Congestion Management and Transit Priority: Toronto’s Confused Approach

Two studies came through Toronto’s Public Works & Infrastructure Committee at its November meeting revealing a less-than-coherent approach to traffic management and transit priority.

The Congestion Management Plan (CMP) takes a city-wide view with two background reports:

The Downtown Traffic Operations Study (DTOS) has a much tighter focus on the core area from Bathurst to Jarvis with a northern boundary of Queen Street, except between University and Victoria where it extends to Dundas Street.  Background reports include:

Most striking about these reports is the fundamentally different way in which they approach their subject.

DTOS is very much about action, making specific changes “on the ground” to the way streets operate with the goal of improved capacity.  This includes a more sophisticated form of Transit Signal Priority (TSP) that would take into account not just the presence of transit vehicles, but whether they were ready to proceed through intersections.

By contrast, the CMP spends a great deal of time talking about the need for technology upgrades and for co-ordination among various agencies – City Transportation, TTC, Utilities, Emergency Services – to the point one might ask if any of them ever talk to each other today and, by extension, how much “congestion” there is simply in agency-to-agency co-ordination.  TSP gets pushed to the back burner here with a suggestion that it be granted only when vehicles actually need it, although how exactly that would be achieved given TTC’s chaotic approach to line management is anyone’s guess.  More to the point, the TSP facilities now in place were funded by the TTC, but City Transportation now proposes to hobble their usefulness.

Two Approaches to Transit Priority

The Downtown Traffic Operations Study reflects the need for transit vehicles to move through intersections when they are ready to do so, but that green time can be wasted while transit vehicles load passengers.  The emphasis is on giving transit priority when it needs it.

This is introduced in the covering report to the DTOS and detailed in the project description:

Project 15: Transit Signal Priority Strategy

Implement a revised Transit Signal Priority strategy at a number of key locations in the downtown area. The new strategy will consider serving TTC patrons when the traffic signal is red as oppose to serving when the traffic signal is red or green. Implementing this change is expected to continue to provide transit priority, while reducing driver frustration and reducing the possible conflict between vehicles and pedestrians boarding and alighting streetcars.

Project Description:

Implement a revised Transit Signal Priority (TSP) strategy at a number of key locations throughout the DTOS study area given the following scenarios:

  • Revise TSP strategy at locations with existing TSP; and
  • Implement TSP at selected locations without existing TSP.

Initially, efforts should be focused on locations with high boarding and alighting volumes to maximize the effect. The objective is the optimization of people flow and maximization of safety. Therefore, the transit signal priority strategy should focus on managing signal timings and providing priority to transit vehicles with the following guiding principles:

  • Traffic signals should display green when the transit vehicle is ready to proceed through the intersection; and
  • Traffic signals should display red during transit vehicle boarding and alighting.

This approach increases transit passenger safety and minimizes the percent time general traffic is blocked by transit boarding and alighting. Project rollout will occur through the following steps:

  • Step 1 – Determine optimal TSP locations along King Street and Queen Street in coordination with the TTC;
  • Step 2 – Complete preliminary and detailed study design;
  • Step 3 – TSP Implementation; and
  • Step 4 – Complete before/after study to evaluate the TSP success.

The Congestion Management Plan is less aggressive about TSP:

1. Transit Signal Priority – To date, transit signal priority has been implemented in the City at approximately 350 intersections on an unconditional basis (i.e., buses and streetcars receive priority based on presence only). In this project, potential enhancements to traffic signal priority are investigated. Such enhancements could include conditional priority that only provides priority to transit vehicles that would benefit from priority treatment (e.g., behind schedule, experiencing ‘bunching’, etc.). This would reduce the impact of transit signal priority on general traffic operations, and provide opportunities for a wider application of transit signal priority on the City’s arterial street network.  [Page 11]

There is a self-evident bias suggesting that transit gets in the way of other traffic, and that it should receive priority only on an as needed basis.  Although this is superficially similar to the aims of the DTOS, the essential difference is that the CMP defines “need” as bunching or late operation, not the general movement of transit vehicles.

