02 June 2012

Is Demand-Based Planning a Myth?

Original photo by qviri
With over 80% of riders using Caltrain to commute to their jobs during rush hour, one would think that the service would be planned around where people live and where people work, using cold hard numbers from the census.  That's known as traditional demand-based planning: provide service where the most people will use it.  It's not rocket science, and demand-based planning is used all around the world to plan excellent rail service.

But not here on the peninsula.

In a contrarian argument made circa 2005, Caltrain's operations staff claimed that demand-based planning is a myth. (14 Mb PDF file)   At the time, Caltrain was crowing to its industry peers about the success of the Baby Bullet.  The keys to success included "Questioning Traditional Planning Processes" and "Trusting Your Intuition".  Numbers don't matter, just go with your gut!

In the years since, there has been plenty of hard evidence that the Baby Bullet has severely reduced ridership at many locations, especially in Santa Clara County.  Indeed, data from the 2010 census can be correlated to the latest Caltrain ridership data without ever looking at a timetable to reflect quite accurately which Caltrain stops are under-served and falling short of their ridership potential.

Maybe demand-based planning isn't such a myth after all.  Maybe numbers don't lie.  Here's hoping that objective, quantitative metrics will play a central role in planning future blended operation scenarios with high-speed rail.  This stuff is too important to trust anybody's intuition.

112 comments:

  1. Do you dare to question the dedicated, hardworking staff at Caltrain?
    Do you dare to question that Baby Bullet is the panacea that all other agencies should adopt?

    Sacrilege…

    ReplyDelete
  2. The circa 2005 “demand-based planning is a myth” presentation is extremely one-sided, as if *nothing else* drove, or will drive ridership increases???

    Who was this presentation made to? I don’t recall ever seeing this presentation at any Caltrain Board Meetings. Who/what is this John Hugunin/Earth Tech? Did Caltrain tour around the country to other agencies crowing about their “success” Baby Bullet? All rail agencies should follow Caltrain’s lead if they are to increase ridership… screw the local riders and smaller stations… customers will just flock to a bullet station even if they have no means to get there…

    ReplyDelete
  3. That 80% of Caltrain traffic is rush hour riders is not a good thing. It means that the off-peak service is substandard, and this means that more equipment and labor are needed per rider than on more even systems.

    ReplyDelete
  4. Alon: That 80% of Caltrain traffic is rush hour riders is not a good thing.
    I think Caltrain is better than ACE, Metrolink, Metra...etc those commuter rail runs only commute hours or "Relatively" few train on midday.

    Good thing for Caltrain is "RELATIVELY" higher percentage of weekend ridership. You will find ridership statistics from APTA (American Public Transport Association) website.
    (2012-1Q) http://www.apta.com/resources/statistics/Documents/Ridership/2012-q1-ridership-APTA.pdf

    When you calculate the ridership ratio between Monthly and Average weekday, you will see system utilization. (Higher number means more weekend ridership)

    Here are some data for Feb 2012(unit 1000 people)

    BART 9663K(Monthly), 383K(Average Wkday) 9663/383=25.2
    Caltrain 1088K(Monthly), 42.4K(Average Wkday) 1088/42.4=25.7
    Metrolink 941K(Monthly), 43.1K(Average Wkday) 941/43.1=21.8
    LIRR (NY)7585K(Monthly), 324K(Average Wkday) 7585/324=23.4

    It is surprising that Caltrain carries "Same level of" weekend ridership than BART, under such poor time table. Please note that BART runs every 20 min in weekend.

    Caltrain management should realize this number and take action to create potential demand for revenue. Adding weekend train require no capitol investment and no technical challenging.
    However, Caltrain management need to change their mindset from "COMMUTER" train to "RAPID" transit, even before electrification.

    Adding hourly express train from 9AM to 6PM will be the first step. When they tried weekend Baby bullet in JAn-2011, Caltrain explained 10% (~1000) additonal rider will offset the cost increase of 4 trains.
    Caltrain already archived this goal only few month after implementation. Why they don't repeat same thing to add more revenue? Now is time for next step.

    ReplyDelete
    Replies
    1. Serious question: would riders be more interested in an hourly express train, or in a second local train per hour, providing 30-minute frequency to every station all day? Obviously the answer is different depending on who you ask - people in Mountain View and Palo Alto are going to answer differently from people in San Mateo and San Antonio - but one set of people may be much larger than the other.

      Delete
    2. RE: "Good thing for Caltrain is "RELATIVELY" higher percentage of weekend ridership."

      I think it tells you how bad the weekday share is!

      40k trips/day with one end "serving" the SF CBD? Pathetic! SF has local bus routes about as good.

      I think the weekend/weekday ratio just tells you the number of poor decision makers (or carless, or very time insensitive) who ride no matter how bad the service.

      Also BART weekend/off-peak is badly dragged down by Pittsburg, Dublin, Fremont (and soon Warm Springs and San Jose) where the trains run empty and run twice as often as justified.

      In contrast, SF-Oakland and SF-Berkeley and SF-San Leandro can be standing room only at any hour of the day or night. Just like SF-Palo Alto ought to be if Caltrain weren't a joke.

      Delete
    3. As one who occasionally leaves work early to take the train from Palo Alto to 4th and King mid-day, I'd rather have locals at consistent 30 minute intervals, than an express followed by a local at some other interval. Once per hour express trains are only useful if they happen to match your schedule and destination needs. With the varying gaps in interval that are inevitable with one local and express scheduled per hour, the time benefit of the express train will be lost for many (if not most) passengers.

      Delete
    4. RE: In contrast, SF-Oakland and SF-Berkeley and SF-San Leandro can be standing room only at any hour of the day or night. Just like SF-Palo Alto ought to be if Caltrain weren't a joke.

      This may be true. But comparing BART's "Core" business and Caltrain is not fair comparison. If you see "Mindset" of people in San Carlos, their standard is still one of "commuter rail".

      To add more weekend service, we should encourage Caltrain management to look for above (other heavy rail or international) , not a holizontal(=other commuter train). From my standard, Caltrain timetable is very poor. However, Caltrain is much better than other "Commuter rail" in USA. They cannot chage immediately but step by step.

