Table of Contents
- Summary
- Introduction
- Definitions and terms
- Findings
- Recommendations
- How does the Toronto Transit Commission measure and report on service?
- What TTC operators say about bunching
- What TTC operations supervisors say about bunching
-
Technical Backgrounder: How the experienced wait time is calculated
- Appendix: How do other transit agencies measure and report on service?
- References
1. Summary
Taking public transit should not be a gamble: it should be a given that your bus or streetcar will show up when you expect it to. The public expects that information about transit reliability is being monitored, reported to the TTC Board, and corrected if needed.
Yet the TTC’s current “On-Time Performance” metrics, which get shared in monthly CEO reports to the TTC Board, hide more than they reveal about the state of service and reliability that transit riders experience. The TTC can't fix what it does not measure.
This report uses TTC on-time performance statistics obtained via Freedom of Information request as well as bunching calculations gathered by TransSee from TTC real-time vehicle location data to demonstrate the actual condition of TTC reliability and how the TTC could report on service in more accurate and transparent ways. TransSee is a real-time bus tracking and next vehicle predictions website: www.transsee.ca.
Report Findings
Between September 1 to November 16, 2024:
- Riders waited 50% longer than scheduled on 10 TTC routes due to bunching.
- Riders waited 30% longer than scheduled on 41 routes due to bunching.
- Only 10 routes during the evening rush hour met the TTC’s goal of being on time 90% of the time.
- Bunching is not limited to the routes selected by the TTC for a bunching reduction pilot program.
- Managing service will help maximize investments in more frequent service and transit priority measures.
- The way TTC reports on time performance as an average, which flattens a month’s worth of data in all time periods and across all routes, hides how poorly some routes perform.
The TTC’s 2025 budget makes welcome investments in TTC service and introduces "pilot programs to reduce bunching and gapping" on the most unreliable routes.
TTCriders makes timely proposals in this report for the TTC to make the most of the pilot initiative by establishing new, transparent ways to report on service.
Recommendations
- Change how service and on-time-performance metrics are measured and create a public dashboard to report on these metrics
- Report regularly on headway performance, so the public can evaluate the success of the TTC’s bunching pilot
- Hire additional operations supervisors to implement a bunching reduction pilot
- Change how service is managed with adequate resources
- Communicate honestly with riders about service management
- Report on the accuracy of real-time predictions shared with apps
2. Introduction
“We can’t improve service reliability without first accurately measuring it.”
- Massachusetts Bay Transportation Authority
Have you ever waited for your bus or streetcar to show up, only to have two or three of them arrive at the same time -- much later than you expected?
This is called “bunching” and occurs when delays from traffic congestion or long loading times at bus stops accumulate.
When a bus gets delayed in traffic and buses are not being managed to keep appropriate gaps between vehicles, service reliability can spiral. As a bus falls behind schedule, delays can grow because it has to pick up more riders along the way who have been waiting longer than scheduled. Eventually, other buses may catch up to the first bus that became delayed and show up at a bus stop simultaneously or within a short time interval, which is known as bunching. The flip side of bunching is gapping, because two buses arriving simultaneously or within a narrow window of time are preceded or followed by a longer gap in service than what is scheduled. When vehicles bunch up, the first vehicle is typically overcrowded while the vehicles behind it have fewer people on board.
“Bunching” is not something that the TTC publicly reports on, even though it has the data to do so and its Official Service Standards have targets for “headway performance.”
The 2025 TTC budget sets out to establish "pilot programs to reduce bunching and gapping on 10 of the most problematic routes with enhanced on-street presence.” A staff presentation later listed 11 priority routes that will be targeted through the pilot:
- 7 Bathurst
- 24 Victoria Park
- 25 Don Mills
- 29 Dufferin
- 100 Flemingdon Park
- 165 Weston
- 506 Carlton
- 512 St. Clair
- 924 Victoria Park Express
- 925 Don Mills Express
- 929 Dufferin Express
This is a welcome initiative that will help maximize investments in more frequent service and transit priority measures.
