The TTC Board met on February 21, 2017. The following items of interest were on the agenda:
- CEO’s Report
- VISION Service Management Program Update
- Track Reconstruction on The Queensway
- Donlands Station Second Entrance
- Customer Liaison Panel
- Customer Satisfaction Survey
- Electric Bus Charging Stations
- Transit Museum Proposal
Introducing his CEO’s Report, Andy Byford noted that the 2017 edition of the Customer Charter had been launched including 37 “time bound commitments” for improvement. This all sounds quite good, and each year TTC management get to check off most of the boxes on their “to do” list. The problem with such lists is that they can be very inward facing, self-congratulatory exercises where “achievement” consists of little more than waiting for the sun to rise tomorrow.
Some items simply require than planned events occur, and in many cases the timing is beyond the TTC’s control once the wheels are set in motion with a contract to a third party. The TTC expects to receive hundreds of new buses and a few dozen new streetcars in 2017, but actually achieving this is largely in the hands of the manufacturers.
Some items involve the roll out of new procedures such as new Wheel-Trans eligibility criteria or the conversion of Station Collectors to “Customer Service Agents”. The problem with this type of goal is not just that it happens, but that it happens well. Are the new processes and duties well thought-out as improvements to the system and for the riders, or to suit a management plan such as cost control?
Some items confer little visible benefit to anyone such as the plan to “convert an additional 3000 bus poles to the new design”. The most useful information at a bus stop, the schedule, has been removed by the TTC on the assumption that most riders have a smartphone with which they can retrieve this information. As a consolation, the TTC plans to “pilot solar-powered passenger information displays to provide real time information on vehicle arrivals at stops without utility power”. That will give them some kudos for going green, and it will simplify installation because they will no longer have to depend on a utility hookup. However, the real benefit would be in a large expansion of the number of displays actually installed and working.
Byford acknowledged that the scorecard is disappointing, but claimed that the underlying performance is good including on a year over year basis. He is aiming for either a “green” or upward trend, and said that January stats, to appear in his next report, are demonstrably better than December’s. One cannot help being skeptical about such claims, but this begs two other questions:
- If there is a year-over-year improvement, why are the majority of the scorecard entries red? They were not a year ago. What has changed? In a few cases, the target has been reset higher, but then was missed. However, most of the targets are the same now as they were a year ago.
- In some cases, the target is based on a percentage improvement over the previous year. This will inevitably run into the classic “efficiency” problem we hear about in budget debates: there is an easily achieved first round of improvement, but replicating that is much more of a challenge. Also, some improvements are achieved by relaxing the environment in which the target must be achieved such as the reduction of short turns by provision of more generous schedules.
The report gives a black eye to the concept of target achievement with an “all red” status for the Key Performance Indicators (KPIs) related to Customer Service.
A few of these are near misses, such as the trains per hour value for Line 4 Sheppard, but most KPIs miss their mark by wide margins. There was a discussion of delay types by the Board members, and it was quite clear that management has a more finely-grained breakdown of delay statistics. An important consideration is that some delays are “controllable” in the sense that actions by the organization (e.g. maintenance quality) can affect them, while others are “un-controllable” (e.g. passenger illness incidents). One might argue that there is some overlap in that crowded, irregular service creates conditions where passengers are more likely to suffer panic attacks or faint, but this too requires that the stats be broken down to co-relate the time of incidents with conditions that could have contributed to them.
Every time service reliability stats come up for the subway, members of the Board ask for a breakdown of the types of incident. What they get is an ad hoc verbal report. This information should be included in the CEO’s report, as least periodically, so that it is clear where the problems lie.
The issue with many types of delay, even those beyond the TTC’s control, is not simply to count their number, but to review how well the organization responds and what might be done to reduce the effects on riders. For example, knowing that crowded peak period trains trigger health problems, the TTC now stations paramedics at key points to minimize response times to such incidents. Would that a response at the political level to the lack of system capacity were as simple.