      Another encourage to Caltrain is bi-directional peak demand. This makes better utilization of crew/equipment. So, Caltrain should have "Much better" farebox recovery than other commuter rail. Why not?

      Delete
    5. Anonymous : Your analysis effectively comparing weekday vs. weekend Feb. 2012 ridership showing Caltrain’s weekend patronage to have an unusually high strength compared to weekday ridership poses the question Why? Caltrain’s weekend expresses certainly help. An additional reason might be due to the exactly hourly local train departures at every station throughout the weekend could be significant.
      Shifting all locals to two instead of five passenger car trains (Or occasionally 3 cars while absorbing the U.S. commuter train industry standard 3,3% schedule pad.) would save 8 minutes from the present SJ−SF local train-in-motion-period enabling all-stop rush-period trains to fit into a moderately adjusted express train schedule. Using only Bombardier cars for local trains and building two 15 inch-high (above the current 8 inch-high platforms) 200 square-foot wheel-chair accessible platforms at every station would minimize wheel chair delays and make only one conductor per train practical. Ticket display clips within reach of every seat would also ease the work-load of a single conductor per train. Two car local trains would save 31% in traction power fuel and free-up 2 to 3 cars per train for heavily patronized express runs.(A 990,000 pound train at 79 mph has 78 kw hours of kinetic energy which requires 4+ gallons of diesel fuel. Assuming 137 mega-joules per gallon available from diesel fuel a 38% heat to mechanical energy conversion efficiency and an average 8% transmission energy loss during acceleration through the entire speed range. Therefore one full speed acceleration from a stop costs over $12 just for fuel when subject to a $3/gallon fuel charge.)
      Additional express trains scheduled to reduce local train crowding could be paid-for with the shorter-train operating savings. (Before the turn of the last century when the 5:20, 5:25, and 5:30 pm zone expresses departed SF before the 5:35 local where I counted 180 and 220 passengers at the peak loading point near Millbrae on two separate checks. Neither the conductor no the collector were busy more than half the time.) Caltrain could then truthfully post individual schedule notices for each direction at every station stating that a least one train is will arrive every hour at the same number of minutes past the hour between early in the morning to late evening every day of the year. Given the high proportion of occasional potential users especially during non-rush-hour periods an easily remembered departure schedule should make a significant ridership increase probable during non-rush-hour periods.
      Running two-car locals combined with more-convenient wheel-chair boarding infrastructure and easily seen ticket clips would give Caltrain a good shot at increasing revenues faster than operating costs by allowing the foregoing infrastructure changes to enable improvements in local schedule adherence, speed, and consistency while simultaneously increasing non-rush-hour express-train frequency

      Delete
    6. @ John Bacon… Weekend ridership has always been quite significant, with some weekend trains carrying more passengers than weekday peak hour trains. Even 25 years ago, when the Saturday schedule was less frequent, I would ride from SJ to SF, observing 500-600 passengers getting on just one train. The weekend Bullets may help but their passenger loads are generally average among weekend trains. Not sure what the “exactly hourly local train departures at every station throughout the weekend” has to do with the ridership?
      If anything the hourly frequencies may deter ridership.

      “Shifting all locals to two instead of five passenger car trains (Or occasionally 3 cars…)”

      Have you ever been on weekend Caltrain? Many weekend trains are packed, especially on Baseball game days.

      Supplementing the local service with additional local and express service would provide quite a boost to weekend ridership.

      “Ticket display clips within reach of every seat would also ease the work-load of a single conductor per train.”

      What??? Caltrain is a proof-of-payment system, conductors *do not* sell tickets on board anymore. The conductor is there to ensure the safe operation of the train and to act as fare inspector doing *random* checks for proper fares/tickets.

      Delete
    7. If I recall correctly, there actually are ticket display clips on some of Caltrain's Gallery cars. They're just not used for anything.

      Delete
    8. Those are left over from the pre POP days when conductors would check all fares/tickets. Once checked, the conductor would place a ‘hat-check’ in the clips so that he would know that that customer has already been checked.

      Delete
  5. Ideally, Caltrain needs both local and express train every 30 min during weekend before the electrification. Every 20 min frequency is needed after electrification.

    Adding express train into current weekend schedule is easier for Caltrain management to accept. Running hourly express train between SF-SJ requires 3 sets of equipment and crew. (64 min run time for oneway + 16 min turn around time: 64+16+64+16=180 min)
    Adding hourly local train requires 4 sets. (96+24+96+24=240min)

    For revenue point of view, we need 33% more revenue (per train) from local train. Clearly, Caltrain management have no confidence in adding local train into existing schedule. Adding express train will be easier for them to justify.

    ReplyDelete
    Replies
    1. Adding an hourly local only requires 3 extra sets, when you think about it right: today's schedule requires 4, whereas a half-hourly schedule could have 96+9+96+9 = 210 and use 7 total.

      An hourly express would require 6 (96+10+64+10 = 180), if Caltrain could schedule the expresses and locals appropriately.

      Delete
    2. Actualy, today's schedule requires 5 sets. The train that arrives at San Jose at :51 past the hour doesn't turn around to become the departure at :00 past the hour, instead the arriving train goes to the yard and the departing train is a fresh train from the yard. In fact, as far as I can tell, the typical weekend workday for Caltrain crews is taking their train from the yard, to San Jose, doing one round trip to SF, then taking the train back to the yard.

      Delete
    3. I have no convenient place to bang my head right now. Sigh.

      Anyway, if Caltrain were run like a 21st-century first-world railroad, half-hourly locals would require 7 sets vs. 6 for an hourly local and an hourly express. (In such a hypothetical Caltrain would be electrified already and then half-hourly local service south to Tamien would require 6 sets.) Typical schedule would then be taking a train from the yard or from the last crew, doing two roundtrips, handing it over to the next crew, and going home.

      Delete
    4. The train that arrives at San Jose at :51 past the hour doesn't turn around to become the departure at :00 past the hour.

      Crew don't have to turn around but train set can turn around.

      Delete
  6. Alon, Thank you for calculation. I think 9~10 min turn around is not technically difficult, but current turn around time of Caltrain are very~very long.