This report explains why tackling bunching is so important. Using a combination of real-time vehicle location data collected by TransSee and TTC on-time performance statistics obtained via Freedom of Information request, this report demonstrates the actual condition of TTC reliability and how the TTC could report on service in more accurate and transparent ways. TransSee is a real-time bus tracking and next vehicle predictions website.
This report was created by members of TTCriders, based on analysis and input from Darwin O’Connor (transsee.ca) and Steve Munro (stevemunro.ca). Transit operators and operations supervisors were consulted to learn from their experience.
3. Definitions and terms
Headway
The scheduled time between vehicles.
Bunching
Bunching occurs when a bus falls behind schedule enough that the bus behind it catches up. Conversely, bunching can also occur if a bus is running well ahead of schedule and catches up with a bus in front of it. Instead of arriving at evenly-spaced intervals (e.g. 1 bus arrives every 5 minutes), buses arrive in clusters with long waits in between (e.g. 2 buses arrive together every 10 minutes).
Gapping
Gapping describes the long wait times that occur between bunched clusters of buses. Instead of arriving at evenly-spaced intervals (e.g. 1 bus arrives every 5 minutes), buses arrive in clusters with long waits in between (e.g. 2 buses arrive together every 10 minutes).
Missed trip
The TTC defines any vehicle leaving more than 20 minutes late from an end as a “missed trip.”
Short Turns
When a vehicle is turned back and taken out of service before reaching the end point of a route.
GTFS = General Transit Feed Specification. An industry standard for publishing transit schedules.
APC = Automatic Passenger Counter
API = Application Programming Interface

4. Findings
Click here to view a spreadsheet of real-time vehicle location data collected and analyzed by TransSee from October 1 - November 16, 2024.
Bunching Data||https://docs.google.com/spreadsheets/d/1Y5glvZBukgYxnZhoPpz0x3AioD8clV3GGqQFCJnWAqg/edit?gid=89825827#gid=89825827
Click here to view a spreadsheet of TTC on-time performance statistics for October 1 - November 16, 2024 obtained via Freedom of Information request.
OTP Data Googlesheet||https://docs.google.com/spreadsheets/d/e/2PACX-1vQwiKQY05Kzrqqkei1HEed0HLheiDfgseiNkp1RM6He0eXabxr0Wgie42y0-KUbfAzgxpe3JCNH4lHY/pubhtml
Click to view charts created by Steve Munro sorted by night service, streetcars, and buses in the NorthEast, NorthWest, West, East, and South, based on the above TTC on-time performance statistics obtained via Freedom of Information request.
OTP-Night||https://assets.nationbuilder.com/ttcridersto/pages/4545/attachments/original/1737523505/OTP_ByRouteandPeriod_Night.pdf?1737523505
OTP-SC||https://assets.nationbuilder.com/ttcridersto/pages/4545/attachments/original/1737523504/OTP_ByRouteandPeriod_SC.pdf?1737523504
OTP-NE||https://assets.nationbuilder.com/ttcridersto/pages/4545/attachments/original/1737523502/OTP_ByRouteandPeriod_NorthEast.pdf?1737523502
OTP-NW||https://assets.nationbuilder.com/ttcridersto/pages/4545/attachments/original/1737523501/OTP_ByRouteandPeriod_NorthWest.pdf?1737523501
OTP-West||https://assets.nationbuilder.com/ttcridersto/pages/4545/attachments/original/1737523499/OTP_ByRouteandPeriod_West.pdf?1737523499
OTP-East||https://assets.nationbuilder.com/ttcridersto/pages/4545/attachments/original/1737555694/OTP_ByRouteandPeriod_East.pdf?1737555694
OTP-South||https://assets.nationbuilder.com/ttcridersto/pages/4545/attachments/original/1737523504/OTP_ByRouteandPeriod_South.pdf?1737523504
Bunching findings
Bunching is not limited to the selected by the TTC for a bunching reduction pilot program. Managing service will help maximize investments in more frequent service and transit priority measures. City data from the King Street Transitway and RapidTO on Eglinton East, Kingston, and Morningside shows that transit priority measures support reliability.