As I have discussed at length elsewhere, the TTC is supposed to be moving to a headway-based line management policy where “on time” has little to do with the schedule, or more accurately, being “on time” should consist of having a reasonable spacing between vehicles.  Just because a vehicle is “on time” at one location does not mean that it won’t benefit from priority especially when downstream traffic or loading conditions are worse.  A car may be “on time” but carrying a gap and be overloaded.  An additional hold at a traffic signal will simply worsen the problem.

It is important that the concept of schedule-based priority not be misapplied.  Where headways are wide and scheduled times actually mean something to riders, there could be an argument for selectively dropping priority.  However, such locations, by definition, see few transit vehicles that would interfere with other traffic, and so the benefit to general road users would be small.

Where service is frequent, it is spacing that counts, not on time performance. On short headways, the extra spacing introduced by “missing the light” can be critical in extending headways and creating bunching.  The last thing we need is a signal system that fights transit vehicles.

Information from the existing Vehicle Monitoring System and reported through NextBus can be, at the scale of traffic signal operations, significantly out of whack.  The polling cycle for the VMS is 20 seconds, and a further delay is introduced by the frequency with which NextBus updates its own database.  The signal system must know that a transit vehicle is at a stop “now”, not where it was 30 seconds ago.  A further wrinkle is that extra services, diversions, etc., must be “known” by the signal system in its effort to determine whether a vehicle is “on time”.

I cannot help thinking that a great deal of effort would be required to implement such a system with limited benefit.  If the intention is to relieve the TTC of the need to actively manage its service, this is the wrong way to proceed.

Downtown Traffic Operations Study

The DTOS recognizes that it is not enough to simply say “there is congestion, make it go away”, but that street operations are affected by a variety of factors, and differing goals could lead to different definitions of “success” in addressing problems.  The study cites eight leading causes of congestion:

  • The Gardiner Expressway, specifically in the backlogs of traffic wanting access to it and backing up onto local streets.
  • Illegal parking, loading, maintenance and construction.
  • Legal activities from the same list, plus building construction.
  • Lack of signal co-ordination and out-of-date timings.
  • Interaction between transit vehicles, passengers and transit signal priority.
  • Loss of capacity to turning vehicles, intersection blockages and vehicle-pedestrian conflicts.
  • Poor lane configurations.
  • Special events of varying lengths including emergencies, festivals, parades, etc.

Each of these contributes to congestion on streets which, in turn, affects transit along with other traffic.  A table of “hot spots” for each of these categories (page 7 of the main report) will not surprise anyone familiar with downtown traffic.

Another important issue is the growth of traffic in the shoulders of the peak periods by thirty to sixty minutes such that the “rush hour” is now three to four hours long.  A goodly chunk of the peak periods attempts to use streets with off-peak regulations for parking and turns.

From the public consultation round earlier this fall, a list of about 50 projects emerged.  This was whittled down to 17 as having the most promise, and they are broadly grouped in four categories.  These are explained in detail in the Project Descriptions.