    Why is that?
    Strong labor union?
    Diesel engine need some cooling down?
    No such mind set for Caltrain management in San Carlos?

    Southwest introduced short turn around time for airline industry. Why Caltrain needs longer turn around than airplane?

    ReplyDelete
    Replies
    1. The MBTA and Metro-North both turn in about 5 minutes at the outer end of the trip when they need to. So even with diesel engines, short-ish turnaround times are possible. For cultural reasons, they take forever to turn at the city end. Both railroads also have way more tracks at the city terminals than they need, so the cultural issue here is really the preference for scheduling the long layover to be at the city end rather than at the suburb end.

      Delete
    2. Adirondacker1280006 June, 2012 08:41

      Those dastardly engineers, conductors and assistant conductors. Imagine the never wanting to use the bathroom and maybe get something to drink when they get to the terminal. And do things like turn in cash. Outrageous..

      Delete
    3. Anon, Southwest's goal is to make a profit, Caltrain is to provide a commuter service. If Caltrain was actually interested in providing an efficient service that was cost-effective, they would shorten the turnaround times and increase trainset utilization. They could then reduce the size of the 4th and King terminal, which is one of the biggest wastes of space I've seen- seriously, one track per train per hour?

      Delete
    4. Long turnaround times are convenient when you have unreliable equipment, as they allow time to take equipment out of service and replace it. I wonder if that's the real reason?

      --Nathanael

      Delete
    5. Unreliable equipment? Long turnarounds allow you to swap out unreliable equipment or even repair it.

      --Nathanael

      Delete
    6. Doesn't explain why Northeastern railroads routinely have short turnarounds at the outer end. Most outer ends don't have any place to swap equipment and so lengthening the turnaround doesn't add to equipment reliability, but New Haven does, and still Metro-North has short turnarounds there.

      Delete
    7. Hmm. Good point, Alon, long turnarounds at non-maintenance / non-storage locations have no plausible purpose.

      --Nathanael

      Delete
  7. @Adirondacker: Caltrain conductors haven't had cash to "turn in" since they stopped selling tickets a decade or so ago. And when service is frequent, it would probably make more sense in many cases to "leapfrog" crews, that is for a different crew (e.g. from a prior arrival) take the incoming crew's train back out.

    ReplyDelete
  8. Conductors in more civilized places in the world still sell tickets and still take cash. Conductors anywhere in the world still need to use bathrooms, take breaks etc.

    ReplyDelete
    Replies
    1. True. Germany isn't civilized. SEPTA is.

      Please, travel.

      Delete
    2. Yes, they do all of those things in other parts of the world, and they also to manage to run a service that doesn't waste time (and losing revenue) sitting around terminals.

      Delete
    3. Adirondacker1280008 June, 2012 12:52

      Much more efficient to be running trains around empty than to have them sitting at the terminal, unstaffed.

      Delete
    4. This comment has been removed by a blog administrator.

      Delete
  9. Demand is obscenely unpredictable -- I would say that any demand prediction has a variance of at least 100% -- but there isn't really an alternative to demand-based planning. You can tweak your model to show that demand is pretty much whereever you think it's going to be, however, so maybe it's more honest to just go with instinct.

    To me, instinct says "Go to the popular places, then connect them to the places with lots of people. With frequent service."

    --Nathanael

    ReplyDelete
    Replies
    1. There is no modeling involved in figuring out where people live and work. It's just cold hard data, and it's available to anyone who seeks it.

      Delete
    2. You still have to model whether those people are going to use the train if you offer the service. Plus there's also induced demand. Just using verbatim census numbers of residences and places of work as a basis to decide where to put services doesn't sound much more reliable than the intuition of somebody's who's been working with the transit system for thirty years.

      Delete
    3. You could also do the OPPOSITE of modelling where the people are going, and gut service to real and significant job destinations such as Santa Clara, Sunnyvale, and California Avenue.

      Because of, you know, your Special American Transportation Planner Expertise achieved through 30 years of not being qualified to even get an interview for a job for planning or running a transit network anywhere else in the world.

      Gut feelings! Like those ones that tell us that, in spite of actual inspections of actual data, that Saddam Hussein is secretly in possession of shitloads of Weapons of Mass Destruction. Damn the facts! Damn the data! Damn the demographics! INTUITION and BOLD LEADERSHIP is what it takes to boldly decide where the trains run.

      Delete
    4. Strawman...

      Delete
    5. I agree, Richard and Clem, that these particular Caltrain proposals do not pass the smell test -- not only are they not based on data, they don't fit with the instincts most of us would have.

      If you're going to do something deeply counterintuitive, you had better have some data on your side.

      --Nathanael

      Delete
  10. From the this is how we at Caltrain implement proof-of-payment department…

    Beginning this past week, for some bizarre reason, conductors are to check all tickets as customers board all trains in San Francisco. There are 400 people waiting at the gates, we open the gates 15 minutes before departure and conductors check everyone’s ticket.

    The only time I would see them do this is after a baseball game which also they could screen out the too drunk baseball fans. I have also seen them sometimes do this late evenings on non-baseball game days.

    Conductors are still required to do random checks at some other point(s) during the trip. One conductor told me that when they do a random check some customers get annoyed since they have already been ticket checked.

    Correct me if I am wrong, but isn’t the whole idea of proof-of-payment supposed to be random checks for proper fare?

    ReplyDelete
  11. Caltrain monthly ridership is based on revenue/tickets sold, using formulas that go back 30 years; to the Caltrans/Southern Pacific days. They do a detailed on-board passenger count once a year in Jan./Feb. in which teams of surveyors log the on’s/off’s at each station for every train for each weekday and Sat./Sun.

    Caltrain could actually acquire realistic ridership data and provide more detailed/useful monthly reports. TVM’s know where they are and they can log the station data of origin and the destination zone for all tickets sold. There should be years of this data acquired in the TVM database. Now with clipper, the clipper readers know where they are and should provide tag-on/tag-off origin and destination data for all users. Its not rocket science, why doesn’t Caltrain provide more detailed ridership data?

    ReplyDelete
    Replies
    1. Somebody somewhere deep in the bowels of the accounting department has that information. Otherwise they wouldn't be able to check the cash versus credit card versus debit card transactions at the the TVMs. Every railroad has been doing that since they began carry passengers.