Between September 1 to November 16, 2024:
Riders waited 50% longer than scheduled on 10 TTC routes due to bunching (routes in bold will be part of the TTC’s bunching reduction pilot program):
- 320 Yonge Night Bus
- 509 Harbourfront
- 945 Kipling Express
- 300 Bloor-Danforth Night Bus
- 504 King
- 100 Flemingdon Park
- 54 Lawrence East
- 510 Spadina
- 35 Jane
- 32 Eglinton West
Riders waited 30% longer than scheduled on 41 routes due to bunching (routes in bold will be part of the TTC’s bunching reduction pilot program):
- 320 Yonge Night Bus
- 509 Harbourfront
- 945 Kipling Express
- 300 Bloor-Danforth Night Bus
- 504 King
- 100 Flemingdon Park
- 54 Lawrence East
- 510 Spadina
- 35 Jane
- 32 Eglinton West
- 25 Don Mills
- 29 Dufferin
- 506 Carlton
- 952 Lawrence West Express
- 505 Dundas
- 52-Lawrence West
- 941 Keele Express
- 53-Steeles East
- 114 Queens Quay East
- 36-Finch West
- 202 Cherry Beach
- 929 Dufferin Express
- 996 Wilson Express
- 301 Queen
- 960 Steeles West Express
- 306 Carlton
- 24 Victoria Park
- 925 Don Mills Express
- 7 Bathurst
- 165 Weston Rd. North
- 501 Queen
- 63 Ossington
- 503 Kingston Rd
- 43 Kennedy
- 304 King
- 96 Wilson
- 95 York Mills
- 116 Morningside
- 508 Lake Shore
- 41 Keele
- 89 Weston
Two routes that the TTC plans to focus on in a bunching pilot do not appear in the top 41 worst routes between September 1 to November 16, 2024:
- 512 St. Clair
- 924 Victoria Park Express
On-time performance findings
The way TTC reports on time performance as an average, which flattens a month’s worth of data in all time periods and across all routes, hides how poorly some routes perform. The most recent monthly TTC KPIs report reported that for the October 6 to November 16, 2024 period, bus on-time performance was 83%, while streetcar on-time performance was 69%. During the September 1 to October 5, 2024 period, the TTC KPIs report reported that bus on-time performance was 82%, while streetcar on-time performance was 72%.
 |
 |
Bus OTP from TTC KPI report (Sept 1 - Oct 5, 2024). |
Streetcar OTP from TTC KPI report (Sept 1 - Oct 5, 2024). |
 |
 |
Bus OTP from TTC KPI report (Oct 6 - Nov 16, 2024). |
Streetcar OTP from TTC KPI report (Oct 6 - Nov 16, 2024). |
In contrast, according to the TTC’s On-Time-Performance data obtained via Freedom of Information request, which breaks out routes and time periods, between September 1 to November 16, 2024:
Only 10 routes during the evening rush hour met the TTC’s goal of being on time 90% of the time:
- 905 Eglinton East
-
127 Davenport924 Victoria Park Express
- 104 Faywood
- 17 Birchmount
- 903 Kennedy Stn-Scarborough Express
- 48 Rathburn
- 507 Long Branch
- 126 Christie
- 131 Nugget
17 TTC routes were on time less than 50% of the time during the same time period during the evening rush hour:
- 185 Sheppard Central
- 171 Mount Dennis
- 50 Burnhamthorpe
- 61 Avenue Road North
- 202 Cherry Beach
- 100 Flemingdon Park
- 91 Woodbine
- 925 Don Mills Express
- 508 Lake Shore
- 941 Keele Express
- 952 Lawrence West Express
- 79 Scarlett Rd.
