Short-term Traffic volume data (traffic counts and turning movements) from the FLOW database and other data sources.
- Introduction
- What is counted?
- Where does it come from?
- How often is data updated?
- Where can I access the data?
- How is the data structured?
- Relevant Tables
- Useful Views
- Cycling Seasonality Adjustment
The City of Toronto collects ad-hoc traffic volume data for projects and service requests. The traffic data collection program serves many internal transportation projects and operations teams for project planning, capital planning, engineering design, project analysis, and operational functions like signal timing.
The most common traffic studies conducted are the Turning Movement Count (TMC) and the Automated Traffic Recorder (ATR) count. TMCs observe movements of motor vehicle, bicycle, and pedestrian volumes at intersections. ATRs observe volumes, speeds, and vehicle classification of motor vehicles travelling along a section of road.
Other studies include pedestrian delay and classification, pedestrian crossover observation, stop-sign compliance, queue-delay, cordon count, and radar speed studies.
- Type of road user: car, truck, bus, bicycle, pedestrian, other
- Intersection approach leg: N / S / E / W
- Type of movement:
- Motor vehicle: Through / Left / Right
- Cyclist: cyclist volume by approach leg
- Pedestrian: pedestrian volume by leg of intersection crossed
- Location Identifier (SLSN Node ID)
- 15 min aggregated interval time
- 15 min aggregated volume per movement (turning and approach) by:
- vehicle types
- cyclists and pedestrian counts are approach only
- No regular data load schedule
- Data files collected by 2-3 staff members
- Manually geo-reference volume data to an SLSN node during data import process
- Counts are typically conducted on Tuesdays, Wednesdays, and/or Thursdays during school season (September - June) for 1 to 3 consecutive days
- If collected data varies more than defined historical value threshold by 10%, the collected data will not be loaded
- Volumes are available at both signalized and non-signalized intersections
- Each count station is given a unique identifier to avoid duplicate records
- Data will not be collected under irregular traffic conditions (construction, closure, etc), but it maybe skewed by unplanned incidents
- Volume
- Direction
- Volume
- Speed
- Direction
- Speed bin
- Volume
- Vehicle classification
- Direction
- Vehicle classification
- Volume
- Location Identifier (SLSN Link ID)
- Direction
- 15 min aggregated interval time
- 15 min volume
- typically aggregated by direction, although data may be available by lane
- The counts represent roadway and direction(s), not on a lane-by-lane level
- No regular data load schedule
- Manually geo-reference volume data to an SLSN node during data import process
- Typical ATR counts 24h * 3 days at location in either 1 or both directions
- Each PCS/ATR is given a unique identifier to avoid duplicate records
The City of Toronto retains a traffic counting contractor who conducts data collection. They schedule and install temporary counting equipment, or send staff into the field to observe volumes, and process data.
Data are collected through various technologies. Originally, data were collected by field staff who would manually observe and record volumes. Pneumatic road tubes were introduced to record motor vehicle volumes, speeds, and classifications. More recently, counting has shifted to video observation.
Once the City receives data from the contractor, staff load the data into our database. Until recently, staff would load data files into a legacy Oracle database through an application called "FlowLoad". Data would then be retrieved through a user interface application called "Flow", where the data were formatted into nice reports. In recent years, "Flow" was replaced by MOVE.
Traffic counts are conducted ad-hoc, usually on request for a specific project need. As such, the data is not necessarily systematically collected. We do not have comprehensive coverage across time or space.
TMCs are processed automatically, nightly, once made available from the contractor. ATRs are loaded manually by staff once data files are received from the contractor.
Internal to the Transportation Data & Analytics team, data flows from legacy Oracle database, nightly to MOVE (flashcrow RDS), and is then replicated to the bigdata RDS.
Look in the traffic schema for all ad-hoc data tables.
The database is structured around three types of tables: metadata, count observations, and reference tables (spatial, temporal, or categorical).
- Turning Movement Count (TMC)
countinfomics: metadatadet: count observations
- Automated Traffic Recorder (ATR)
- Spatial reference
arterydata: an internal reference system that maps a count to a location, used by both TMC and ATR tables
- Other reference
category: reference table for traffic count type or data source, used by both TMC and ATR tables
The following diagram shows the relationship between the above-mentioned tables.
A human-friendly interpretation of studies. Grouped by colocated arterycodes and with single-day ATR "counts" into continuous study days.
Find at traffic.studies.
