The TTC Board met on March 26, and considered a meaty agenda that begins to address some important policy issues.
Updated March 29, 2015 at 3:45 pm: The presentation on One Person Train Operation (OPTO) given at the meeting has been added along with comments.
Updated March 24, 2015 at 8:10 am: After this was published, the TTC posted the CEO’s Report.
In a previous article, I wrote about the Spadina subway extension project update. This will undoubtedly be the main attraction both for board members and the media. Other items of interest include:
- An overhaul of system key performance indicators (KPIs)
- A door monitoring system for Toronto Rocket trains and one person train crews (Updated March 29)
- Revision and consolidation of the resignalling contract for the Yonge-University line
- A study of express bus routes
- CEO’s Report
Over many years, TTC management has reported on various aspects of the transit system through tracking indices, charts and tables, and there has always been a desire at the board level to make this information as simple as possible. Two basic management problems arise from this practice:
- Key information that might be evident at a more detailed level is lost in averaging of observations over a broad reach, or by the choice of the wrong measurement factor. The sun rises every day, but this tells us little about the weather.
- Targets for “good performance” tend to reflect current practice with little sense of why a metric may sit at a specific value or how it can be improved. This is further complicated by averaging effects where significant changes at a detailed level are lost in the summary data.
A further problem is that metrics used by the TTC might, or might not, be comparable to data from other organizations in the industry, and external benchmarking is difficult. A recent industry review of subway operations, for example, ranked TTC highly for “efficiency”, but this was a direct result of lower than average maintenance cost. There was no examination of the efficacy of the level of maintenance or the quality of service resulting from TTC practices, nor any comparison system-by-system of accounting practices that could skew the numbers being reported. Surface networks and their role as part of a total transit network including the supply of passengers were completely ignored. The TTC routinely cites this report as showing how well they are doing, but omits any discussion of the context for their rank and performance relative to other systems.
Long absent from published TTC reports is any indication of vehicle reliability, a fundamental point of comparison with peer systems and an important issue internally. How do TTC fleets stack up against those in other cities whose equipment has a similar duty cycle? Do successive generations of equipment show an improvement in reliability not just against older vehicles “as is” but to their historic performance figures when they were of comparable age? What constitutes a good maintenance plan and spare equipment ratio, and how will changes in these translate to better service? Can any improvements be tracked and demonstrated?
Readers here will know of the TTC’s quarterly performance reports on a route-by-route basis most recently reported for the fourth quarter of 2014. I will not reiterate my critique of these measurements here beyond saying that they represent a complete abdication by the TTC to provide anything remotely close to reliable service, or at least to report on service quality in a meaningful way.
The detailed presentation of new metrics is not yet online. I will update this article after the board meeting.
TTC management proposes that the board approve the next step in a move to One Person Train Operation (OPTO: a new acronym for readers to learn) with a trial on the Sheppard Subway (aka Line 4) of a door monitoring system. Here is a description of the proposed system:
The TDM System consists of four strategically placed CCTV cameras that are installed on the subway platform to provide live clear video of all 24 train doors while the train is in the station. The video images are collected and transmitted by means of a wireless system to the subway train. Similar equipment is installed on each train to collect, process and feed the video images to a monitor in the operator’s cab.
The monitor is installed in the operator’s cab in such a manner that it does not obstruct the operator’s view of track level and the signal system while considering ergonomics for the operator. The view on the monitor is split in four providing a live view from each CCTV camera on the subway platform.
The TDM System provides the operator with the ability to view the subway car doors from a forward facing position in the cab while the train is in the station and as the train is leaving the station. The operator will have the ability to discern objects or people caught in the doors. The system automatically turns on and off as the train enters the station and does not require any action by the operator to operate. [pg. 4]
The specific request in this report is a change order to Bombardier under its current TR car supply contract to design the modifications required in TR cabs for the video displays and for revised door operation controls.
