How the “Small Stuff” Can Wreck the Best Intentions

A comment arrived from a regular reader, David Crawford, about how the TTC mishandles the details of its service and information. This really belongs in a new thread, and I’m sure it will spawn a trove of examples from other parts of the system.

The TTC is, probably necessarily, a very complex organisation but seems to lack the ability or ‘culture’ to look at projects or services as a whole. This is apparent in major ‘conflicts’ between priorities where the priority to reduce overtime results in short-turns and ‘maintaining the schedule’ rather than thinking of the priority to provide ‘good customer service’ and thus ‘running to headways’. This inability to ‘think things through’ is also apparent in much smaller ‘projects’.

More minor examples are the strange absence of links to NextBus on the TTC schedules pages (because the ‘schedule folk’ don’t talk to the “Nextbus folk” or resent Nextbus?) and in the recent temporary splitting of the 72 bus route.

Splitting the route may have been a good idea but giving it a new number (172) led to (unplanned) problems. Many bus stops had the Route Number painted on them, someone had to go out and add a “1″ before the “72″, the Nextbus displays at the King Station were not working for months (and may still not be), someone was sent out to fasten new 172 route schedules to poles, but was not told to remove the ones for the 72 – which no longer ran to those stops. In addition, because the 72/172 is on a ‘temporary’ routing downtown there were also large signs along the route noting that the 72 ran there – these too had to be changed to read 172. Furthermore, there are (old) buses that do not seem to be able to display a route 172 sign so they still actually run as 72s – though are, in fact 172s. There were also references on the TTC website to the 72 bus – getting to the Distillery for example – when the new (temporary) route was the 172.

It would have been far easier and less confusing if they had simply created another branch of the 72 that would only run between King and Commissioners. When the route numbering was changed it would also have made sense to have bitten the bullet and officially changed the routing so that the bus will go permanently to the King and St Andrew subway stations. This temporary change has clearly resulted in far more ridership as the 72/172 is now a very viable alternative to the (too often short-turned) 504 streetcar.

The lack of direct links between the TTC schedule pages and Nextbus is very annoying, and you have to know your way around to use the information.  That said, there are a few challenges in presenting the information.

Both the desktop and mobile versions of the site provide the “stop number” which can be used with a text message to retrieve predicted arrival times for vehicles.  The desktop version includes a clickable “?” that is supposed to call out a help panel, but on the 504 display I tested, this has no effect. I believe the idea is that the pop-up should tell you how to use the stop id in a Nextbus call.

There is no reference to Nextbus on the general Schedules and Maps page, but it is included with much other info on a Connect With the TTC page.  If one is using either a desktop or a smart phone, it is not necessary to know the stop number as there are other paths into Nextbus, but their availability is not explained by the TTC.

Other apps, of course, use the same data feed and have their own benefits and drawbacks, but they all depend on having something beyond a basic cell phone.

As for the 172 and the Pape diversion in general, the truly annoying feature was that it was so prominently advertised in King and St. Andrew Stations while information about the 509 Harbourfront bus (a much more frequent and well-used service, especially by people unfamiliar with the area) went for a long period with only minimal signage. Visitors from other planets might be forgiven for thinking that 72 Pape / 172 Cherry are vital routes without which the city would collapse.

The problem for the TTC is inconsistency.  One day a change or diversion will get saturation coverage, another day it’s a small paper notice if you look in just the right place.

As for the 504, yes, it’s a mess. I am working on a large collection of vehicle tracking data for a series of articles on the evolution of service on this route over past years. What is abundantly clear to anyone riding the system is that managing for headway is a myth, and short turns to keep operators on time are more common than ever. Meanwhile the King route supervisors cluster outside the John Street Tim Horton’s in an obvious display of wasted manpower.

17 thoughts on “How the “Small Stuff” Can Wreck the Best Intentions

  1. Steve said:

    “As for the 504, yes, it’s a mess. I am working on a large collection of vehicle tracking data for a series of articles on the evolution of service on this route over past years. What is abundantly clear to anyone riding the system is that managing for headway is a myth, and short turns to keep operators on time are more common than ever.”