- 154 Curran Hall
- 43 Kennedy
- 505 Dundas
- 15 Evans
- 118 Thistle Down
5. How does the Toronto Transit Commission measure and report on service?
“On-Time Performance”
Every month, the TTC’s CEO reports on the “On-Time Performance” of buses, streetcars, and subways. The TTC’s stated goal is to have 90% of subway, bus, and streetcar trips meet its on-time performance standard.
But the metrics used by the TTC to track “On-Time Performance” do not paint an accurate picture of the service that transit riders are actually experiencing.
For a Wheel-Trans bus to be considered on time, the bus must arrive within 20 minutes of its scheduled arrival.
For a TTC subway to be considered on time, the train must be “within 1.5 times of its scheduled headway” at the end terminals. “Headway” means the amount of time between train arrivals at a station. This means that if trains are scheduled to arrive every 5 minutes, trains could arrive 7.5 minutes apart and still count as being “on time.”
On-time performance for TTC buses and streetcars measures vehicle departures from “end terminals.” To be considered on time, buses and streetcars must leave from an “end terminal” within a 6-minute window: “within 59 seconds earlier or five minutes later than their scheduled departure time.”
The TTC’s official service standard is to have 90% of all trips depart end terminals on-time and to have 60% of all trips arrive at end terminals on-time. However, the monthly KPIs that get reported on publicly only report on departures from end terminals.
What’s wrong with the way TTC tracks “On-Time Performance”?
The TTC does not report publicly on whether buses, streetcars, and subways stick to the schedule along the route.
What this means for riders: If your bus is 30 minutes late picking you up at your stop, the TTC might still count your bus as being “on time” if it left the origin terminal of its route within 6 minutes of its scheduled departure time, and arrived at the end terminal within 6 minutes of its scheduled arrival time.
Because the TTC is only measuring schedule adherence at the terminals, the solutions they have implemented to improve “On-Time Performance” metrics are not making service more reliable and could make service worse from a rider’s perspective. One approach is adding more time to the schedule (also known as adding “run-time,” or “padding” the schedule) to give transit operators more leeway to make it from beginning to end on time. But this only means that service becomes slower and less frequent, and does not address the real problem: Whether the bus is reliable along the rest of the route.

The TTC reports on system-wide averages, instead of data by route and time period
The charts produced by Steve Munro and data above based on route-by-route data broken out by time periods show how divergent route-by-route performance is compared to system-wide averages. Using an average for the entire system over the course of a month hides how poorly some routes are performing.
“Headway Performance”
When service is frequent and headways, or the scheduled gaps between vehicles, are shorter, riders care more that the service is evenly spaced than whether it is early or late by a few minutes. The TTC’s official service standards use three different measurements for headway performance, depending on how frequent the service is. The TTC’s goal is to have 60% of all trips meet this headway performance standard.
For service that arrives every 10 minutes or longer, the vehicle is considered on time when it arrives within 1 minute early and no more than 5 minutes late. When vehicles are less frequent, being on time is even more important especially when riders are trying to make connections to other routes.
For service that arrives every 5 to 10 minutes, a vehicle is considered on time when the “headway deviation” is less than 50% of the scheduled headway. For example, if your streetcar is scheduled to arrive every 6 minutes, the next streetcar will be considered on time if it arrives between 3 minutes and 9 minutes after the previous one.
For service that operates more frequently than 5 minutes, a vehicle is considered on-time when the “headway deviation” is less than 75% of the scheduled headway. For example, if your bus is scheduled to arrive every 4 minutes, the next bus will be considered on time if it arrives between 1 minute and 7 minutes after the previous one.
What’s wrong with the way TTC tracks “Headway Performance”?
The TTC does not publish headway performance metrics as part of its monthly CEO reports about on-time performance. This lack of reporting combined with the fact that the TTC’s goal is to have only 60% of all trips meet this headway performance standard has allowed bunching to become a problem with little accountability.