The Board authorize a Change Directive to Bombardier Inc. (Bombardier), in the amount of $2,734,822.98, including taxes, for the additional engineering design to facilitate installation of TDM equipment and modification/relocation of door control systems on the TR trains for OPTO. [pg. 2]
Note that this does not include the procurement of the station camera systems, nor the installation/modification of equipment on the trains themself. A related problem is that the TR trains do not now operate on the Sheppard line. This brings us to the following proposal:
Once Bombardier completes the engineering design a proposal will be requested from Bombardier for installation on one prototype TR train for testing. Line 4 Sheppard Line was selected as the pilot for the OPTO concept because there are only 4 trains operating during customer service hours.
The T1 trains presently operating on Line 4 will require replacement with TR trains that are Automatic Train Control (ATC) equipped prior to implementation of ATC on Line 1 YUS in 2020 because access to Line 4 is from Line 1. The required conversion of a six car TR train to a four car train for the OPTO pilot will be the subject of a future Board report. [pg. 5]
This has all the earmarks of a project that will proceed in bite-sized steps with one funding request after another, but no sense of an overall program. The scheme to shift TR trains onto Sheppard will take equipment that was originally purchased for service improvements on the YUS and redirect it to the Sheppard line. The current fleet plan calls for 10 additional trains (which have already been ordered from Bombardier) that would enter service from 2019 to 2031.
The same plan shows T1 equipment remaining on Sheppard for the indefinite future. The “need” to convert the line to TRs is a very recent change.
One might reasonably ask why the TTC orders trains so far in advance of actual need, especially when there will be a separate order for new trains for the BD line in the mid-2020s. The philosophy, of course, is “get ’em while they’re hot”, in other words, order more trains at a lower price as part of a big order. There are, however, limits to this and the related costs of housing trains we will not actually need for well over a decade. The “savings” may not be quite what they seem, and in the process the TTC acquires a bloated fleet.
In the medium term, the proposed trial implementation looks like this:
Upon completion of the design for the TDM and door control modifications, approval will be obtained to modify one TR train as a prototype. Upon completion of the modifications, testing of the TDM equipment and door controls will take place followed by evaluation processes as prescribed by the Concept of Operations. It is planned to have one TR train available for service in OPTO by the end of 2015.
The associated cost savings and increases related to OPTO will be available at that time for development of a business case for roll out of OPTO for the remaining 3 trains on Line 4 in 2016 and for submission of the 2016 – 2025 Capital Program for roll out of OPTO on Lines 1 and 2. [pg. 5]
This plan becomes even stranger with the following reference to funding:
Sufficient funds are available for the pilot project on Line 4. Modifications to the remaining TR and all T1 trains for Lines 1 and 2 are unfunded. Sufficient funds will be submitted in the 2016 – 2025 Capital Program upon successful completion of the prototype TR train on Line 4. [pg. 2]
The TTC’s fleet and signaling plans call for the BD line to receive new equipment and convert to ATC in the early 2020s, certainly in time for opening of the Scarborough Subway Extension. It is quite strange that TTC would contemplate a retrofit of the T1 equipment for OPTO on the BD line when these cars would probably be less than five years from their retirement. Moreover, design work for the TR trains would have to be redone for the T1s because these have completely different cabs.
There may well be a business case for implementation of OPTO on the TTC system, but this report underplays the technical and financial issues the TTC could face in the actual implementation. It is striking that:
The TR trains do not have provisions for installation of the TDM System or relocation/modification of the door control system. A significant amount of structural work, electrical work and software changes are required in the TR train. [pg. 6]
One could argue that preliminary work is needed just to reach a decision point on proceeding with this conversion now, or as part of a future project involving new trains and station equipment. However, the history of TTC projects is that this could be “in for a penny, in for a pound” with the goal of OPTO overwhelming any technical issues of implementation.
Why, for example, does the TTC not begin with the premise of OPTO and ask “how much will this actually save”? That translates to a sustainable level of spending on technology change from operational savings and at least an order of magnitude figure for capital spending. Of course, if the savings from OPTO are consumed by capital costs, they are not available as offsets in the operating budget.
At the very least, the TTC should examine various scenarios for OPTO implementation including staging conversions to coincide with provision of new trains whose controls are designed for this option from the outset.
Meanwhile, operators on Sheppard should be prepared to walk the length of a four-car train over 40 times per day (an 8 hour shift with the need to change ends every 11 minutes, about 3.7km). They will be very fit.
At the meeting, Chief Operating Officer Mike Palmer gave a presentation on OPTO (One Person Train Operation).