    Funny, you would think headway on a car as frequent as the King one, would be priority 1 or 2 along with service to the terminus point, yet you seem to be saying both are clearly being sacrificed on the alter of some priority likely effective invisible to the rider.


  2. So Steve, contract award for Yonge VIVA BRT. How excited are you?

    Steve: When I see York Region with a plan for very substantial service improvements, then I will start to get excited. Meanwhile, it’s another construction project. For my readers north of Steeles, this may sound dismissive, but as a general rule I am tired of seeing huge amounts of capital spending with no increase in operating funding from Queen’s Park.


  3. They should just add a tab with “live map” with the route and location of all vehicles/stops updated in realtime…easily doable with the nextbus feed.

    They should also change the “next scheduled” column to “next predicted”…which is really all I care about as a customer…they can schedule 100 to show up every minute, but if there is a delay and they aren’t going to show up till 20 minutes from now I’m not going to go to that stop…


  4. The TTC is very unpredictable.

    I don’t think the 72/172 split on nextBus was that bad. I think they didn’t have route 172 on nextbus for a week, not a month. The signage was another problem. The 72 Pape signs remained up for over a month. Management at division level never seem like they informed anyone that the route is going to be split. Yet they got it right on the new TTC experimental map.

    When the 52 Lawrence West restructuring came in effect along with the new 195 Jane Rocket, they got it right. NextBus had it that day when these service came in effect. The TTC website was updated right away but their system maps still show the discontinued 58 Malton route.

    The whole Lawrence west route restructuring wasn’t advertised so well. 52G to Martin Grove doesn’t specifies if it’s The Westway/Martin Grove or Dixon/Martin Grove. They can at least add to the destination sign, via The Westway or via Dixon since the route next splits in two directions. They led to a lot of confusion on which bus runs on Dixon and to the airport. Any customer service would have pointed this out.


  5. Steve regarding short turns and the 504,

    Recently every morning I have seen cars turning back east at Spadina with a ton of cars also turning back at Parliament or Church.

    Looks like they are trying to keep service going downtown. Has anyone ever thought of a tripper service to the borders of downtown in order to keep service frequent in downtown and between subway stations i.e. 504 Parliament and 504 Spadina?

    Steve: Frequent service in the near downtown may look good to route supervisors stationed there, but it doesn’t do anything for people trying to get into downtown from points east and west of the turnbacks. Once I start to look at the tracking data, I will be very interested to see whether the problem is that the scheduled running time is far too short for what cars really need, or if something else is at work here.

    It’s ironic that the city just implemented extended hours of parking/stopping restrictions but they only affect the shoulders of the peak, not the peak itself.


  6. It would be helpful if every bus or streetcar stop had a “stop number”.

    Steve: Do you mean posted on the stop itself, or available online? There is a big lag between rolling out the stop numbers on the system and actually telling people what they are out on the street.


  7. I was wondering. Before the Yonge subway, the Yonge streetcar (and trailer) train ran up and down from Glen Echo Road (south of York Mills at old city limits) to Union Station. This was before GPS, smartphones, and on-board communications. I was a toddler then, so don’t remember or cared about headways or schedules. I do remember the drivers having to stop the streetcar, and running to a pay phone (using a dime from the coin dispenser) to get in contact with control.

    Didn’t they have the same problems today with emergencies, accidents, breakdowns, etc.? Shouldn’t we have the capability to handle such incidents better these days? Maybe the TTC should go back to the “olden” days and start all over again. We have the hindsight today, so we should do better.


  8. I think there are two things going on to cause such apparently easy-to-correct problems to be uncorrected.

    One is simply the organizational structure that the TTC uses, which is the old command and control style where you are told exactly what you are supposed to do, and you do it. You are not paid to think about what you’re doing; neither is there any expectation on the part of management that you will come up with any ideas for improving things and passing them up the chain. This is actually the old industrial model used by car manufacturers. The “new” model, as shown by Toyota, involves the worker and expects that workers will come up with improvement ideas and pass them on. The workers, likewise, know that this is expected of them, and the ideas will be acted on.