      Delete
    2. For some reason many here have some visceral dislike of ticket barriers, but the advantage of these is you have a much more controlled count of passenger numbers, both at the journey start and end points. With the addition of smart cards, you have even more detailed info about passenger journeys (transfers from other rail lines, bus transfers, etc.). With modern rolling stock equipped with TIMS (train information management system), passenger loads can be measured and recorded. All of this can be put on-line for the traffic department to use in planning improvements in service. With automation you reduce labor costs, and staff like conductors can focus on their critical task, which is ensuring the safe and on-time operation of their trains, rather than worrying if that hippie in car 3 is trying to dodge a fare.

      Delete
    3. "For some reason many here have some visceral dislike of ticket barriers"

      Funnily enough, "many here" have a visceral dislike of pissing the public's money away, of writing blank checks to rent-seeking defense contractors who manipulate requirements and bidding processes, and of making transit less convenient, slower, more expensive, and more user-hostile.

      But hey, if it allows us to track every one of the little people as they go about their business, and to deliver a cut to Cubic, Inc, every time anybody gets on of off any vehicle, what's not to like?

      Like up and be counted, citizens! It's your duty. An extra few minutes in a queue is the least service you can provide the fatherland corporation.


      "With automation you reduce labor costs, and staff like conductors can focus on their critical task, which is ensuring the safe and on-time operation of their trains"


      The "critical jobs" of conductors is to draw wages and pay UTU dues. The job is obsolete and has been for 50 years.

      "rather than worrying if that hippie in car 3 is trying to dodge a fare."

      What is it with the deluded people who are so upset with the prospect that somebody, somewhere might be getting away with something, somehow that they utterly lose all perspective and do things like pay $500 million (and counting) to a corporation in order to still have fare evasion?

      Fare control is (or ought to be, and sometimes, in civilized parts of the world can be) about costs and benefits, about realistic engineering, about value for money. Just as rail and road infrastructure design and rail and road vehicle design and transit network design can be, and, in places outside the Bay Area, sometimes is.

      Delete
    4. Interestingly enough, given your "line up and be counted, citizens" rhetoric, in actual Communist countries, namely the USSR, the commuter trains had effectively a POP system with random inspections. They only started to install faregates on the suburbn rail system in Moscow after the fall of the Soviet Union and the coming of capitalism. Likewise, the bus, trolleybus, and tram had a POP system but eventually switched to faregates as well. As in, there's actually a turnstile in the front of every bus in Moscow that every passenger has to go through. This obviously slows boarding considerably.

      Delete
  12. Richard:
    While fare gate barriers are ineffective with today’s 8 inch platforms after Caltrain is rebuilt with level boarding with higher platforms, fare gates to verify payment and record trip lengths would be a superior system to the current proof-of-payment system for the following reasons:
    (1) Using ticket barriers in order to gather rapid accurate feedback on peninsula riders’ travel patterns is particularly important for Caltrain given the many uniquely varied sources (40%+ reverse commute for example) for their patronage to an unusual degree compared to most commuter operations in the U.S. Frequently gathering reliable statistics can clearly show customer response to service quality changes. Ridership sources, or their sudden disappearance, are particularly important parameters to know where, as is certainly the case on the peninsula, the automobile is a viable alternative for most passengers.

    (2) A fare payment scheme that levies heavy fines for not getting the proper ticket from a TVM appears to be an unusually hostile system for an hourly service for infrequent users; a likely occurrence on trains which serve well under 10% of their potential market. One’s choices are particularly grim when considering boarding the last train in the evening after the station TVM doesn’t work for you.
    (3) Since when is a labor-intensive fare verification scheme cheaper than a machine based system? A barrier system with turn-styles was developed during the early 1920’s for the New York Subway in order to survive with the nickel fare until 1947.

    ReplyDelete
    Replies
    1. Caltrain could gather “rapid accurate feedback on peninsula riders’ travel patterns” if they were to drop the current customer hostile zone system and adopt station to station fare pricing. This can be achieved with a TVM based, barrier free system. No barrier equipment acquisition costs, no labor intensive maintenance costs for gates/turnstiles and no station agents needed to aid customers when there are problems with the turnstiles. Simple, easy to use barrier free system…

      The only reason for the current fare zone system, implemented in 2003, was so that on Caltrain the fare from Millbrae to SF would be comparable to the BART fare from Millbrae to SF. The current zones are customer hostile and discourage shorter trips. Since zones are way too long (13 miles), the base fare AND the per-zone fare has to be higher than necessary. Station to station pricing would result in a lower base fare and much more reasonable and FAIR pricing between ALL stations. Under the current system a customer can pay the same fare to go just 2 miles as a customer that is going 25 miles!!!

      Station to station pricing would result in increased ridership because of now reasonable fares between every station.

      Delete
    2. Reality Check13 June, 2012 16:48

      Fare gates are way more costly than POP. You've got to pay people to watch & service the costly-to-acquire-faregates. Those people will typically be unionized and will want decent pay and benefits and will need things like booths to sit in, monitors and buttons to watch and operate, etc., etc. What's more, there will still be faregate jumpers anyway.

      A short excerpt from an item in the (Los Angeles-area) Transit Coaltion's May 21 Weekly Transit eNewsletter:

      "At the Metro Board Executive Committee meeting last Thursday, Kymberleigh Richards noted on the fare gate locking proposal that the subway currently has 16 stations, five of which have two sets of fare gates, while 7th/Metro has three. That's a total of 24 gate sets, which the staff report says will be staffed by Sheriff's Assistants. 24 gate sets times 20 hours per daily span of service, times seven days a week, totals 3,360 hours per week of gate attendant wages, or 174,720 hours per year, times the hourly rate Sheriff's Assistants get paid. Richards questioned whether there is enough uncollected fare revenue to cover the expense of locking the gates."

