
[mailto:transport-...@googlegroups.com] On Behalf Of Jerry Schneider
Sent: Wednesday, September 29, 2010 1:13 PM
To: transport-...@googlegroups.com
Subject: Re: [t-i] Visualization of capacity on PRT networks

[Jerry Schneider]

At 11:07 AM 9/29/2010, you wrote:
>Jerry,
>
>Do you have a more exact link to this study than:
>

>my prtquick.htm - here it is

<http://faculty.washington.edu/jbs/itrans/big/SWEfeasibilitystudy.pdf>Large
PRT feasibility study released by SkyWeb Express, 44
<http://faculty.washington.edu/jbs/itrans/big/SWEfeasibilitystudy.pdf>pp.

[Marsden Burger]

Robert,

 

 

Again, everything old is new again....

 

You might not have seen my recent post under another thread, but the question you are asking, “How does light rail compare directly with small vehicle systems?” is the same question I was being bombarded with repeatedly in 1979 as the lead for North America for the Cabintaxi joint venture.  To the best of my knowledge, nothing has significantly changed since then.

 

Started in 1976, the federally funded Southwest Corridor Alternatives Analysis, led by the City of Chicago Department of Public Works Research and Development Division; with the support of the Regional Transportation Authority, the Chicago Transit Authority, and the Chicago Area Transportation Study (the State of Illinois), included Cabintaxi (a small vehicle automated guideway system which is capable of both personal rapid transit and small vehicle advanced group rapid transit) in a direct comparison of bus, heavy rail, light rail, and automated guideway transit (AGT).  Those groups made up one of the most competent planning teams in the nation at that time.

 

At the time of the study, I was an employee of the City of Chicago and a member of the Public Works R&D team.

 

The results were that AGT (in the form of Cabintaxi) was significantly better than any other alternative.  However, the reality was, it could not be built by our City because it had not been built and operated anywhere in the world.  We knew this before we started the analysis.  AGT was taken into the study because the R&D Division felt it important to consider an important option for the future and identify its pros and cons – higher ridership at lower cost, cons never demonstrated.  

 

 

The AGT portion of the Southwest Corridor study became buried in the next level of analysis with the selected alternative.  However, it provided a good real world comparison of a corridor analysis between light rail and small vehicle systems, with light rail loosing out to the more convenient service of small vehicles.  Comparing light rail directly to small vehicle systems down the same corridor with the same hinterland (no advantage given Cabintaxi to the idea of network coverage and a broader ridership shed) small vehicle systems would attract 35% more rides than light rail from the given population.

 

I tried to capture the essence of this reality as a quick handout for English speaking planners in the simple document that I put together in 1979.  The transportation corridors and the CBD are stylized to allow those that only think in transportation corridors to better get the comparison potentials.  I will repeat it here for those that were not following the other thread.  Little has changed since 1979.  The relationships are still valid.

 

Best wishes,

 

Marsden

 

 

 

 

 

 

---

 

 

 

 

 

-----Original Message-----
From: transport-...@googlegroups.com [mailto:transport-...@googlegroups.com] On Behalf Of Robert Williams
Sent: Wednesday, September 29, 2010 4:45 AM
To: transport-...@googlegroups.com
Subject: Re: [t-i] Visualization of capacity on PRT networks

 

  I was actually proposing something much more simple than a one-on-one

--

[Rob Means]

FOR IMMEDIATE RELEASE: September 29, 2010
CONTACT: Bob Williams, 408-733-1102
10/10/10 EVENT HIGHLIGHTS ADVANCED TRANSIT

Personal Rapid Transit (PRT) Project Billed as Next Example of Advanced
Transit --

"Bridging Milpitas: the next example of modern transit" is sponsored by
the Advanced Transit Association group in northern California
(ATRA-NorCal) as a public outreach event to inspire the first U.S.
implementation of Personal Rapid Transit. PRT systems, which are as
easy to use as an elevator, promise efficient demand-responsive public
transit networks with automated non-stop service available 24 hours a
day.