Among the commitments made in the Customer Charter is the provision of statistics on the reliability of all surface routes. In fact, no stats have been published since the first quarter of 2015. This situation was originally caused by a change in the metric used by TTC management to report line reliability with the two new measures being on-time performance at terminals and the count of short turns.
Customer satisfaction numbers rebounded in the fourth quarter of 2016, and the summer dip was directly a result of the cock-up with failed air conditioning units on Line 2 Bloor-Danforth. The amazing point is that even with that huge disincentive to ride the subway, that the TTC managed as good a rating from Line 2 riders as it achieved. The CS report results appear in a separate section later in this article.
Rick Leary, who is responsible for surface operations, made a short presentation about how performance has evolved. This information is actually more informative than the KPIs in the CEO’s report because the KPIs are consolidated at too high a level to allow meaningful interpretation.
Short turns on bus routes have been substantially reduced since 2014, although the stats in the chart above do converge toward the end of 2016 and early 2017. This may indicate that the relatively low hanging fruit has been picked, and that further reductions will be less dramatic and focused on problem routes.
Streetcars saw their big drop from 2014 into 2015 as schedules were adjusted, and then further into early 2016. However, the effect appears to have leveled out by fall 2016.
On the surface system as a whole, six routes account for 40% of the short turns according to Leary. This suggests it is time to publish tracking numbers on these problem routes rather than diluting their condition with every other route in the system. Of course, if the TTC were publishing route-by-route numbers as they claim in the Customer Charter, we would know these details already.
Running time improvements have been implemented on some surface routes, but the effect on on-time performance has been minimal. The problem, as shown in many route analyses published on this site, is that on time departure from terminals is rare regardless of the schedule. Indeed, if there is too much “padding” in the scheduled travel time, there is an incentive to take a longer-than-planned break at the end of the line and make up the time enroute rather than plodding across the city. Running time changes should produce a reduction in short turns, but these stats were not shown.
Construction both for the Spadina subway extension and for water main upgrades on Queen Street took their toll on surface routes. Some schedules were adjusted to compensate for these conditions, but service suffered nonethless. Of particular note was the spillover effect on King of the Queen diversion for which no provision was made in the King schedules, nor in any transit signal priority adjustments.
The work on Queen, contrary to this slide, was not “unscheduled”, and the TTC made provision in schedules for this beginning on May 8, 2016. However, the project ran long, and the “standard” schedules were reintroduced on October 9 before work actually completed. The chart above shows various effects:
- On time performance sat at the 40% level through much of 2015. On this schedule, through service was provided from Neville to Long Branch.
- In January 2016, the route was split with the Long Branch section (aka 501L) running independently of the main part of the route except for late evenings and overnight. Running times were adjusted to better match reality, and the on time performance shot up into the 70%+ range.
- Concurrent with the onset of the diversion, on time performance fell back to 2015 levels even though the schedules had “extra” running time. This suggests that the “extra” was insufficient.
- On October 9, the reversion to the standard schedule shows some improvement, although this was also concurrent with a cutback of east end service to Woodbine Loop for construction at Neville.
- Service resumed over the standard route on December 10, and on time stats are back up to early 2016 levels.
Sidewalk reconstruction in the same area is planned for 2017. Sidewalk rebuilds should not trigger transit diversions, but we shall see.
Updated Feb. 27, 2017 at 12:05 pm: The schedule changes effective March 26, 2017 have just been released by the TTC and they include a resumption of the 501 Queen diversion between Spadina and Shaw. This is expected to continue until October 2017. Quite frankly, it is appalling that this project will take so long and disrupt streetcar service again. Whether the City will do any better at transit priority for this diversion remains to be seen, but I am not hopeful.
[I am currently working on articles about the Queen streetcar and bus operation, both to compare the modes where there has been a replacement in 2017, and to look at the overall route operation through 2016.]
Other projects that will cause diversions in 2017 are:
- Watermain and track replacement on Dundas from Yonge to Church. This will include complete replacement of the triangular intersection at Victoria and Dundas Square. Whether the granite paving setts on Dundas Square will be retained, I do not know, but will inquire.