  • Co-ordination and Communication
    • A team of senior members from various agencies to “develop, co-ordinate and prioritize initiatives” to improve traffic.  Notable by its absence from the agency list is the TTC.
    • Road user education.
    • Personnel to improve intersection operation such as controlling/preventing blockage at intersections.  “Traffic Assistance Personnel” would “guide” traffic and “discourage” blockages, but they would have no authority to enforce bylaws.  That would remain with the Toronto Police Service.  This has “turf war” written all over it, sadly.
  • Traffic Regulation and Management
    • Amend parking and turning regulations to reflect the longer rush hour periods.  The revised hours would be 7:00 to 10:00 am, and 3:00 to 7:00 pm.  The change would apply to Queen and King between Bathurst and Jarvis, and to Adelaide between University and Yonge with possible expansion to other areas after evaluation of the effects.  See the list of new restrictions to be enacted for details.
    • Improve traffic flow with a targeted enforcement plan to reduce curb lane blockage including more towing.
    • Implement new fines, including a rush hour tier, to discourage lane blockage.  Fines would be on a “fixed” basis so that they could not be negotiated down at Court.  Rush hour fines would be set at $150.
    • Implement 13 off-peak courier delivery zones as a pilot project.
    • Improve management of the effects of road and lane closures on the surrounding area.
    • Develop a Special Events plan to maximize road and transit capacity.
  • Traffic Circulation
    • Yonge & Dundas:  Implement specific zones for couriers and deliveries, taxi drop off and pick up.  Enforce peak period parking regulations.
    • Bay Street:  Prohibit or restrict right turns at Richmond, King and Wellington Streets.
    • Two-way streets:  Convert Wellington between Yonge and Peter Streets, and Simcoe between Wellington and Front to two-way operation.  This will reallocate currently excess capacity on Wellington to eastbound traffic and will partly offset the permanent change to the operation of Front Street at Union Station.  (After the subway construction is completed, the road will be narrower than its earlier configuration in recognition of the high level of pedestrian traffic.)  Making an additional block of Simcoe two-way will link the new Lower Simcoe underpass to the reconfigured Wellington Street.
    • Revise westbound lanes on the Gardiner to improve merge operations for traffic entering from local streets.
  • Intelligent Transportation Systems
    • Add traffic cameras at critical sites to allow monitoring of traffic conditions, identify needed adjustments of signal timings, and spot illegal stopping, parking and contractor activity so that enforcement can be dispatched to deal with the problem.
    • Revise Transit Signal Priority at key locations so that signals are red while transit serves passengers at stops, but green when needed to move buses and streetcars.  Implement TSP where it is not now installed.
    • Richmond Street:  Revise signal operations so that traffic queues are detected and signal timings are adjusted to spread out the queue.  Incoming traffic to downtown would be throttled by the signal system when there is only limited capacity for it to be accepted.
    • Entertainment District:  Revise signal operations to reflect the fact that this area does not have traditional periods of traffic behaviour, and that different schemes are needed depending on the day and season to reflect actual conditions.

Quite notable here is that signal priority for transit is only one item in a long list.  In very congested areas such as downtown, “priority” does not mean much if the traffic isn’t moving, but it should allow transit to move when a vehicle is ready to do so.  This is more challenging on streets with very frequent service because there is always a bus or streetcar “there”, but operators need a way to signal “I need a green now”.

Also notable is the fact that the majority of these projects are “action items” that will directly make changes.  Some do involve design phases, and post-implementation reviews are an integral part of all of them.  However, the DTOS has a strong emphasis on actually addressing congestion problems.  An underlying issue, of course, is that there is only so much capacity to go around even in the most optimized situation.

Congestion Management Plan

Like the DTOS, the Congestion Management Plan identifies broad classes of problems contributing to congestion.  There is a lot of overlap in the type of item between the two studies, but the CMP’s scope is city-wide and location-specific recommendations are not part of its mandate.

The CMP’s list is:

  • Bottlenecks.  The city is monitoring various areas where bottlenecks cause congestion.  Although several specific tactics such as signal timing studies, bus lanes and notification of downstream delays (typically on expressways) are mentioned, there is little sense here of a system-wide strategy.  Particularly ironic is the inclusion of cycling lanes as a tactic to reduce congestion.
  • Collisions and breakdowns.  This section concentrates on central monitoring of problems through traffic cameras.  Some of the responses already cited under “Bottlenecks” are repeated here.
  • Parking and stopping.  “… the parking requirements of the City of Toronto’s vibrant business community must be balanced with the roads needs of vehicles, public transit and others …”  Work now underway includes reviewing parking restrictions and increasing enforcement; adding traffic cameras; determining the needs of couriers, business and industry.  This does not exactly sound like a “call to arms” in the same way that the DTOS makes specific proposals.
  • Weather.
  • Road maintenance and building construction.
  • Traffic signal timing.  There are two elements to this category: updating the control system and performing detailed studies of major corridors and intersections.  The CMP proposes that the traffic control system be brought up to date with a new computer system and with a move to wireless communications between controllers and the central system.
  • Special events.