      Delete
  13. Long fare zone lengths will encourage those with bargain fare rates to ride and those faced with high fare rates per mile to drive. In either extreme case the transit agency would likely have a lower yield per potential customer than a pricing system where fare levels are proportional to distance traveled. TVMs that print tickets at the end of the transaction should be able to easily handle the additional complexity involved in pricing all station pairs on the line on the basis of what the traffic will bear which in most cases is directly proportional to distance traveled. There are profitable exceptions: BART began a $3.80 surcharge for BART's SF Airport Station users without any serious drop in ridership. Perhaps Caltrain, the CHSR, BART and the First Avenue Trolley could profitably share a combined San Jose Airport Transfer Station directly beneath the most centrally located airplane passenger terminal.

    ReplyDelete
    Replies
    1. This comment has been removed by the author.

      Delete
    2. Before everyone gets too carried away with notions of expanded uses for TVMs consider this under-reported fact: the July 1st fare increase for paper media had at its core a motivation to reduce the use of current TVMs so they can last a little longer. Unhappily, there seems to be little enthusiasm in San Carlos for the implementation of anything new and creative to drive revenue and ridership, especially flexible fare structures which would rely on "modern" fare collection systems and not the old-tech fiasco (Clipper) that MTC imposed on the Bay Area.
      Oyster cards anyone?

      Delete
    3. Adirondacker1280012 June, 2012 19:36

      According to Wikipedia both Clipper Cards and Oyster Cards use
      MIFARE DESfire chips in the cards. So the significant difference between a Clipper Card and an Oyster Card is what? Besides the logo on the outside of the card?

      Delete
    4. @ A12800 Congratulations! You've missed the point entirely. The card is not the "system", it is just one component. The significant difference is the implementation of the system, and the creativity of those that specified it. It's why Oyster can calculate whether you have used the equivalent value of a day pass then stop further charging, versus the dumb-as-a-rock Clipper system that can't even manage 8-rides logically.
      Wikipedia indeed!

      Delete
    5. Adirondacker1280013 June, 2012 05:53

      Silly me mistaking "Oyster Cards anyone" for, ya know, those wallet sized things ya put in your wallet..
      It didn't occur to me that a system designed to bill for transportation services in Greater London would be particularly useful in the Bay Area or anyplace else outside of the Greater London area.

      Delete
    6. Apparently Clipper now limits your daily VTA swipes to the cost of a day pass.

      Delete
    7. Metrolink in Southern California used to have a zone system, but they switched to pure distance-based fares with monthly passes. They have TVMs that print paper tickets, and until recently, they had "cut a corner" style 10-ride ticket validators too. That system is a good way to prove both Caltrain and BART wrong when they claim they can't change their fare system. Oh, and they operate trains with one conductor, and occasionally have sherriff's deputies ride the trains to do fare inspections and provide security.

      Delete
    8. @Peter So says the VTA/Clipper web-page: "Unlike the paper day pass, which is purchased in advance, Clipper automatically grants you a day pass once you've paid the same amount in Clipper Cash as the day pass costs. Once you've earned a day pass, Clipper will stop deducting fares from your Clipper Cash balance and all your rides for the rest of that day will be free of charge. Clipper automatically caps your day's fare, so you may see less than the standard fare deducted if you're close to the daily maximum."
      Which suggests there absolutely no technical reason why Caltrain could not emulate this approach...... or is there?

      Delete
    9. Don't forget the other "bonus" delivered by the $500 million Clipper system: you're required to waste time queuing to get OFF the train ("tag off or get completely screwed, and all the money goes to Cubic hah hah hah hah hah).

      The nice and civilized thing about paper ten rides is that you bought them in advance, kept them in your wallet, walked up to the validator once as the train arrives, and then, when you get to the other end, where-ever it is, simply leave the train directly and go about your business. Just like buying a ticket for anything else.

      The Clipper "tag off" BS adds a minute or two (or more, eg at 22nd in the evening) to nearly every trip of the 50% of riders who don't have monthly passes. (Which note are still "monthly", even though there's no reason that can't get 30 days starting in any day. And which note can no longer be used as partial trip credit with "Zone Upgrade". Just one Clipper screw after another.) To save a couple minutes per trip for half your riders by any other means is going to cost serious money.

      Clipper! The gift that keeps on giving. And giving. And giving.

      Delete
    10. Reality Check14 June, 2012 18:11

      No way to add value to (or get) Clipper cards at Caltrain stations. Ya gotta love all the convenience! We're not worthy! We're not worthy!

      Delete
    11. Tagging off doesn't have to be this difficult. They could add more validators at busy stations; Singapore pays S$960 (about US$750) per validator nowadays.

      Delete
    12. What Alon said. The problem is that Clipper is being mind-bogglingly poorly implemented. There's two shining examples of best practice for this sort of card -- London and Hong Kong -- can't you manage to copy one of them?

      No, apparently you can't. Perhaps this is due to the pile of different squabbling agencies (London has one central transportation agency which controls everything).

      --Nathanael

      Delete
    13. Best practices are in Hong Kong, Tokyo (and increasingly other Japanese cities - though Japan as a whole has agency integration problems), and Singapore. London is the city that fined passengers who took out the chip from the card to make an ersatz Oyster watch, on the model of Hong Kong's Octopus watches.

      Honestly, to avoid Japan's problems with needing bilateral agreements for interoperability, this should just be a federal thing. There are already two kinds of electronic money paid by the federal government (EBT for food stamps and Eagle Cash for the military), and problems with small change caused by stores' listing prices without sales taxes. Do an anonymous electronic money card once nationally, integrate it with EBT to make an established base of users, and license it to transit agencies and such at-cost. (Or just make stores accept it. If they can make everyone accept $2 bills, they can make everyone accept electronic money.)

      Delete
    14. Ew, no thanks. The idea that to have two bureaucracies cooperate, you need to set up a third bureaucracy above both of them is exactly what led to the Clipper fiasco. There are plenty of transit agencies around here that, for all their faults, have plenty of experience in the business of collecting fares from customers. But instead of having one of them set the standard and building some kind of federated system, the task fell to the MTC, which is a regional planning bureaucracy with no experience actually operating transit or fare collection systems. And since everything is now centralized, changing anything requires asking the MTC to ask the contractor to implement it, which is why Caltrain only "agreed" to the Clipper system under threat of losing funding, and why Caltrain is to some extent stuck with the silliness of the existing Clipper fare system. And of course, the MTC doesn't acknowledge the existence of non-MTC agencies, which is why Caltrain no longer has discounted transfers to Santa Cruz Metro or MST buses, while the VTA has no problem doing exactly that with their new non-Clipper contactless card system (that is shared with SCMTD and MST, and wasn't as far as I know, imposed by anyone).
      You might also think that an approach of letting agencies set their own standards and trying to get others to adopt them would take a long time. But I have a hard time believing that it would take longer than the 17 year development of Translink/Clipper actually took.