A gaggle of cascading tables that aggregate "counts" into "studies" (counts at the same location that occurred on continuous days) and "arterycodes" into "arterycode groups" (counts that occured at the same location). These intermediary tables are used to create studies.
traffic.arteries_groups- arterycodes describe a count-location
- for ATR counts, an arterycode also describes a direction
- for a two-way street where a count was conducted to observe traffic in both directions, two arterycodes exist for this one road segment
- as such, there can be multiple arterycodes that exist at the same physical road segment
- this table groups arterycodes that belong together at the same location
traffic.arteries_centreline- a mapping of legacy arterycodes to current centreline nodes and segments
- this is the output of the MOVE conflation
traffic.counts_multiday_runs- aggregates single-day ATR "counts" into continuous multi-day "studies"
traffic.arteries_counts_groups- brings together artery groups and count groups
- single-day counts aggregated into continous studies (
count_group_id) at colocated arterycodes (artery_group_id)
Recent TMCs (September 2023 and on) loaded through new mechanisms. Includes 14-hour TMC data. Designed to mimic the legacy data tables for backwards compatibility.
tmc_metadata_legacy- includes additional metadata like centreline, geometry, corresponding study request, and human-readable location name
tmc_study_legacy
Recent ATRs (May 2022 and on) loaded through new mechanisms. Includes speed and volume data only. Designed to mimic the legacy data tables for backwards compatibility.
atr_metadata- includes additional metadata like centreline, geometry, corresponding study request, and human-readable location name
atr_study
This table contains Turning Movement Count metadata only. This table contains the location reference, date, and source for each Turning Movement Count. Each Turning Movement Count is defined by a unique count_info_id.
| Field Name | Type | Description |
|---|---|---|
| count_info_id | bigint | Unique ID for a count linked to det table containing detailed count entries |
| arterycode | bigint | ID number linked to arterydata table containing information for the count location |
| count_type | varchar(1) | Count hours1 during which data are recorded, Routine (R) or School/Pedestrian (P) |
| count_date | date | Date on which the count was conducted |
| day_no | bigint | Day of the week (ISO standard; 1 = Monday, 7 = Sunday) |
| category_id | int | ID number linked to category table containing the text description of the count type or source |
1 - Routine and School Hours
Routine hours are the "typical" hours during which data would be collected. School hours were specifically selected to observe school pickup, dropoff, and lunch periods. We are moving towards continuous collection periods (e.g. 6:00am-8:00pm), but legacy data are still reported during these standard 8-hour disaggregate periods.
- Routine Hours: 7:30 - 9:30 / 10:00 - 12:00 / 13:00 - 15:00 / 16:00 - 18:00
- School Hours: 7:30 - 9:30 / 10:00 - 11:00 / 12:00 - 13:30 / 14:15 - 15:45 / 16:00 - 18:00
This table contains individual data entries for Turning Movement Counts in 15-minute non-continuous increments. This is a "wide" format, where each direction-mode-movement has its own column. For a long (instead of wide) version of this table, see the matview traffic.tmc_miovision_long_format.
| Field Name | Type | Description |
|---|---|---|
| ID | Autonumber | Autonumber function |
| COUNT_INFO_ID | number | Unique ID number for a count linked to countinfomics table containing count metadata (higher-level information) |
| COUNT_TIME | Date/Time | Effective time of counts (time displayed is the end time period) |
| N_CARS_R | number | S/B cars turning right |
| N_CARS_T | number | S/B cars going through |
| N_CARS_L | number | S/B cars turning left |
| S_CARS_R | number | N/B cars turning right |
| S_CARS_T | number | N/B cars going through |
| S_CARS_L | number | N/B cars turning left |
| E_CARS_R | number | W/B cars turning right |
| E_CARS_T | number | W/B cars going through |
| E_CARS_L | number | W/B cars turning left |
| W_CARS_R | number | E/B cars turning right |
| W_CARS_T | number | E/B cars going through |
| W_CARS_L | number | E/B cars turning left |
| N_TRUCK_R | number | S/B trucks turning right |
| N_TRUCK_T | number | S/B trucks going through |
| N_TRUCK_L | number | S/B trucks turning left |
| S_TRUCK_R | number | N/B trucks turning right |
| S_TRUCK_T | number | N/B trucks going through |
| S_TRUCK_L | number | N/B trucks turning left |
| E_TRUCK_R | number | W/B trucks turning right |
| E_TRUCK_T | number | W/B trucks going through |
| E_TRUCK_L | number | W/B trucks turning left |
| W_TRUCK_R | number | E/B trucks turning right |
| W_TRUCK_T | number | E/B trucks going through |
| W_TRUCK_L | number | E/B trucks turning left |
| N_BUS_R | number | S/B buses turning right |
| N_BUS_T | number | S/B buses going through |
| N_BUS_L | number | S/B buses turning left |
| S_BUS_R | number | N/B buses turning right |
| S_BUS_T | number | N/B buses going through |
| S_BUS_L | number | N/B buses turning left |
| E_BUS_R | number | W/B buses turning right |
| E_BUS_T | number | W/B buses going through |
| E_BUS_L | number | W/B buses turning left |
| W_BUS_R | number | E/B buses turning right |
| W_BUS_T | number | E/B buses going through |
| W_BUS_L | number | E/B buses turning left |
| N_PEDS | number | North side pedestrians |
| S_PEDS | number | South side pedestrians |
| E_PEDS | number | East side pedestrians |
| W_PEDS | number | West side pedestrians |
| N_BIKE | number | S/B bicycles from the north side |
| S_BIKE | number | N/B bicylcles from the south side |
| E_BIKE | number | W/B bicycles from the east side |
| W_BIKE | number | E/B bicycles from the west side |
| N_OTHER | number | North side - optional field |
| S_OTHER | number | South side - optional field |
| E_OTHER | number | East side - optional field |
| W_OTHER | number | West side - optional field |
The following image depicts motor vehicle movements. This example shows south approach, or northbound travel, movements.