- Track replacement at Dundas and Parliament.
Diversion of the 505 Dundas service is planned to start on March 26 and continue to October.
Spillover effects are expected this summer from the demolition and replacement of the eastbound ramp from the Gardiner Expressway at York Street. This will likely cause some traffic to move north to King (as occurred for other Gardiner projects) and could complicate the operation of the Spadina & Lake Shore intersection as more traffic will be attempting to exit at this point. If traffic signals are adjusted in an attempt to handle greater flow, this will make this difficult location even worse for 510 Spadina.
Also in 2017, Wellington Street will be under construction for watermain replacement and track reconstruction. This will choke one of the parallel streets to King, albeit westbound only.
Vehicle Reliability and Deliveries
Vehicle reliability numbers, like the service-based set above, do not look good compared to targets with a few exceptions.
The T1 trains (Line 2 Bloor-Danforth) have improved substantially, but for the TR trains (Line 1 Yonge-University-Spadina) the stats continue to fall well below the target value. Several subsystems including cab doors, passenger doors and brakes are the subject of ongoing troubleshooting and modifications (see report at p. 44).
Streetcar stats also are below target, and the best that can be said for the Flexity fleet is that the failure rate has levelled out over much of 2016.
Bus stats are improving because some old vehicles have been retired, but mainly through changes in maintenance procedures.
Part of Rick Leary’s presentation above was a chart showing the number of change-offs for buses and streetcars. This is a swap-out of an in service vehicle that has broken down and cannot be quickly repaired and returned to service on street.
The numbers for buses have been improving reflecting the retirement of the most cantankerous of the older vehicles and an increase in the ratio of spares to scheduled vehicles. (More spares mean a better chance that buses will not be pushed out the door just to “make service” only to break down.) The situation for streetcars is not as good especially considering that the number of scheduled cars has been reduced in 2017 and yet the change-off rate is roughly the same as earlier years when more cars were on the street. Leary noted that although the Flexitys do suffer breakdowns, they are easier to fix and this can be done on the street. That said, the “repairability” numbers need to be cross-referenced with the “mean distance to failure” stats that consistently fall well below the TTC’s target. Either the target is too high, or incidents that actually have a minor effect on service are being counted against the new cars. This suggests that a different metric or better detail is needed. (Does a “failure” result in a long delay including the need to push a car back to the carhouse, or is it a “quick fix” situation?)
Bombardier streetcar deliveries continue at a glacial pace, although this is expected to improve as new production capacity comes on line in mid-2017.
The first two cars for 2017 have been delivered. 4431 has been in service for a few weeks, and 4432 is in its acceptance test phase. On Saturday, February 18, all 31 of the new cars were in service either on scheduled runs or as extras. The typical high point for the in service count has been 25 in early 2017, although this varies from day to day.
That said, the delivery schedule for much of 2017 is unimpressive and the length of time needed to “ramp up” quantities suggests a rather leisurely approach by Bombardier. So few cars will arrive until the fall that they will have little effect on the TTC’s ability to provide service with the old fleet’s declining reliability. They will depend on construction projects that replace streetcars with buses to keep the total fleet requirement within what is available. Byford is pushing Bombardier to produce more than 40 cars for 2017, but given past experience, it would be amazing for them to simply hit their target with over 50% of deliveries aimed at the fourth quarter.
Andy Byford reported that the TTC expects to turn on the traction power for the Spadina extension on April 1, and this will allow testing of trains under their own power.
With respect to the Presto fare care system, a software patch to the add value machines appears to have corrected problems they were having. On board card readers are still an issue, but their reliability is improving as the TTC and Presto work through finding and upgrading software on failing units.
Software is in place to support a daily cap on fare charges (effectively providing a “Day Pass”) once usage reaches a threshold, but the TTC is waiting to roll this out until changes for weekly capping are implemented in a planned summer software update. There will be a report to the TTC Board on the potential revenue effects of automatic fare capping that will not require riders to commit, in advance, to purchasing a pass to receive the savings. No doubt this will be yet another excuse for fulmination by those who believe discounts are the devil’s work and that, in their absence, birds will sing in the trees and transit will be a profitable venture.