OPTO Presentation [skip to page 8 of linked pdf]
One person operation has been in use worldwide for decades, and is appearing city-by-city, line-by-line even on very old systems such as Boston’s. The SRT, albeit with comparatively small trains by TTC rapid transit standards, has operated with one person crews since it opened 30 years ago.
The position taken by the TTC is that the single train operator who will, eventually, not be driving the train, can concentrate on door operations, a move that can improve safety. The presentation cites stats from London where OPTO lines had fewer “door related incidents” than non-OPTO ones, but it is unclear what other factors could have been at play.
One might argue that “safety” could equally be improved by an attitude adjustment of guards who now delight in closing doors on boarding/departing passengers. This is not a case of a last-minute attempt to jam through closing doors, but of failure to pay attention to (or simply ignore) passengers.
Inevitably a move to OPTO brings with it labour unrest. The TTC hopes to avoid this by treating automation and one person control as a chance to redeploy staff rather than to downsize. Whether this is practical will depend a great deal on the duties of the new positions which, if they involve standing around in stations for eight hours at a time, will be none too attractive.
The claim that “future service improvements require no new hires” shows that this is a transitional arrangement, not a permanent one, because any added service will require at least a driver. The cost will be lower than with a driver a guard, but it won’t be zero unless the TTC plans to absorb staff back into driving roles from temporary positions as station monitors. Conversely, the TTC could simply absorb drivers back onto the surface workforce. The TTC needs to clarify its position on train and station staffing because their stated plans are inconsistent.
Among many issues listed for the shift to OPTO is one oddball: tunnel ventilation, listed with the need to “upgrade and repair”. If there are tunnel issues, these should be addressed as basic safety and maintenance matters, and they do not have anything to do with how big the train or station crews might be. Conversely, if OPTO implies a need for better ventilation, then this is a cost and risk that must be identified up front.
As I wrote in the original article, the changes needed to retrofit the older T-1 trains used on Sheppard and Bloor-Danforth (a few sets remain on YUS for storm service because no TR trains have been fitted with de-icing equipment yet) are more complex. Given that the BD line won’t move to a new signal system and ATC likely until the early 2020s, it could make more sense to aim at the next planned fleet upgrade and the Scarborough Subway Extension as targets rather than attempting to retrofit T-1s for a limited remaining lifespan.
The presentation takes a “wait and see” position on conversion of the BD line including the consolidation of conversion with acquisition of a new fleet.
[See also The Evolution of TTC Signaling Contracts]
For some time, the TTC has been engaged in replacing and upgrading the signal system on the Yonge-University subway line. The original section, from Eglinton to Union, opened in 1954, and equipment installed at that time is well beyond its design life. With technology changes, parts are difficult to obtain. The need for a complete replacement is without question.
The TTC launched this process in September 2008 with a contract covering Eglinton to St. Patrick Stations (this covers the 1954 line, plus the southern end of the 1963 University line that overlaps into the control territory of Union Station). The technology to be used was Computer Based Interlocking (CBI), a modern version of the original signalling that is controlled through track circuits to monitor train locations.
Less than a year later, in April 2009, the TTC decided to embark on the implementation of Automatic Train Control (ATC) over the entire line, then Finch to Downsview Station. Although the Spadina Subway project (TYSSE) was already underway, signalling for that extension was not included in this contract. Alstom was the successful bidder for the ATC contract which included on board control equipment for the 39 TR trains then on order.
At this point, three critical assumptions had been made:
- That some of the service on YUS would continue to be provided by T1 trains to which ATC would be retrofitted.
- Work cars and yard areas would not be covered by ATC, but would use conventional signalling to navigate the system.
- That the CBI and ATC technologies from different vendors would be able to co-exist.
The plan for some T1 trains to remain on the YUS ran into problems on a few counts:
- The cost of retrofitting this equipment for ATC proved to be quite high, and
- The TTC wanted to push ahead with replacement of all of the “H” series trains due to reliability problems, and this required shifting many T1 sets to the BD line.
Additional contracts extended the scope of both the equipment and signals procurement:
- June 2011: An additional 21 sets of ATC equipment were ordered for 21 supplementary TR trains (bringing the total to 60 sets) to replace the “H” cars.