Short Turns
A “short turn” occurs when a transit vehicle is turned back in the opposite direction before the end of its route. This means that it technically goes out of service before it reaches the intended destination. The TTC reports on short turns each month in the CEO’s report.
The TTC’s current target is a 1% short turn rate for streetcars and 0.1% for scheduled bus trips (equivalent to about 200 bus turns and 265-275 streetcar turns per week, depending on the service board period). Short turns used to be counted rather than reported as a percentage, and are generally under-reported, according to analysis by Steve Munro.
What’s wrong with the way TTC tracks short turns?
Although short turns can be disruptive to transit riders, they can be an important tool to fix service gaps that arise from emergency road closures or when bunching occurs. The goal of minimizing short turns can be at odds with the goal of reliable service. For example, in the image below from NextBus showing the 7 Bathurst bus, only 2 of 13 buses were travelling southbound. Short-turning one or more of the buses to fill the southbound gaps would have reduced the significant gap in service. Although travellers on a short-turned bus are temporarily inconvenienced, the image demonstrates how another northbound bus would be along shortly to pick them up again, while travellers going southbound on Bathurst would avoid a major wait time.

Screenshot of NextBus, 5:13 PM EST, September 14, 2023 via Steve Munro.
6. What TTC operators say about bunching
TTCriders held a focus group in October 2024 with streetcar, bus, and subway operators to understand what happens when bunching and gapping occurs and ideas for change. Transit operators see schedule reliability as a workplace safety issue: when transit is unreliable, the public takes out their frustrations with operators on the frontlines.
"When I was first operating, if you were 3 minutes [behind schedule], you knew you were going to get a call." - TTC operator
We’re supposed to have supervisors but they’re looking at dots on a screen like a video game, while we [operators] get exposed to that anger and aggression." - TTC operator
Operators describe calling Transit Control supervisors at the Hillcrest Complex like calling "faceless customer support at your bank." Because supervisors are centralized at Hillcrest, they lack knowledge of local routes and cannot always provide advice when support is needed.
In the past, operations supervisors were assigned to each bus division and were located in garages. This ensured that operations supervisors were experts in their area of Toronto and fostered stronger relationships and a feeling among operators that they were being supported by operations supervisors.
Long-time operators described a change in culture among both operations supervisors and operators, with new operators trained in different ways when it comes to schedule adherence. Operators also shared concerns that there are fewer operations supervisors now that the process has been centralized at Hillcrest. Operators described long wait times after calling Transit Control about scheduling issues. When calls do get connected, some supervisors are unable to provide meaningful advice because of the volume of calls they receive or their lack of local knowledge.
7. What TTC operations supervisors say about bunching
TTCriders held a focus group in January 2025 with operations supervisors to understand what happens when bunching and gapping occurs and ideas for change. Operations supervisors are responsible for managing service along bus and streetcar routes.
Operations supervisors confirmed that a reduction in staffing occurred when supervisors were centralized at Hillcrest and removed from working out of the divisions:
Bus division
|
Operations supervisor staffing level at division prior to centralization
|
Operations supervisor staffing level after centralization at Hillcrest
|
Queensway
|
5 operations supervisors
|
2 operations supervisors, 7 days / week
|
Birchmount
|
5 operations supervisors
|
2 operations supervisors, 7 days / week
|
Malvern
|
5 operations supervisors
|
3 operations supervisors Monday-Friday, 2 on weekends
|
Wilson
|
5 operations supervisors
|
3 operations supervisors Monday-Friday, 2 on weekends
|
Mount Dennis
|
5 operations supervisors
*Mount Dennis used to have 1 supervisor dedicated to Jane Street alone.
|
3 operations supervisors Monday-Friday, 2 on weekends
|
Arrow Road
|
5 operations supervisors
|
3 operations supervisors Monday-Friday, 2 on weekends
|
Eglinton Division
|
5 operations supervisors
|
3 operations supervisors Monday-Friday, 2 on weekends
|
McNicoll
|
McNicoll opened during centralization process (2016-2018).
|
3 operations supervisors (Monday-Friday), 2 on weekends
|
Operations supervisors described the challenges that understaffing poses to managing bus service and supporting operators. In one example, a supervisor was responsible for managing the busy 7 Bathurst route in addition to 11 other routes.