ATRA promotes the consideration of Advanced Transit whenever
transportation planning occurs, but does not endorse any vendor,
technology or project. The event will be held at the park adjacent to a
proposed PRT crossing of the railroad tracks in Milpitas. For details of
the proposal, see http://www.electric-bikes.com/prt/ferry.html

"Bridging Milpitas" will include the first public showing of the
Tri-Track prototype (which is two-thirds complete), a 4-passenger
dual-mode "car" designed for regular roads -- and elevated rail travel
at speeds up to 180 mph . (Learn more at TriTrack.net) Also on display,
and available for very short rides by the public, will be a Jpods cab
and elevated rail. Other PRT firms will have educational PRT displays
and representatives to help explain the technologies and answer
questions.

The event is co-sponsored by Sunnyhills Neighborhood Association
(Milpitas) and Jakes Associates, a leader in innovative transportation
solutions. As part of the event, a "community resources" area will
include helpful organizations ranging from the local Farmers' Market to
an electric bicycle dealer. Also expected to attend are elected
officials, Milpitas Council candidates, neighboring residents, and
others interested in modern transit and reducing traffic. Finger foods
and drinks will be provided.

Held under the auspices of 350.org, an international movement to solve
the climate crisis, this event proclaims advanced transit as a way to
reduce our carbon footprint while serving transportation needs for
neighborhoods and cities.

Come to "Bridging Milpitas" located in Parc Metro East Park at the East
end of Curtis Avenue on Sunday, October 10, from 2 to 4 pm. Due to
limited parking, attendees are advised to park in the northeast corner
of the Great Mall parking area and walk to the Parc Metro East park.
For more information and a map, see http://www.350.org/node/18428

###

[WALTER BREWER]

"Boardings are used by the industry but are somewhat misleading, since
many transit riders board twice (coming and going) and so the number
of "different" people served is about half of what might be
understood by the public - good for transit agency PR, maybe not so
good for public understanding. Newspaper reporters never mention this
in their articles."

I suggest boardings is even more misleading, and as Jerry says very
misleading in the press, and even some planning comparisons.

I don't quite understand what the "going and coming" means if this is a one
way comparison. But the effect is corect and probably more so.

If we compare on a real origin to real destination, a mass transit journey,
I'll call it, requires typically 3 boardingss, compared to one for autos and
PRT.

Another unfortunate factor is comparing mass transit boardings with
automobile counts on roads, of which California has extensive data. The road
counts, if occupancy of people is included per vehicle are actually
on-boards. The equivalent of boardings would be to count occupants entering
the freeway but not subtracting those leaving.

For these reasons at least, I suggest passenger-miles is a bettter measure
of transportation mode effectiveness.

Walt Brewer

----- Original Message -----
From: "Jerry Schneider" <j...@peak.org>
To: <transport-...@googlegroups.com>

Sent: Wednesday, September 29, 2010 1:13 PM
Subject: Re: [t-i] Visualization of capacity on PRT networks

[Jerry Schneider]

At 07:06 PM 9/29/2010, WALTER BREWER wrote:
>"Boardings are used by the industry but are somewhat misleading, since
>many transit riders board twice (coming and going) and so the number
>of "different" people served is about half of what might be
>understood by the public - good for transit agency PR, maybe not so
>good for public understanding. Newspaper reporters never mention this
>in their articles."
>
>I suggest boardings is even more misleading, and as Jerry says very
>misleading in the press, and even some planning comparisons.
>
>I don't quite understand what the "going and coming" means if this
>is a one way comparison. But the effect is corect and probably more so.

Going from A to B, coming back from B to A = 2 boardings by the same
person. Most people who use transit make at least 2 boardings per
round trip, some more than that if the make one or more transfers on
their trip. If you start on a bus, transfer to a train, and return
the same way, that counts as 4 boardings (usually). I've also read
that the automatic counters that count boardings often produce inaccurate data.

>If we compare on a real origin to real destination, a mass transit
>journey, I'll call it, requires typically 3 boardingss, compared to
>one for autos and PRT.

I don't know where the 3 comes from.