- March 2012: A contract with Ansaldo for implementation of CBI on the remainder of the YUS plus the TYSSE.
- January 2013: An additional 10 sets of ATC equipment were ordered for the 10 TR trains destined for the TYSSE.
- April 2014: A major restructuring of the Alstom contract was approved to simplify the phased implementation, to extend ATC to the TYSSE, and to provide equipment for a further 10 sets of TR trains intended for future growth in demand. (See the discussion of TR trains for the Sheppard Subway above.)
The actual implementation to date of the two technologies has not gone well, and the projects are behind schedule. In 2014, the TTC retained Parsons (not to be confused with Parsons Brinkerhoff who worked on the TYSSE project management review) to review the project and recommend improvements. This led to the current proposal to consolidate all work under Alstom and cancel the outstanding contract with Ansaldo.
The confusion about signalling technologies is rather strange, and speaks to inconsistencies in the TTC’s scoping of this project:
The current signaling contract arrangement for Line 1 has evolved since its inception in 2008 with a higher than anticipated passenger demand and an increased scope with the inclusion of ATC on TYSSE … [pg. 1]
During the very period when the TTC was telling anyone who would listen that Toronto didn’t need a Downtown Relief Line and that passengers would somehow fit on an upgraded YUS, the need for a better signal system was somehow forgotten, or it was conveniently ignored as a pre-requisite to increase YUS capacity.
An important technical change that appears only in the recommendations is:
Increase $74,580,000.00 for adding the Alstom CBI system, equipping work cars for ATC and interfacing to the Wilson yard signaling system. [pg. 2]
This sorts out a previous design issue that non-ATC rolling stock, notably the work fleet which does operate during revenue hours, would not be controlled by the ATC system and, therefore, would require management by the CBI system and its interface to the primary ATC system. (The same would be true if a T1 train without ATC ventured onto the YUS.)
Obviously there will be sunk costs for the Ansaldo work that cannot be recovered plus whatever cancellation penalties might apply. The TTC expects to remain within the project budget and timeframe because of significant savings:
The key areas of savings that offset the cost of this contract change are:
- Reduced TTC construction costs, both material and labor as significantly less field equipment is required.
- Greatly reduced number of subway closures.
- Costs recovered from the cancellation of the existing CBI contracts.
- Reduced effort in TTC design with one supplier not three.
- Reduced testing and commissioning activities given the simplified solution as the need for independent subsystem testing is eliminated.
- The ability to test the new system during the day without inconveniencing the public, i.e. running the new system in shadow mode and ensuring greater reliability from day one.
The risks associated with current complex contract and technical arrangements are greatly reduced also allowing more confidence in cost and schedule.
The financial impact on future years is significantly reduced as the maintenance costs of the newly proposed solution are also greatly reduced. [pp 9-10]
A detailed technical review of this project is not online, but according to the report is available. I will be asking to see it if only to better inform myself on issues related to signaling changes on the subway.
In line with the difficulties already reported on the TYSSE project, it would certainly be useful to see an updated project implementation schedule and a confirmation of the estimated cost/savings tradeoffs. Needless to say, this should be tracked to completion.
TTC management proposes a study of express routes that will consolidate two requests from the Board. The first in March 2014 asked for a feasibility report on additional express bus routes. The second was the August 2014 “Opportunities” report that proposed various improvements including increased express services on existing routes.
No new services can be added given the current limitations on the fleet until at best early 2016, and this study will report in October 2015. Whether any recommdations from the study find their way into the 2016 budget will depend on the mood of the Board and of City Council who will still be digesting unexpected costs on capital projects.