The five year plan 2014-2018 contains many projects in support of the CMP.  What is striking about the list is the number of items that could, at best, be called administrative, not to mention the need to improve collaboration and communication between various agencies.  Like our road and transit systems, the traffic management system also suffers from outdated technology and a major upgrade is an essential part of this plan.

The first of four major activities in the plan is:

Taking a more proactive approach to traffic management on arterial roads, complementing the current traffic management activities on City expressways.  [Page 1]

This statement implies that the arterials today are more-or-less left to their own devices and are not managed on an active basis.  To some extent, this may be a technological limitation of the old systems, but such constraints can easily become part of an organizational culture regarding what can and cannot be achieved.

The Plan has five major goals:

  • Maximize system efficiency and reliability.
  • Improve safety
  • Improve the City’s ability to detect and respond to problems
  • Improve the availability and reliability of information for the public
  • Reduce the environmental effects of transportation

These are to be addressed through a long list of improvements in eight groupings.

  • Intelligent Transportation Systems
    • Replace software
    • Enhanced signal control to be applied to 25% of intersections, up from 18% now, by 2018
    • More cameras for traffic monitoring
    • Monitor arterials from detectors, fleets, 3rd party (e.g. GPS tracking data from smart phone monitor services)
    • Update communications including better emergency networks and increased use of City-owned optical fibre
    • Replace hardware
  • Congestion and Engineering Studies
    • Auxiliary timing plans for the signal network for use in bad weather, shutdowns due to collisions,  etc.
    • Process for regular updates of corridor coordination plans, signal timings, priorities of changes.
    • Active traffic management study.  This is described as a study of hard shoulder running for buses on expressways.  It is unclear what, beyond the provision of some transit priority, this has to do with the topic.
    • Integrated management of corridors: treat transit and other vehicles as network of parallel roads, not as individual entities.
  • Incident and Event Response
    • Traffic incident management teams: co-ordinate better ways to deal with incidents among agencies.
    • Service patrols to provide roving repair services for minor breakdowns.
    • Public education re clearing non-injury accidents.
    • Universal fire station pre-emption.
  • Construction Co-ordination
    • Smart work zones: add cameras and information signs at work zones to monitor traffic conditions and advise motorists.
    • Electronic system for lane occupancy permits: streamline the application process and establish a database allowing monitoring and enforcement agencies to check for abuses.
    • Lane occupancy permit reviews: are the road blockages justified in terms of their traffic effects, and is construction optimized to minimize the length of time road space is needed.
    • Performance management: make contractors responsible for the traffic level of service around a work site.
  • Curbside Management
    • Parking charges: provide more short term spaces, set a higher cost for long term spaces.
    • Study parking requirements, time of day rules, shifting locations of parking.  This parallels the fine-grained parking review in the DTOS.
    • Smart park: investigate a system to easily tell drivers where parking spaces are available such as through smart phone apps.
  • Support All Modes of Transportation
    • Transit signal priority.  Give priority only to vehicles that are “late”.  See the discussion above regarding the validity and applicability of this concept.
    • HOV lane review:  Review the 8 locations where there are bus-only or HOV lanes to determine whether hours of operation or vehicle occupancy rules should be changed.
    • Bike facilities expansion including a separated bike lane network downtown.
    • Corridor renewal for sustainable transportation:  What this phrase really means is the re-engineering of roads to better move a variety of traffic including vehicles, bicycles and pedestrians.  What is not clear is which of these groups has priority in the outcome.
  • Traveller Information
    • Review the technology available to improve information given to travellers.
    • Install signs to display current travel times where there is sufficient demand and space.
    • Establish an event database for co-ordination among agencies of planned events, including lane closures.  The fact that such a database does not already exist says a great deal about problems today.
    • Improve the City website to provide more “modern” and convenient information, not to mention rebranding.  It is unclear how a better brand will reduce traffic congestion, and moreover the study does not acknowledge that a major redesign of the website is already in progress.
    • Review the use of social media as an information tool.
    • Explore the use of mobile applications for information distribution.
  • Traffic Operations Centre
    • Improve the efficiency of the TOC.  What is striking throughout this report is that the TOC today appears to operate in a vacuum.  It runs 12 hours/day (to be extended to 14), but interagency co-operation does not appear to be a strong point.
    • Expanded co-ordination with Emergency Services:  two-way exchange of information with police cameras and information on accidents, and provision of information to police from City databases on events, lane closures and construction.
    • Co-ordination and data exchange with TTC/GO dispatch centres.
    • Co-ordination with external agencies, notably in the 905 region.
    • TOC co-ordination: establish regular meetings of TOC staff and other agencies to debrief after major events and improve plans.