The King Street transit priority study is now in progress and will be reported to the TTC Board and to Council when there are specific recommendations. Any Board member is, of course, free to read the considerable literature on this that is available online. Changes made to date on King have only been “around the edges” of peak periods to identify locations where extended parking/turning restrictions can yield benefits, but not to address peak conditions directly.
The VISION project started out with the more prosaic acronym CAD/AVL. Thus “Computer Aided Dispatch/Automatic Vehicle Location” became “Vehicle Information System & Integrated Operations Network”. The change of name reflects a shift from the generic mechanics of tracking transit vehicles and assisting with route supervision to a wider range of system operations functions. That, at least, is what the name change implies.
The existing AVL system known as CIS (Communications & Information System) is positively antique and is built on some of the earliest of microprocessor technology. Over the decades, this constraint has limited enhancements to the system through basic constraints such as device storage and the amount of data that could be transmitted in a single message from a vehicle. The “location” technology originally employed was notoriously inaccurate when vehicles did not run close to their planned schedules and routes, and so it gave operators little reason to trust it as an integral part of their operations. (One might observe that some of the foul-ups with and staff attitudes to Presto follow the same track for implementation of unreliable technology.)
In 2009, the method of “finding” a vehicle changed to use GPS resulting in much more accurate and fine-grained position information. (My own programs for analysis of route operations were greatly simplified because they no longer had to filter through CIS data peculiarities to determine where vehicles had actually gone particularly for short turns and diversions that the old system could not track.) The original CIS system plans included analytical functions, but these were never implemented. Only recently has the TTC begun to delve into its mountain of tracking data to learn what is happening “in the field”. Integrating that function in a new system is an important change so that data analysis can be a fundamental part of management, not something grafted on to an old and limited system.
On board vehicles, the VISION system will integrate functions that have been grafted on since the CIS implementation such as the stop announcement system.
The real question behind any technology program, however, is not the name or even the assembly of hardware and software, but the way in which the technology is used. In particular, will there, can there be a change in some of the TTC’s long held beliefs about how service is managed? That will be a critical part of the “culture change” that CEO Andy Byford claims to be leading in his organization.
The operation management function, now scattered across the city in CIS control rooms, will be consolidated at Transit Control in the Gunn Building at Hillcrest. This will give the surface network the same level of centralized management the subway now has. During 2017, the new system will be installed on 15 buses and 2 streetcars for initial testing. Full roll out to the streetcar and bus fleet will be largely complete in 2018.
The TTC hopes to gain several benefits including better management of route and operator performance, improved information for riders, real-time monitoring of vehicle status to track potential failures, and reduced operating costs.
Later this year, I plan to meet with TTC staff to explore how all of the vehicle and service support functions will work and fit together under VISION with the intent of providing a description of the system for readers.
The TTC will rebuild the track on the entire Queensway right-of-way from Parkside Drive to Humber Loop during 2017. The existing ties and ballast structure will be replaced with track in concrete similar to what is used for in street construction with the intent of improving track stability and lowering future maintenance costs. Whether this is strictly necessary is worth asking, but it reflects a shift in focus of to capital investment as a way to reduce day-to-day maintenance efforts.
Other work to be done at the same time includes:
- Reconstruction of the bridge over the Humber River (a City of Toronto project)
- Reconstruction and improvements at Humber Loop (we can only hope that the days of passengers trekking through mud to transfer between streetcars will be at an end)
- Installation of a new substation at Humber Loop
Award of this contract was unusual in that streetcar track projects are normally awarded by the City of Toronto as part of consolidated road, water and transit rehabilitation. This is a large project as track replacements go, and so it has a higher-than-usual cost. On a relatively quiet agenda day, some Board members asked questions about the bids and contractors, and wondered whether the successful bidder could manage the work. If they knew anything about work done over the years, they would know that this company, Sanscon Construction, has been responsible for major track projects around the city and has developed considerable expertise in working with the TTC on this type of project.