A Service Plan for express routes will address:
- the costs and revenues associated with existing express routes;
- an analysis of all existing express services to determine the viability of these services;
- an analysis of instituting peak-period express service on the TTC’s busiest bus routes which do not already have express service;
- an analysis of possible new “rocket” express routes that would directly link major generators;
- the potential benefits of using articulated buses on existing and/or proposed express routes;
- a cost / benefit analysis of different fare structures for express bus route services;
- potential means of alleviating bunching of buses and short-turns on routes being considered for express bus service;
- the implementation of queue-jump lanes, priority signalling, and dedicated lanes as ways to improve speed and reliability on existing and proposed new express routes; and
- a review of other comparable municipalities or transit systems that successfully operate express bus services [p. 4]
A complete review of these routes is long overdue. Many existing “Downtown Express” services exist because of special pleading from Councillors who were on the TTC Board when they were implemented. These routes generally do not appear in the TTC’s annual statistics for its surface system, and there has never been a report comparing the cost of resources devoted to these routes with the benefit they might confer. Could the buses be better used elsewhere? Is there justification for running more service on these or other new express routes? Are there suburban nodes that could support express routes?
Part of the study will look at existing “local” routes to determine whether they could benefit from an express overlay. This is always a tradeoff problem because some riders use local stops at one end or the other of their trip on a route, and the “express” branch is of no use to them. Generally speaking, creation of an “E” branch speeds travel for those who can use it, but hurts those who ride local branches because fewer buses are left to serve them.
This will be an interesting review in particular because it is system wide, rather than a location specific response to “squeaky wheels”.
Earlier in this article, I wrote about the poor quality of the Key Performance Indicators (KPIs) used by the TTC to track aspects of the system, notably service quality. While new methodology may be in the works, this has yet to show up in the CEO’s report.
Service quality is still reported relative to a ±3 minute target of scheduled headway. The subway achieves a very high rating, but it is almost impossible for a route with all day frequent service to achieve a low rating. During the peak period, half of the service could be missing, but the headway would still be within 3 minutes of the scheduled value. On surface routes, the target is a hapless 65% for bus routes and 70% for streetcars. Many routes fail to meet even these targets on paper, and riders can be excused for thinking even these numbers are optimistic especially beyond common short turn points on major routes.
Streetcar service has particular problems, although oddly enough weather issues are not mentioned.
Construction associated with Harbourfront Toronto continued to negatively impact both the performance of the 509 Harbourfront and 510 Spadina routes and overall street car performance.
The March Board meeting will include a presentation from the Chief Service Officer on actions being taken to arrest this decline and drive up all aspects of performance.
Leaving aside that it’s “Waterfront Toronto”, by January the schedules for the routes on Queens Quay included extra time to deal with traffic signals that interfered with the service rather than helping it. Queens Quay represents a relatively small part of the streetcar system, and problems there should not have a large effect on system-wide stats. A major problem through the winter was vehicle reliability, but the TTC does not report on this aspect of its operation.
Elevator and escalator “availability” is reported above targets of 98% and 97% respectively. Although the report speaks of the benefits of improved maintenance on reliability, there is little movement especially in the escalator numbers. This could reflect what is counted as an “unavailable” device.
Ridership in January 2015 is up about 1.9% over 2014, and on a rolling annual basis is up by 1.8%. However, this is lower than the budget target and the drop is blamed on January’s unusually bad winter weather.
The new streetcar roll out plan is described in conflicting ways in this report:
After much discussion and negotiations, which included me [Andy Byford] and my counterpart at Bombardier, we have now received a revised delivery schedule from Bombardier for our new streetcars. We have four state-of-the-art streetcars in operation on the 510 Spadina route; however, a much higher number should have been received by now, but production difficulties with Bombardier have caused significant delays.
The new schedule commits to 30 cars being delivered by the end of 2015, enough to complete the conversion of the 510 Spadina route, plus the 509 Harbourfront and 511 Bathurst routes. This will require Bombardier to reduce production time to five from 10 days per vehicle. [pg. 5]
Until streetcars begin arriving on a regular, frequent basis, we will not know whether Bombardier has managed to overcome its supply chain and quality control problems.
Meanwhile at Leslie Barns, the TTC expects to have “staged occupancy” in June 2015 with project completion by the fourth quarter of 2015. Construction continues on Leslie Street but has finally progressed to the point where the track bed north to Queen and south from Lake Shore is under construction.
The implementation of the Presto fare card continues through the system, but as yet there is no discussion of a new fare structure such as time-based fares or cross-border integration with systems in the 905.
A common problem through the CEO’s Report is that projects are described in text, some of which changes little from one report to another. Projects schedules are not shown, nor are there exception reports to flag significant changes from previous versions.