Operations supervisors have decades of experience managing service and are empowered to use service management tools such as short-turns. Operations supervisors described a culture of fragmented responsibilities between various departments and types of supervisors at the TTC.
Operations supervisors feel they are best placed to implement a pilot program to minimize bunching and gapping because their job is to manage service. They expressed concerns that if another staff position is responsible for carrying out the pilot, these staff would be limited to tracking data and not be empowered to manage service on the spot. They feel that a visible on-street presence of operations supervisors will contribute to better service management as well as public safety and an ability to answer rider questions.
8. Recommendations
1. Change how service and on-time-performance metrics are measured and and create a public dashboard to report on these metrics
Reporting on system-wide averages hides how poorly some routes are performing. The TTC should create a public dashboard that reports on meaningful metrics, including on-time performance statistics by route and time period, headway performance, loading/crowding data, and service frequency (headways), vehicle-kilometres-travelled (VKT), revenue kilometres, service hours, and revenue hours.
On-time performance and crowding data should be disaggregated to report on routes and time periods, not system-wide averages. The TTC CEO’s monthly report should report on exceptions, to highlight poorly performing routes that need attention.
As recent analysis by Steve Munro demonstrates, transit riders are short-changed by metrics that only report on service hours: "If the capacity of streetcar routes in peak periods were restored to 2019 levels, the TTC would need 40 more cars in service. On the bus network, considering only the routes analysed here, they would require about 150 more buses."
A graph presented recently by TTC staff demonstrates the effect that congestion has had on service. Despite an increase in service hours, the TTC’s “bus service kilometres” has not risen substantially since 2013:

Image: TTC staff presentation, January 10, 2025.
The TTC’s KPI reports currently count bus and streetcar service to be “on-time” when a vehicle departs 1 minute early or 5 minutes late. This measure does not account for whether a bus arrives at stops on time over the duration of the route. Routes with vehicles scheduled more frequently than every six minutes could see two vehicles bunched together and leave at the same time, but still get counted “on time” under the current metric. The TTC should publicly report on headway reliability or some other indicator that would measure how reliable service is between departure and end points.
As long as TTC prioritizes being “on time” to a schedule as opposed to managing service so that it is evenly spaced, the problem of bunching will linger. The TTC’s current metric disincentivizes keeping service evenly spaced. The service management pilot project is an opportunity to experiment with managing service to a headway (e.g. keeping spacing between vehicles even, such as every 5 minutes).
TTCriders submitted a Freedom of Information request to obtain on-time performance data, broken out by route and time period. This information should be available online for anyone to access.
The TTC has occasionally produced bus crowding heat maps to illustrate reports using its “APC [Automatic Passenger Count] Ridership Tool” (see images below). This granular crowding data should be publicly available on the TTC’s website, not only to particular apps. These heat maps reveal that TTC has more granular data available internally than what it makes available through public data; Currently the TTC only makes three crowding ranges available (0-10%, 10-90%, 90-100%).
2. Report regularly on headway performance, so the public can evaluate the success of the TTC’s bunching pilot
A pilot to reduce bunching and gapping is a welcome initiative that will help maximize investments in more frequent service and transit priority measures. However, if the TTC does not begin to publicly report on headway performance, it will be difficult for the public to evaluate if the pilot has been successful. The TTC’s own internal target of 60% headway adherence should be the bare minimum goal for the pilot to meet. Like on time performance data, headway performance data should be available publicly and disaggregated by route and time period so that averages do not hide poorly performing routes.