This is a very long plan, and it is uncertain how much of it will be endorsed by Council and included in future budgets.  A great deal of it is at a high level, not at the road-and-intersection level where people experience congestion.  It is as much about “management” as it is about solving congestion.

As I said earlier about the DTOS, the one glaring omission through all of this is a recognition that even with the best design and technology, there is a finite amount of road capacity shared out among many classes of travellers.  At some point, the difficult decision is whether some travellers will have to make do with less so that others can get more.

The answer will be different downtown with its strong transit, pedestrian and cycling culture, and in the suburbs.  A solution for one is not a solution for all, and any plan attempted on that basis would be doomed.  Particularly challenging will be the “in between” areas with the some aspects of the old and new cities, and the effect of change as land use in suburban areas evolves placing more stress on a road system designed decades ago.

What the DTOS shows is the need for micro-level study and design.  This can be tedious both because each street is different, and because a great deal of political and neighbourhood input will be required.  The CMP is a view from on high, and as such may be less successful except on very broadly defined targets.

12 thoughts on “Congestion Management and Transit Priority: Toronto’s Confused Approach

  1. The DTOS seems to be a good plan: it has been researched well, and the plan of action is clearly communicated.

    The other study seems really vague, and seems like another study to sit on a shelf. It seems destined to only get implemented if the solution can be a quick low-effort fix.

    I am curious how the plan to convert Simcoe to a 2-way street aligns with the plan for improved bike lanes on Simcoe as part of the Richmond-Adelaide cycle track study.


  2. Back in 1990, I was fortunate to be the subject of a story on CBC’s Venture about a time tracking tool that I developed. During the story, we were driving around and I pointed out that traffic would flow more smoothly were in not for the two-second “all-red” feature that is a part of Ontario traffic lights. Presumably, the original idea was that this feature was meant to clear intersections and make them safer.

    It is a feature that as far as I know is unique to Ontario. There would probably be howls of protest if it were eliminated, but the simple fact is that the all-red feature adds to congestion because traffic in all directions is stopped for a period of time. And the shorter the traffic light cycle, the greater the percentage of time that vehicles are not moving.

    All-red seems to be a sacred cow. I wish it were on the list for re-consideration.

    Steve: The all-red was implemented because there were too many collisions with people running the end of the amber on one street colliding with people jumping the green on the cross street, especially when one or both were making turns. It is supposed to have reduced accident rates, although two decades on, the only way to prove this would be to switch back and wait for the crash (or absence of one). Motorists are used to this behaviour, and those who presume that they have a few seconds to finish their turns would be in for a big surprise.


  3. You wrote,

    “There is a self-evident bias suggesting that transit gets in the way…”

    This sort of short-sightedness might have a systemic cause: Too many of the contributors to this report might not be residents of this city and therefore are not sufficiently invested in the city to make the right contribution. The fact of their residency is no matter for debate and I only speculate that such a problem might exist among the authors of this report. However, such a fact should not surprise in light of what I know (mostly anecdotally) about the proportional representation of city residents among the Police (13%) and Fire (10%) departments. If this sort of lopsidedness exists among city staff — even if less extremely — then we are all in deep trouble in the long run.

    I hope I’m simply mistaken on all fronts.


  4. I find it bizarre that these kinds of studies pit transit against cars instead of optimizing for overall capacity on a corridor. It defies belief that on Spadina, for example, a streetcar holding 50 people has to wait for five cars to turn left every time it gets stopped at a light. Why won’t the city consider signal priority based on net person-minutes of waiting time and just ignore the type of vehicle altogether? Separating people-in-personal-vehicles congestion from people-in-transit-vehicles congestion is nonsensical.