Reconstruction of The Queensway east from Parkside Drive to Roncesvalles including all of the in-street special work for access to the carhouse and the intersection at King-Queen-Roncesvalles is planned for 2019.
The Board approved a proposed location for a second exit from Donlands Station, one of several older stations where exits are being added for fire safety. The importance of this report is less the topic than that it marks the end of a very fractious history between the TTC and the community.
When new exits were first proposed for Donlands and Greenwood Stations, the news came out to affected property owners through the media, and there had been no advance consultation. The entire process, including misleading and contradictory claims about “standards” applied to the site selection, came under fire from the City Ombudsman, and the TTC completely revamped their process for undertaking projects of this type. The new exit selection is the result of that process, one which is now standard for TTC construction proposals.
The range of options considered is shown in the chart below (more details are in the report). This allowed the community to work through alternatives understanding the technical constraints, and the TTC to appreciate and react to community concerns about the effect of the work on their neighbourhood. Option E shown below is the one recommended for construction.
In October 2017, the Board passed a motion by Chair Josh Colle asking for a report on the establishment of a Youth Advisory Committee to the TTC. Staff are now recommending that representatives of this demographic group be added to the TTC’s Customer Liaison Panel. This is a rather odd panel set up in 2012 to advise management on TTC issues.
The Customer Liaison Panel was established in 2012 and was formed from a recommendation from the Customer Service Advisory Panel. The Customer Liaison Panel membership is composed of 8 to 12 members and TTC strives to reflect the diversity of its customers on the membership panel. The Customer Liaison Panel includes one standing member appointed by the TTC’s Advisory Committee of Accessible Transit (ACAT). The Panel’s role is to assist in developing and delivering TTC’s strategic aims on customer experience, assist in understanding customer priorities and promote dialogue between customers and the TTC.
One might forgive Chair Colle for proposing an entirely new group because the operation of this panel has been completely out of view since its inception. Information is available on request, but unlike the very public Advisory Committee on Accessible Transit (ACAT), there are no regular public reports to the Board indicating the topics the committee dealt with or what its advice might have been. Moreover, the panel appears to be driven more by management requests for feedback than by issues it initiates, unlike ACAT which sees itself as having a strong advocacy role.
The contrast with Metrolinx is intriguing in that their Customer Experience Committee reports out regularly to the Board. The report itself is a bit odd in that it spends more time talking about management initiatives than actual customer feedback, but there is at least a report.
The challenge for TTC management with any publicly visible advisory group is the difficulty of exploring proposals without turning this into a media and political forum that could compromise the idea of exploring “advice” from riders. However, there is no point of having an advisory committee if its work isn’t published, especially dissenting opinions.
The TTC conducts a rolling survey of its riders by telephone with 1,000 respondents every quarter. The poll is conducted weekly to build up the quarterly total without the problem that a major event (a storm or a major subway outage) could skew the results if it happened to occur before a single large poll.
Values for overall satisfaction in the past two years averaged slightly higher than the previous three, but not by much. As a target, the TTC would like to achieve an 80% rating consistently as this is what is found in other cities with transit systems that are acknowledged to be “excellent”. They bump into this number occasionally, but not on a sustained basis.
Of particular note, regular users of the system tend to rate it more poorly than occasional riders. This could reflect a different demographic or travel pattern, and breaking down that difference would be a useful addition to this report. Are occasional users more likely to be on a high quality service like the subway? Do they tend to travel when the system is less crowded? How is their experience influence by the time and type of trips they make?
If frequent user ratings are often found below 70%, the TTC has a lot of work to do.
Rider satisfaction is also affected by the mode of their travel and even by line when a factor is isolated. Streetcar riders have been less satisfied than those of other modes during many surveys, but not consistently. A key item missing from this breakdown is a sense of the range of values and how tightly clustered (or not) the rankings are around the average.