3. Hire additional operations supervisors to implement a bunching reduction pilot
The TTC’s pilot program to reduce bunching and gapping with enhanced on-street presence is being implemented using “7 existing resources,” without the staff role being specified. Operations supervisors feel they are the best-placed staff to implement a pilot program to minimize bunching and gapping because their job is to manage service. If another staff position is responsible for implementing the pilot, these staff would be limited to observing and recording data that the TTC already collects, and would not be empowered to manage service on the spot. This approach would have a limited impact and limited lessons to apply city-wide.
4. Change how service is managed with adequate resources
Because the TTC does not publish on-time performance, headway performance, and other metrics on a route-by-route basis, improving service reliability has not been a priority. The TTC has all of the data and tools it needs to manage service by instructing operators to pause or slow down slightly to restore even spacing between vehicles, but needs a culture change and more resources to hire additional operations supervisors. Transit operators support bringing operations supervisors back to bus divisions, so that supervisors have knowledge of local routes and issues.
The TTC has piloted service management in the past. For example,The TTC piloted a new service management model in 2008 after public complaints about 501 Queen streetcar service. The TTC also ran a pilot in 2014 to reduce short-turns and increase reliability on the 512 St. Clair streetcar and the 29 Dufferin bus.
5. Communicate honestly with riders about service management
Communicating to transit riders through audio and visual announcements about how and why service is being managed will be essential. Tools to manage service, such as short turns or pausing at a bus stop to get back on schedule, can otherwise be highly disruptive and frustrating to riders. For example, Transport for London uses pre-recorded audio and visual announcements to communicate with riders when an operator is being instructed to pause, which mitigates tension between transit users and operators. Improving real-time predictions by adopting a benchmark and reporting on accuracy will also build trust in TTC reliability.

How Transport for London communicates with riders about service management: “The driver has been told to wait at this bus stop for a short time to help even out the service.”
6. Report on the accuracy of real-time predictions shared with apps and update GTFS more frequently
Many riders use apps to plan their routes and leave on time to catch the bus. Nothing is more frustrating than showing up according to an app’s instructions, only to have missed the bus. More accurate real-time information is one of the number one things that riders say would get them to take public transit more, according to quarterly surveys by one transit prediction app. The TTC should adopt a benchmark and report on the accuracy of real-time predictions. For example, TransitApp has developed the ETA Accuracy Benchmark, a methodology for measuring real-time accuracy.
In addition, the TTC should put more effort into providing better data when vehicles aren’t following the scheduled route. When there are long term unscheduled detours that can last weeks a new GTFS schedule should be published and applied to the real-time prediction systems. The MBTA is a leader in this area that could be looked to as a model; they generate new schedules a couple of times per week to include unplanned detours. This could be done for planned subway shuttles as well. For short term detours the TTC should look into applying the features available in the BusTime and the GTFS real-time Trip Modifications to make the details available to all transit prediction apps.
9. Technical Backgrounder: How the experienced wait time is calculated
TransSee is a real-time bus tracking and next vehicle predictions website: www.transsee.ca.
Approximately every 30 seconds TransSee collected the GPS location of all TTC vehicles, as well as which route and trip they are on from the TTC open data feeds.
Using the GTFS schedule and stop location data the TTC also provides TransSee can calculate and record when each vehicle arrives at each stop. When this happens it also calculates the average scheduled time between vehicles for that stop (headway) as well as the actual time since a vehicle from the same route served that stop (gap). These two values are what is used in this calculation. When the gap is more than 40 minutes, 4 times the scheduled headway or the vehicle appears to be going backwards then the vehicle stop isn't included in the calculations.
The "Average scheduled headway" is the average scheduled time between vehicles for all vehicle stops on the route. The "Average actual time between vehicles" is usually only a little higher because most service reaches the stops, but because it is an average, it does not reflect bunching.
The "Additional wait time weighted by odds of experiencing it" column is calculated next. The formula is sum((gap-headway)*gap)/sum(gap). This is the total difference between the gap and headway, which is how much additional time a passenger would wait for a vehicle beyond the scheduled headway for gapped vehicles weighted by the gap. If the vehicle is bunched the value would be negative, but because the gap is low this value would get less weighting. The difference is used so variations in headway don't appear as excess wait time.