      Delete
    15. As far as I can tell, there's exactly *ONE* Clipper Add Value machine in Santa Clara County (at Diridon). What an effing joke.

      Delete
    16. Every Walgreen's has Clipper machines. Just ask to fill up or buy a new card at the customer service desk.

      Delete
    17. Q: You're at a station. A train station. You wish to catch a train. You wish to buy a ticket. How do you do it?

      A: Locate a Walgreen's.

      Fucking brilliant.

      Delete
    18. A: You buy a ticket at the ticket vending machine, which for all its flaws, is at every station.

      amandainsjc asked if there was any Clipper Add Value machine in Santa Clara county. The answer is: Yes, at every Walgreen's. Is this a pain sometimes? Yes. Do other systems have this same issue? Yes.

      I have found myself running around the city of Boston looking for a place with CharlieCards for a friend, and I wish I could consistently count on finding them at one particular store like I could for Clipper in the Bay Area. On the commuter rail it can be even worse (they don't take CharlieCard anyway) and you have to hunt around for a place selling tickets on the side. Even in major Northeast Corridor stations like Providence.

      Have you, "Anonymous", ever ridden Caltrain?

      Delete
    19. There have been two Clipper Add Value machines at the Palo Alto Station (which, last I looked, is in Santa Clara County) for several months. Perhaps unsurprisingly, these machines have been non-functional far more often than not (often showing some sort of Windows error), which does not exactly inspire confidence. Oddly, the pair of Add Value machines at 4th & King have been reliable since installation. Walgreens is a perfectly usable alternative, I suspect there is one within 1/2 mile or so of just about every Caltrain station. Of course, the big advantage of Walgreens is that they are widely available to the large numbers of Clipper users who rarely visit a Caltrain, BART, or Muni station...

      Delete
    20. The Clipper machines at 4th & King are from 80's. With their bizarre UI that makes makes you page through all the zones, then wait two minutes for "dialing", makes you insert the clipper card twice and swap it with the credit card, and then how do you complete your transaction? by hitting "Cancel" to cancel it!!!

      When I need to add rides, I now do it in Muni Metro where their clipper machines can complete the transaction in 30secs with a much nice experience.

      Delete
    21. Anyone contemplating using a Transit Flex Account debit card will quickly find out what existing users already know; you can't use those debit cards to pay for Clipper at Walgreens as their merchant code is disallowed for pre-tax transit accounts. And yes, the Clipper machines at 4th & King are quaint, aren't they? Just leave plenty of time before your train departs if you need a top-up.

      Delete
    22. You're at Redwood City Caltrain ... nearest Walgreen's is over a mile away in San Carlos, nowhere near a Caltrain station. I used to use Commuter Checks to load Caltrain passes onto my Clipper Card there. You wait around for them to page an employee trained/authorized to do Clipper transactions. More than once, the guy asked if I can come back, maybe in the evening or tomorrow, because their system is down or malfunctioning or not communicating or something. Brilliant. Transit in America: only the very best, for the very best, by the very best!

      Delete
    23. Matthew, I don't want to defend the MBTA, but there was a CharlieCard add-value machine at every station I have used, and when I wanted to get a card, I could get it at the station I was using, I forget which one.

      Yes, it's not the best practice, which is that every station should let you get a card, from a machine rather than a person. (Ironically, New York does just that with the MetroCard.) I have no idea why otherwise good systems don't do that - e.g. Singapore requires you to buy and recharge EZ-Link from a person, or at least did in 2005.

      And Anon, the point of federalizing this is that there's no need to ask an MTC to change things. Just put in one standard, and issue cards and let independent vendors stick the chip in a watch or keychain or cellphone. Transit agencies should be worrying integrating fares, not about managing their same-as-cash-but-faster payment options. There's an alternative, yes, but then you get Japan's situation, where not only are fares almost never integrated, but also your JR West card can't be used in Hokkaido even though JR West and JR Hokkaido both have bilateral agreements with JR East and they all use the exact same technology.

      Delete
    24. I found myself at North Station after checking in a bunch of shops and smaller stations. Sometimes customer service reps will have cards, but they were all hiding behind a glass bunker and not willing to respond. So I remembered someone telling me that sometimes they stack piles of CharlieCards on top of the machine. I reached up there, and found a single CharlieCard sitting there, but it was clearly not new. In fact, it had $13 on it. So, that story had a happy ending :)

      I had originally been looking in shops because, in their infinite wisdom, the MBTA makes it only possible to load a weekly pass onto a CharlieCard from the machine they give out to 3rd party stores. It's not possible to do it using the machines inside of stations. It's a frustrating thing that I finally had to learn by asking some random person how they managed to get a 7-day pass on their CharlieCard.

      Delete
    25. @Alon, the situation in Japan is not technology limited (as you inferred), but due to the fact that there are so many railway and bus companies, many that are fiercely independent, often are competitors with each other (esp. in Kansai or Tokai), or are reluctant to invest in interchangeability that may provide little benefit to themselves. For example, JR West has relatively few Kansai based business customers that would use their card in Hokkaido, and vice-versa, compared to the large business travel market between Tokyo and Sapporo- an overwhelming number of Suica users use the card for the (profitable for JR Hokkaido) Airport Rapid service, and little else. Tourists in Hokkaido who use the metro area trains covered by Kitaca are few in number, anyway. Regarding fare integration, there is fare integration between JR companies- they use the same online reservation system, MARS. Smart cards are intended for local and commuter use, as they have a two hour time limit per entry/exit. Intercity travel is done with conventional paper tickets or a separate IC card system on the shinkansen, which has a different fare structure from conventional lines. Private railway companies have little to no incentive to have fare integration with JR, especially if they are competitors. That said, true nationwide compatibility is on the horizon, once the smaller operators and holdouts get on board.