S_[CARS|TRUCK|BUS]_LS_[CARS|TRUCK|BUS]_TS_[CARS|TRUCK|BUS]_R
Notes:
- Exits can be calculated by summing associated movements.
- U-turns are currently not available in
bigdata.
At the time of writing, bike totals are reported only by the number of cyclists that enter the intersection from a given approach/leg. Turning movements are currently not available in bigdata.
Pedestrians are counted based on the side of the intersection they cross on. The example below shows S_PEDS or pedestrians crossing on the south side of the intersection. Note that they could be travelling either east or west in this example.
Pedestrians are only counted when they cross the roadway, meaning that pedestrians who turn at the intersections without crossing the roadway are not counted.
For 3-legged or "T" intersections, pedestrians have typically not been counted on the side of the intersection without a crosswalk, even when present in large numbers. The count in these cases will be given as zero. Going forward however (circa late 2024), the intention is to count that sidewalk as though it was a crossing of a typical 4-legged intersection.
Similar to TMC Metadata (countinfomics), this table contains the location reference, date, and data type/source from all sources other than Turning Movement Counts.
See TMC Metadata (countinfomics).
This table contains individual data entries for all counts or sources other than Turning Movement Counts.
| Field Name | Type | Description |
|---|---|---|
| count_info_id | bigint | Unique ID number for a count linked to countinfo table containing count metadata (higher-level information) |
| count | bigint | Vehicle count |
| timecount | Date/Time | Effective time of counts (time displayed is the end time period) (except for ATRs, where time is the start of the count) |
| speed_class | int | Speed class codes indicating speed bins associated with the 'prj_volume.speed_classes' table. speed_class=0 refers to non-speed counts. |
This table contains the location information of each volume count.
| Field Name | Type | Description |
|---|---|---|
| arterycode | bigint | ID number referred to by countinfomics and countinfo |
| street1 | text | first street name |
| street2 | text | second street name |
| location | text | full description of count location (do not use PX references, not consistent and can change without warning from upstream sources) |
| apprdir | text | direction of the approach referred to by this arterycode |
| sideofint | text | the side of the intersection that the arterycode refers to |
| linkid | text | in the format of 8digits @ 8digits, with each 8 digits referring to a node |
It's very important to understand the humble arterycode. The arterycode identifier system is an internal legacy location reference system that describes intersections and segments of the Toronto street network where a count has occurred. Arterycodes do not describe the entire Toronto transportation network.
Given an arterycode, you can find the corresponding modern-day location by cross-referencing with traffic.arteries_centreline.
This is a reference table referencing the count type or data source of each entry.
| Field Name | Type | Description |
|---|---|---|
| category_id | int | ID number referred to by countinfomics and countinfo |
| category_name | text | name of the count type or data source |
| Category Name | Meaning |
|---|---|
| 24 HOUR | Volume ATR |
| RESCU | RESCU |
| CLASS | Vehicle Classification ATR |
| SPEED | Speed ATR |
| MANUAL | |
| PERM STN | Permanent Count Stations |
| BICYCLE | Bicycle Volume ATR |
| SPD OCC | |
| SENSYS SPEED |
-
traffic.tmc_miovision_long_format- Takes the wide TMC tabletraffic.detand transforms it into a long format designed to be integrated with miovision-derived TMCs as inmiovision_api.volumes_15min_mvt. -
traffic.artery_locations_px- A lookup view between artery codes and px numbers (intersections). Created using MOVE'straffic.traffic_signalslookup. -
traffic.artery_objectid_pavement_asset- A lookup view between artery codes and objectid. Used, for example, to link an arterycode to pavement asset information in vz_analysis.gcc_pavement_asset. This view uses the intermediate tablegis_shared_streets.centreline_pavement_180430which was last updated three years ago and it will be updated via issue Update pavement assets #620.
A model was developed to adjust cycling counts for before after evaluations of new infrastructure based on sparse counts. It can be found in the cycling_seasonality folder