The "Time between vehicles weighted by odds of experiencing it" is the "Additional wait time weighted by odds of experiencing it" added to the "Average scheduled headway".
The "Percent of additional wait time vs scheduled headway" is the "Additional wait time weighted by odds of experiencing it" as a percentage of the "Average scheduled headway".
10. Appendix: How do other transit agencies measure and report on service?
|


|
✅Chicago’s interactive dashboard allows users to see reliability data at a system-wide level or for each route. Chicago reports on:
- % of bunched intervals (% of times when the interval between 2 buses is 60 seconds or less)
- % of big gaps intervals (% of times when the interval between 2 buses is double the scheduled interval and greater than 15 minutes)
- Excess wait time: average actual wait time vs. average scheduled wait time.
|
|
 |
✅Boston’s interactive dashboard allows users to see reliability data for every bus and rapid transit route. Data over the previous 30 days and data for peak and off-peak time periods is reported.
✅Compared to the TTC, which uses the same on-time performance measure for all routes, Boston uses a narrower window for more frequent routes:
“Frequent routes have buses that come every 15 minutes or less. We measure their reliability by calculating the percentage of buses that depart no more than three minutes late.
Infrequent routes have buses that come more than 15 minutes apart. We measure their reliability by calculating the percentage of buses that depart no more than one minute early and no more than six minutes late.”
|
|


|
✅San Francisco’s interactive dashboard allows users to see reliability data at a system-wide level or for each route.
✅ San Francisco reports on the percentage of scheduled transit service hours delivered.
✅San Francisco reports on the operational availability of elevators
|
|


|
✅New York's interactive dashboard allows exploration of bus speeds, additional time waited at transit stops or on vehicles, on a route-by-route and borough-by-borough basis or at a system-wide level:
"Additional Bus Stop Time (ABST) is the average added time that customers wait at a stop for a bus, compared with their scheduled wait time."
"Additional Bus Travel Time (ATT) is the average additional time customers spend onboard the bus beyond their scheduled time."
|
11. References
501 Queen Streetcar Route: Status Update (October 2008): https://cdn.ttc.ca/-/media/Project/TTC/DevProto/Documents/Home/Public-Meetings/Board/2008/October-23/Supplementary_Reports/501_Queen_Streetcar_.pdf
501 Queen Streetcar Route: Status Update (February 2009): https://cdn.ttc.ca/-/media/Project/TTC/DevProto/Documents/Home/Public-Meetings/Board/2009/February-18/Reports/501_Queen_Streetcar_.pdf
TTC Service Standards And Decision Rules For Planning Transit Service (May 2024): https://cdn.ttc.ca/-/media/Project/TTC/DevProto/Documents/Home/About-the-TTC/Projects-Landing-Page/Transit-Planning/Service-Standards_May-2024.pdf
TTC Monthly KPIs Report, November-December 2024: https://cdn.ttc.ca/-/media/Project/TTC/DevProto/Documents/Home/Transparency-and-accountability/Reports/KPI-Reports/Nov-Dec-Monthly-TTC-KPIs-Report.pdf
TTC Monthly KPIs Report, January 2025: https://cdn.ttc.ca/-/media/Project/TTC/DevProto/Documents/Home/Transparency-and-accountability/Reports/KPI-Reports/January-2025-Monthly-KPI-Report---External.pdf
TTC 2025 Budget: https://cdn.ttc.ca/-/media/Project/TTC/DevProto/Documents/Home/Public-Meetings/Board/2025/January-10/1Recommended2025TTCOperatingBudget20252034-CapitalBudgetandPlan15YearCIPandREIPUpdate.pdf
TTC 2025 Budget slide presentation: https://cdn.ttc.ca/-/media/Project/TTC/DevProto/Documents/Home/Public-Meetings/Board/2025/January-10/2025_Budget_Presentation_TTC_Board.pdf