      Delete
    26. as you (correctly) inferred...

      Delete
    27. London may be fascist about the use of the chips, but OysterCard (1) is available in every Tube station and most National Rail stations in London; (2) is refillable in *every* station, excepting bus and tram stops; (3) actually has a coherent fare structure and implements it properly; and (4) since TfL started pressuring everyone else involved, basically works on any service in London. (TfL had already forced them to use the same fare structure, because London's had the power to do that for ages.) Oh, and (5) you can get multi-day TravelCards on Oyster in any station with a ticket agent, the same conditions where you could get paper TravelCards.

      Contrast the insane implementations we've seen throughout the US, where it has been made substantially more difficult to use the "smart card" than to use paper tickets.

      --Nathanael


      --Nathanael

      Delete
    28. Alon, the problem with setting a standard on a federal level is that you have to set the standard before there are many implementations, and then how do you know that the standard is good enough? Transit fare collection can be a bit complicated, especially if you want to implement things like being able to add value online, and having card readers not permanently connected to the network (because they're on buses, for example). It just seems like the better solution is to have the local agencies implement it, and leave regional bureaucracies like the MTC to just force the local agencies to all agree on some standard and interoperate.

      Delete
    29. Like ISO/IEC 14443, nah couldn't do that.

      Delete
    30. According to Wikipedia, Clipper uses Mifare DESFire, which is compliant with ISO/IEC 14443. As do Orca and CharlieCard. So, we have perfect standardization and interoperability, then? That standard only addresses things like the physical and link-level protocol, and probably how to store data on the cards. The difficult parts are what data to actually store, if you want to have a universal standard farecard. And there of course there's the backend to deal with, if you want things to interoperate on a large scale.

      Delete
    31. Adirondacker1280027 June, 2012 13:20

      How come I can't use my Sears card in Walmart? Why doesn't my supermarket loyalty card work at the drugstore? How come the receptionist at my doctor has to scan my insurance card and my debit card? Why can't I scan my debit card to get loyalty discounts at the drug store that won't recognize my supermarket card? Why should the NYC subway care if there are people in the Bay Area who want to use their Metrocard to ride Muni?

      Delete
    32. Transit agencies in different cities aren't competing with each other. Nor are their smartcards about loyalty plans; that's what unlimited monthlies are for, and there's a huge difference between having a single standard that works for everything and a season pass for New York that also works in Boston.

      Think of it this way: how come I can use Georgia quarters in Rhode Island?

      Delete
    33. Adirondacker1280027 June, 2012 17:16

      Feel free to use your Rhode Island quarters to pay your bus fare in Georgia. I betcha there's even way to use quarters to pay your MARTA fare.
      Doesn't being able to use your Rhode Island quarters to pay bus fares when you are in Georgia solve your problem?

      Delete
    34. Not really. It means I need to carry correct change, and conduct a much longer transaction than just tapping a card.

      Delete
    35. Adirondacker1280028 June, 2012 16:47

      So how much should the government spend to let you avoid a minor inconvenience? If it's so terribly inconvenient feed the quarter to the card vending machines that will be at the airport and train station.

      Delete
    36. There are lines at train stations. It can be a couple minutes, for no good reason. Remember, the economics of having the feds step in is dramatically different from that of a transit agency doing the same. For a start, if the system hooks into Eagle Cash, it can build upon the defense budget, which habitually pisses multiple billions of dollars per year on redoing completed bids and doesn't even notice the money is gone.

      Delete
    37. Adirondacker1280028 June, 2012 19:43

      The good reason is that your time is worth less than the the cost of the IT project to make it work.

      Delete
    38. My time times how many people in the same situation who'd end up hogging ticket machines?

      Delete
    39. Adirondacker1280029 June, 2012 02:56

      So they install another ticket machine at the airport and another one at the train station. TVM's are moderately cheap. And you have the option of using your quarters directly on the bus.
      How much of a surcharge are you willing to pay to make your card usable?

      Delete
    40. Nobody wants me paying quarters on the bus - not the other bus passengers (even under POP, an excessive line could form), certainly not me.

      As for TVMs, most don't even sell smartcards (I always used magnetic stripe cards on BART). Yeah, they can get more, but the cost adds up. Add up also the cost of doing the same smartcard implementation in each city separately and you get up to a lot.

      Delete
    41. Adirondacker1280029 June, 2012 15:07

      Everybody used quarters on the bus until a few years ago.
      How much of a surcharge are you will to pay to amortize the never ending IT project?

      Delete
    42. Probably would make me a lot likelier to actually ride the bus, and also speed it up via faster boarding so that they'd need fewer operating hours for the same frequency.

      Delete
    43. Adirondacker1280029 June, 2012 17:24

      How often do you ride the bus in Atlanta? How many out of towners arrive in Atlanta without their cars? How much does them using quarters instead of a pass slow down boarding for Atalantans? If you arrive in Atlanta without your car how likely are you to rent one instead of riding the bus? Me being able to use my Sears card in Walmart would speed up the checkout line. Not having to carry around 12 loyalty cards would make my keychain much lighter. Using my smartcard to unlock the car and the front door would make a keychain unnecessary. Walmart doesn't find it economic. Nor do I for my car or front door. How much of a surcharge are you willing to pay to be able to not use quarters on your infrequent out of town trips? Or how much are you will to pay to carry one card instead of two on your frequent out of town trips?

      Delete
    44. Not that American retailers ever do this, but you can use your Suica at various stores in Tokyo, and I won't be surprised if they include department stores run by the private railroads, which want you to use PASMO instead. Electronic money is not the same as loyalty programs.

      Delete
    45. Adirondacker1280030 June, 2012 20:32

      I never have to pay to use my front door or car door.
      How much are you as someone who never leaves New York City, willing to pay so someone who has a Greater Glens Falls transit card can swipe on the bus in Altoona? Or New York City?

      Delete
    46. Well, if the side-effect of Greater Glen Falls and Altoona being compatible is that LIRR and NYCT are compatible, I'd be willing to deal with the side effect of other people's mobility being improved, yes. It's not as if we're *not* paying for new farebox systems everywhere already, so paying for compatible ones doesn't seem like an obviously horrendous and plainly useless extra expense.