    I have been reading several books, by Demming and others, about the struggles of the North American car industry to remodel itself. When I read about how the worker was expected to do as told, I immediately saw a connection with the way operators are supposed to run to schedule, and not to question or take initiatives. Of course never having worked in the TTC, this is amateur organization behaviour analysis only.

    Changing the whole underlying setup is a hard job, perhaps impossible. It certainly can’t just be done from the top, if it can be done at all. And it is a bigger problem than just the TTC’s–General Motors has been trying to re-invent itself how many years? Even went bankrupt and got bailed out. And is still struggling to remake its corporate culture.

    The other problem causing this is the “cost savings” and “efficiencies” being forced on the TTC by the Commission and TTC’s upper management (which, to be fair, has to go along with the Commission, or leave (as Gary Webster demonstrates). The TTC can make the job of putting up route change information “more efficient” by giving the job to one person instead of two, giving that person more changes to make in a day, and so on. Perhaps the TTC worker who made the changes and put up the signs knew there were other things that ought to be done, but they weren’t part of the day’s work orders and the worker had plenty more to do that was ordered. Perhaps the TTC worker was working in such a rush that there was no time to notice the other things that ought to be done. It’s all very “efficient” and “cost conscious” until all the little details that are so inefficient and costly to take care of get skipped, and pile up in a most obvious and inconvenient way.


  9. These are all great points. I can’t help but feel, however, that the problem is simply that upper management does not care. If they did care, they could issue the following directive to all drivers immediately for zero cost:

    If you see the run ahead of you as you arrive at a terminal point, you are to hold at the terminal point for one scheduled headway unless otherwise directed by Control, on pain of being disciplined if you leave earlier than required by this directive.

    Now, I don’t claim this is a great solution, or anywhere near complete, or even a good attitude: do this or be disciplined. Something much more collaborative with the drivers on the street who know what is really happening would be better. And the need for more investment in the system is clear.

    But the point is that it is offensive for parades of vehicles to travel back and forth all day, and it would be trivial to break up the parades at terminal points by simply ordering the drivers to hold until there is no parade. Perhaps the parade would re-form by the time the vehicles reach the other end of the line, but at least people near terminal points would have somewhat better service. I’ve mentioned before, in jest, the idea of community protests against bad service which would consist of lying down across the tracks in front of the exit from a loop and permitting streetcars to leave only on headway. In that form it’s not a serious suggestion, but it really would improve service, probably on the whole line.

    If even simple measures won’t be taken, there is no reason to believe that spending a bunch of money on some new vehicle tracking system (or any other improvement) will result in any improvement: the new system has to be really used in order to do anything.


  10. “I’ve mentioned before, in jest, the idea of community protests against bad service which would consist of lying down across the tracks in front of the exit from a loop and permitting streetcars to leave only on headway”

    You know, I’d kind of love to see this. It would be the weirdest and cleverest protest ever. It would require people with well-timed watches to coordinate, as suddenly all the protesters get up and out of the way and take a break while the streetcar leaves on headway.


  11. Tess Kalinowski had an interesting article in the Toronto Star today discussing the “Big Idea” of expanding express bus service (which has been called for by Coun. Colle as well).

    It is important to note that, whatever the technology and vehicle mode and service type may be, line management, legacy and culture are still big issues that affect service.

    I think that it is strange that the TTC has 3 different express bus networks and would rather see them consolidated into one Rocket Network … but if the line management issues aren’t going to be managed, it won’t make much difference beyond being a little less confusing.

    Cheers, Moaz

    Steve: I had thought of writing about express buses, but was preoccupied working on the Neptis review. What I am waiting for is a TTC report on the economics of the express routes, especially the premium fare services. The TTC does not report performance data for these routes, and it’s my gut feel that they are losing their shirts here. The vehicles could be better used on regular fare services elsewhere. But I will sit back and wait to see what sort of numbers the TTC claims.


  12. I drive up and down Dufferin St in both peak periods to/from home and work and this morning I saw 3x artics in a row southbound 8:00am @ Rogers Rd. First was packed, next two were empty. What a joke. Also, back in Feb when Eglinton Ave between Duffern and Allen really got bad I saw 12 (yes 12) 32’s between Marlee and Dufferin. You could crawl on your knees faster than ride any of those buses.