Subway riders tend to be a more contented lot, but results are not the same for lines 1 and 2 (Yonge and Bloor-Danforth respectively) notably during the summer of 2016 when many cars on Line 2 did not have working air conditioning. By the fall, the line 2 results bounced back, but the TTC now has a large group of riders who are sensitized to shortfalls in maintenance. A repeat performance in 2017 would undermine a great deal of the brave talk of a revitalized, customer focused organization.
When the survey digs into which elements of a journey most affect riders’ experience, three topics come out on top regardless of mode: wait time, comfort and trip duration. It is no surprise that the subway users rank trip duration more highly than wait time as subway trains tend to show up fairly quickly during all service hours. The situation is much different for riders on surface routes where scheduled headways (the time between vehicles) are wider, and arrivals are affected by chronic bunching and not-on-time performance.
Worth noting in the chart below is that most of the key drivers for users of the conventional system did not achieve an 80% or better ranking.
Wait time satisfaction scores are generally good for the subway, but are far below the 80% target for the surface modes, particularly bus routes.
Travel times, by contrast, elicit fairly consistent rankings for all modes. This suggests that the TTC would profitably work on wait time issues as this is a source of displeasure to riders. Planning studies generally penalize wait time because riders feel it is both unproductive and unpredictable, and comparative evaluation of transit proposals can founder on transfer and wait penalties.
Board member Joe Mihevc presented a draft motion that was sponsored by the Canadian Urban Transit Association (CUTA) asking that the city endorse a policy of technology harmonization in vehicle charging stations. This arises from a current situation where manufacturers are developing different, incompatible versions that will make adoption of charging for electric buses more difficult for the transit industry.
Just where electric buses might be in Toronto’s future remains to be seen because the technology is still in its demonstration and proving stage. The TTC will receive a large quantity of diesel buses in the next few years, ironically to replace the less than reliable hybrid buses and some older diesels. This plan is made possible by the Federal infrastructure program, yes, that same government that got us into large scale investment in hybrid buses with targeted subsidies many years ago.
During this discussion, board member and Deputy Mayor Minnan-Wong raised the issue of electric buses for the TTC and cited the May 2016 study Columbia University Electric Bus Analysis For NYC Transit. This analysis provides an overview of the current status of electric bus development, and a calculation of the potential savings both in operating costs and in emissions. There are maintenance savings to be had, but these depend on the technology working well overall. The weakest link in the chain for Toronto’s hybrid buses has been the batteries, and this is still an evolving technology.
Part of the cost comparison in the study was a consideration of the reduced health care costs that would arise from converting New York’s fleet of 5,700 buses to electric operation. These have to be taken with a grain of salt in the Toronto context for a few reasons:
- Health care is substantially more expensive in the USA than in Canada
- The savings in health care, such as they might be, will not accrue to the City of Toronto which would be buying the buses. This would be an argument for transit capital subsidies, but the degree to which this will bring actual commitments at the Ontario or Federal level is unknown.
- The benefit presumed to arise from conversion to electric buses could be masked by the much larger number of diesel engines in trucks within the urban area.
The Board asked staff to report back on the subject, but given current plans and the state of technology, this option is not high on the TTC’s priority list for at least half a decade.
Transit Museum Proposal
Toward the end of the meeting, two overlapping proposals landed in front of the Board by way of a proposal from Board member and Councillor Joe Mihevc for a transit museum. A related request was a scheme to repatriate historic TTC vehicles from other locations to Toronto.
These are really two separate issues, although the TTC’s centenary in 2021 is the hook from which both hang.
The TTC and the City of Toronto are not exactly flush with spare change, and Mihevc recognized that any “ask” of public funding would be a stretch just after getting through the difficult 2017 budget (with worse to come in 2018). The problem then becomes how something like this would be funded. This should not be a small-scale amateur effort, and if it launches on that basis, the project will be doomed.
Chair Josh Colle proposed an amendment to the scheme that would take into account work already underway on the establishment of a Museum of Toronto that will likely be at Old City Hall once the Provincial Courts move out to a new building in a few years. This would still leave the question of how vehicle displays would be handled and whether the central courtyard of the old hall is actually big enough or should be used this way.
The whole matter was packed off to management for a report.