>Another unfortunate factor is comparing mass transit boardings with
>automobile counts on roads, of which California has extensive data.
>The road counts, if occupancy of people is included per vehicle are
>actually on-boards. The equivalent of boardings would be to count
>occupants entering the freeway but not subtracting those leaving.
>
>For these reasons at least, I suggest passenger-miles is a bettter
>measure of transportation mode effectiveness.

How would you get the data without tracking ons and offs for each and
every rider?

[WALTER BREWER]

### Comments ###

Walt Brewer
----- Original Message -----
From: "Jerry Schneider" <j...@peak.org>
To: <transport-...@googlegroups.com>;
<transport-...@googlegroups.com>
Sent: Wednesday, September 29, 2010 11:36 PM
Subject: Re: [t-i] Visualization of capacity on PRT networks

> At 07:06 PM 9/29/2010, WALTER BREWER wrote:
>>"Boardings are used by the industry but are somewhat misleading, since
>>many transit riders board twice (coming and going) and so the number
>>of "different" people served is about half of what might be
>>understood by the public - good for transit agency PR, maybe not so
>>good for public understanding. Newspaper reporters never mention this
>>in their articles."
>>
>>I suggest boardings is even more misleading, and as Jerry says very
>>misleading in the press, and even some planning comparisons.
>>
>>I don't quite understand what the "going and coming" means if this is a
>>one way comparison. But the effect is corect and probably more so.
>
> Going from A to B, coming back from B to A = 2 boardings by the same
> person. Most people who use transit make at least 2 boardings per round
> trip, some more than that if the make one or more transfers on their trip.
> If you start on a bus, transfer to a train, and return the same way, that
> counts as 4 boardings (usually). I've also read that the automatic
> counters that count boardings often produce inaccurate data.

### We are agreeing in principle. I had assumed the comparison was for the
one way part of a journey.
The usual pattern for an all mass transit journey is board a vehicle to get
to a station for the primary leg. At another station board another vehicle
to real destination. Thus 3 each way. 1 of thes would be by auto if you
drive to the first mass transit station. (I know a commuter why has cars at
both ends.) Press releases etc etc tend to only count the major leg.
To be sure some happen to live near the major part line and might have only
2, or in rare cases 1 boarding per journry.
I'm aware of some cases where a LRT line has replaced direct bus trips.
Typically then a bus takes the rider to the LRT and maybe from LRT to real
destination. Thus the record shows 3 instead of 1 for mass transit credit.
###

>
>
>>If we compare on a real origin to real destination, a mass transit
>>journey, I'll call it, requires typically 3 boardingss, compared to one
>>for autos and PRT.
>
> I don't know where the 3 comes from.

## See above ###

>
>>Another unfortunate factor is comparing mass transit boardings with
>>automobile counts on roads, of which California has extensive data. The
>>road counts, if occupancy of people is included per vehicle are actually
>>on-boards. The equivalent of boardings would be to count occupants
>>entering the freeway but not subtracting those leaving.
>>
>>For these reasons at least, I suggest passenger-miles is a bettter measure
>>of transportation mode effectiveness.
>
> How would you get the data without tracking ons and offs for each and
> every rider?

### San Diego mass transit Agencies do just that. The detail records, which
Agencies make difficult for the public to see indeed show on-boards, and
calculate pass-miles. On a sample basis that is, not all trips. ###

[Kirston Henderson]

----------
From: Marsden Burger <Cabint...@msn.com>
To: <transport-...@googlegroups.com>
Subject: RE: [t-i] Visualization of capacity on PRT networks
Date: 29, Sep, 2010, 4:36 PM

At the time of the study, I was an employee of the City of Chicago and a
member of the Public Works R&D team.

The results were that AGT (in the form of Cabintaxi) was significantly
better than any other alternative.  However, the reality was, it could not
be built by our City because it had not been built and operated anywhere in
the world.  We knew this before we started the analysis.  AGT was taken into
the study because the R&D Division felt it important to consider an

important option for the future and identify its pros and cons - higher

ridership at lower cost, cons never demonstrated.  

    And nothing has really changed today.  In every set of system selection requirements from city officials, you find the requirement that the selected system must have been used in some city in daily revenue service, a requirement that rules out anything new.