      Delete
    47. Adirondacker1280002 July, 2012 17:44

      NYC Subway and the LIRR are already compatible. You swipe your Metrocard to get on the subway and when you get on the LIRR train you take your ticket out of the same place you stash your Metrocard. Most people can deal with the concept of there being different cards for different vendors.

      Delete
    48. That's the exact opposite of compatible. For a place where fares and tickets are actually compatible, transfers between the M├ętro and RER are free, and if you keep your t ticket (or use the smartcard) you can insert it into the next system's turnstile when you transfer and it'll work.

      Delete
    49. You don't even have to swipe anything or go through a turnstile to use NYC's RER, the trains are right across the platform at the express stations.

      Delete
    50. You have to exit the system and pay a separate fare to use the LIRR.

      Remember, the point of the RER is not just that it runs express. It's that it uses mainline rail to give the suburbs reasonable service, interlining through trunk lines in urban tunnels to create high-frequency urban service. Same thing the LIRR has with service to Jamaica.

      Delete
    51. Adirondacker1280004 July, 2012 06:59

      You have to exit the Metro or RER to use Transilien just like you have to exit the subway to use the LIRR/NJTranist/Metro North/PATH. Or exit the system for your flight to Tokyo.

      Delete
    52. "As for TVMs, most don't even sell smartcards."
      And there's the problem right there.

      Nobody is going to switch to smartcards until they are just as easy to get as the old "tokens" which subway systems used after they stopped directly taking coins.

      This is just OBVIOUS.

      --Nathanael

      Delete
  14. can't you manage to copy one of them?

    Being able to get from Kowloon to Central or Barking to Wembley isn't particularly useful if you are in San Bruno.

    ReplyDelete
  15. Marc-okay, I haven't taken Caltrain @ Palo Alto for a while, I was just looking at the Clipper website. But seriously, they need to be all over VTA and Caltrain stations. You shouldn't have to go find a Walgreens just to load value on it. Sorta inconvenient. It is not perfectly usable when you are on foot and need to catch trains and make connections.

    ReplyDelete
    Replies
    1. You can also reload over the Internet or over the phone.

      Delete
    2. Re: "You can also reload over the Internet or over the phone."

      You've never actually tried that, have you?

      You know, the three day waiting period for your funds appear and all.

      Why not fill in an application form and show up in person at a consular station as well?

      I can't believe anybody even attempts to defend this clusterfuck. Clipper is absolutely horrible for everybody except Cubic, the sole-source no-compete vendor.

      Delete
    3. Reality Check23 June, 2012 22:54

      Why can't those Clipper f*ckers at MTC/Cubic figure out how to make your funds not disappear off the face of the earth for 3 days? The World's Finest Professionals indeed!

      Delete
    4. Because their core competency is elsewhere.

      Just like Boeing's core competency is making planes and not LRVs, and it shows.

      Delete
    5. Add a "not invented here" unwillingness to learn best practice.

      There are organizations which can seamlessly move into any business because the people running the organization are willing to do the scutwork to learn the new business well enough to hire the right people, and willing to spend enough money to hire those people, and then willing to supervise it until it gets done right. These are rare organizations.

      Then there are attitudes like "We're Boeing, a train is just like a plane, right?". Or "We don't need to see how anyone else implemented this, because we're the greatest transit system in the world / our city is unique".

      --Nathanael

      Delete
  16. Fare collection is the core "competency" of Cubic. That's their entire business.

    ReplyDelete
    Replies
    1. Going by Wikipedia, Cubic has 8,000 employees, of whom only 1,500 are employed by Cubic Transportation. So most of what it does is various support systems for the military.

      Delete
    2. Not to split hairs, but "Cubic" is the parent corporation and "Cubic Transportation Systems" is the subsidiary. Cubic Transportation Systems has been in the transit farebox business for 35 years (at least that is what their web page says).

      Delete
    3. So "rigging contracting processes" and "rent seeking" would be the Core Competencies of the parent military-industrial corporation; while "lobbying HARD against Proof of Payment" would be a sub-speciality of the Ripping Off Transit Agencies subsidiary.

      See, you CAN have it both ways! Military AND industrial. Useless AND expensive. Capitalist AND welfare recipient.

      Delete
    4. Keep in mind that Cubic only took over Clipper a couple years ago. For the decades before that, it was developed by an Australian company called ERG. They are also responsible for the Myki system in Melbourne, which has been, if anything, even more disastrous than Clipper, partly because it replaced an already existing integrated ticket system that was working just fine.

      Delete
  17. What Caltrain needs right now: More frequent service in SC County, and better reverse peak in general. In fact, implementing this efficient 90-train schedule (AM, PM) can have multiple benefits.

    - First thing to do: Change the current zone fare system. Probably reduces ridership by a few thousand riders.

    - Reduction in service earlier hours: Not enough riders are up that early.

    - Usage of 2- or 3-car consists to account for lower demand and an improvement in acceleration.

    - Traditional Peak: Baby Bullet Service reduced to 4 traditional peak trains: strategically placed to add capacity and get riders to the SF CBD. 6 intermediate stops, each with 45 sec dwell times, ensures a very nice 1-hour trip time. All other trains make local stops in SC County to provide 3 tph service. One train makes all local stops, the other 2 make limited stops in SM County to get riders to SF faster. Improved service from 8:00-10:00 A.M. and 7:00 to 9:00 P.M. to account for riders that commute during shoulder peak. BB trains do not overtake at Bayshore and slow down SM County local riders anymore.

    - Reverse peak: Service revised to 4tph, 2 limited, 2 local. 4 tph is enough to serve the demand, and the BB trains stop at all the job centers in SC County.

    - Atherton and College Park permanently closed due to lack of ridership potential. Hayward Park closed due to schedule limitations and proximity to both Hillsdale and San Mateo. Broadway temporarily not served until after electrification due to schedule limitations and proximity to Millbrae. Tamien is served by 10 trains in the traditional peak direction only.

    - All other stations have 2 tph, except Bayshore (and 22nd St to a certain extent) in the traditional peak due to schedule limitations and lesser demand.

    ReplyDelete