  5. Would love to know where the 350 intersections in the city of Toronto supposedly implemented with transit signal priority are, whether the TSP is functional, when it operates, and how it modifies the cycle. I find the number a little hard to believe given experience with TSP or lack thereof on the streetcar network.

    Steve: From my riding around town, there are several intersections where TSP used to operate, but does no longer. This is supposed to be fixed by a change from loop detectors in the pavement to pole mounted sensors that will recognize streetcars. I am still waiting to be convinced of this technology, but if they can actually keep it working this will be an improvement.

    There are several different approaches to TSP depending on the location. Notable by their absence are the intersections on Spadina where TSP exists only for streetcars making turns with the assistance of electric switches (the switch controller signals the need for a turn phase). If the switch is operating manually, then there is no TSP. There is of course no TSP, not even green time extension, to ensure that streetcars can slip through an intersection at the end of a cycle.


  6. On transit priority I have seen many wrong moves. NB Bathurst has priority. I take EB College. The light stays red for College if [there is a] NB Bathurst car. Countdown timer for pedestrians is at 0 for 15 or more seconds before Bathurst goes to yellow or red.

    If as soon as NB streetcar got there, Bathurst was red it would allow loading time and College to move.

    The other is St Clair. Lansdowne short turn only has priority at the beginning of the cycle. Gunns Rd has at both ends. Transit white before the green and also after.

    Steve: One example you give shows how a move to recognizing when a streetcar actually needs priority would be useful. A car standing at a stop loading does not require priority, but it might benefit from extra green time if it is ready to leave soon enough.


  7. Just wondering how flexible next bus is…it can’t be that hard to update the system on a 5 or 10 second basis compared to 30….likewise including user counts from any streetcars with the footpad monitoring, or door monitoring (on the new vehicles I assume this could be added in fairly easy to the complete order).

    Steve: The problem is not NextBus, but the TTC’s antique vehicle monitoring system on which the polling cycle is 20 seconds. Add in whatever update cycle the TTC uses for its own database, and the polling cycle of NextBus to extract this data from the TTC, and it’s easy to get to 30 seconds or more. I have seen Nextbus a minute or more behind the real world when technical problems behind the scene slow down the processing or communication between systems. The TTC finally has money in its capital budget for a new vehicle monitoring system, very, very long overdue. Now if only they will buy something off the shelf rather than falling into the “not invented here” trap of so many Ontario boondoggles.

    It would be good if the city took a more active approach to dealing with delivery trucks … something like an automated system where they say the streets that they will be delivering to during the day (or in the next 5 minutes), and the city provides them with specific locations – perhaps by reserving on street metered parking spots for them for them to park at.

    Single user congestion charges on the QEW during rush-hour would go a long way to improving downtown traffic as well … likely cutting it by a quarter to a half … just target anyone coming off one of the 4 downtown ramps by themselves … 3$ per vehicle.


  8. Until I see transit vehicles (and emergency vehicles) actually moving BEFORE any other non-transit vehicle, it will not be transit priority in my books.

    Instead of the two-second all red signals, why don’t they give transit the priority signal they deserve instead. After the two seconds, then give the left turners their turn. Usually, I have seen, it takes a while (inertia dampers damaged, Captain?) before vehicles move once they get their signal to actually move.

    If the left turning vehicle has to wait as the transit vehicle makes their crossings, at least they’ll actually be half way into the intersection before actually doing the turn anyways (which is the way I was taught).


  9. Advance greens like all major cities in North America. So painfully obvious and a reason why driving in the US is so much easier.

    Steve: Advanced greens are used in some locations, but their primary benefit comes where there is an unbalanced demand and the opposing direction can operate with less green time. Also, the presence of an advanced green can induce left turn demand at an intersection by making them easier. This is bad for streets that are not wide enough for a left turn bay because it blocks through traffic by adding to the turn demand.


Comments are closed.