  13. While you mention NextBus: Given that the NextBus feed is already becoming the canonical source of status data for the TTC network, adding other volatile data useful to riders like status of escalators and elevators — at ride time rather than only at planning time, and information that can change while in transit — the NextBus XML feed would be a useful place to present this information (in addition to the existing ones, of course).

    A quick scan of the API docs (specifically the ‘Messages’ section) suggests that the data feed could presumably accommodate information as to the status of elevators and escalators for each station stop. Although this would best be integrated into specific feed elements for the station stops, this could be a useful interim approach until NextBus iterates the API to handle volatile accessibility information.


  14. I’d like to see more universal placement of stickers on the stop poles indicating the Nextbus code. Only select downtown routes seem to have them. Outside of the downtown, one has to root around to find that they are listed on the TTC website.

    The other problem I’m discovering is the disappearance of a vehicle — sometimes temporarily from the GPS and sometimes altogether. This is most annoying with the Yonge 97 which has now been split and service cut on the part of the route with a large number of apartment dwellers/transit riders, that is Eglinton to Lawrence. When vehicles are coming, at best, once every 20 minutes and a 97A/F disappears from the GPS it’s frustrating. Where has it gone? Is it just stuck in a tunnel so it’s just that the GPS signal has been blocked for a few minutes or has it been short turned?

    Last night I thought I was waiting for a northbound 97B that was 5 minutes away. After an extra 10 minutes I checked and then the wait was 38 minutes. Thank goodness I know how to use the stop numbers or I’d have been waiting almost an hour for that 97 bus!

    Steve: The issue of completing the stop pole stickers has come up at TTC board meetings too. I know the plan is to get this done, but when, for the thousands of stops around Toronto, is another matter.

    As for vanishing buses (and streetcars), this can happen in two ways. The most common is for an operator to flag the vehicle as out of service, and this can cause it to disappear from predictions. Also, a vehicle may either not be in the GPS feed at an instant, or may misreport its location (this is most common where signals may be shielded by structures, or downtown where reflections from office towers cause some streetcars to claim that they are in the middle of the lake or at their cottage in the country).

    One basic thing I have to do when digesting this data is to filter out the bogus readings and interpolate between what remains. I have the benefit of looking at data months after the event, and it is easy to distinguish between a vehicle that simply went astray briefly and one that definitely went out of service. Nextbus has to make this decision in real time before it might actually know what is happening. I believe that they err on the side of omitting vehicles in predictions if they cannot be certain the car or bus is still actually going to arrive at the stop.


  15. 007 said:

    “I drive up and down Dufferin St in both peak periods to/from home and work and this morning I saw 3x artics in a row southbound 8:00am @ Rogers Rd. First was packed, next two were empty. What a joke.”

    Well at least it would appear the one Artic was large enough to carry the load. Begs the question of how much more ridership would appear if the real headway (the one experience by riders not the theoritical average). I know for myself: figure total trip time, wait time at a stop and stuck in traffic time counts twice as much time in motion, especially when there are large variances.

    There must be some reasonable way of controlling this type of thing. This generates a large needless harm to the rider, and means that a large amount of money is being expended while not receiving the associated revenue. Means on this line at this time probably resulted in a real headway of 30 minutes where the intended one was 10.


  16. At least there is a link to NextBus on the site. There is absolutely ZERO reference to Google Maps ANYWHERE on the TTC’s webpage, with the exception of a press release from a few years ago! Search “Google Maps” if you don’t believe me.

    I had a Twitter conversation with @TTChelps a few months back. He said that the TTC has its own trip planner. I pointed him to where they not only have their own trip planner, but Google Maps AND NextBus in tabs right next to it! He thanked me for the information and that was that. Real customer service eh?

    Sometimes I think we should allow other transit systems to operate in Toronto, just to give the TTC some much needed competition in order to get them to be less stubborn.


  17. I note that on the TTC’s list of bus routes they have it broken down into 4 categories … the long list, blue night, community buses and downtown express routes.

    I wonder why they don’t break out the Rocket buses into a separate list from the long list. They could also highlight routes with “Frequent Service” sections.

    Cheers, Moaz


Comments are closed.