Description
The TSB is an independent agency, created by an Act of Parliament (the Canadian Transportation Accident Investigation and Safety Board Act). Its mandate is to advance transportation safety in the marine, pipeline, rail and air modes of transportation by:
conducting independent investigations, including public inquiries when necessary, into selected transportation occurrences in order to make findings as to their causes and contributing factors;
identifying safety deficiencies, as evidenced by transportation occurrences;
making recommendations designed to eliminate or reduce any such safety deficiencies; and
reporting publicly on our investigations and on the findings in relation thereto.
Rail Incident (Occurrence) consists of two types:
Reportable – Class 3
Non-reportable - Class 1,2,3,4,12
Originator
Transportation Safety Board (TSB)
Use Cases
As a Rail Safety Officer, I want to identify locations of frequent train derailments so that train operator can be notified to take special precaution.
As a Rail Incident Analyst, I want to be able to analyse trends and patterns of rail incidents in order to identify hot spots.
Source(s)
Description | Source | Format | Date | Update Frequency | Metadata | Cost | Link | Remarks |
---|---|---|---|---|---|---|---|---|
TSB Rail Investigation | TSB | CSV | 2018 | Monthly | Free | EGIS Data Folder: TSB_Rail.mdb | -Use preprocessed MDB file as provided. |
Remarks
TSB Rail Incident dataset is a non-spatial tabular dataset. Preprocessing to geocode the occurrence location is required using the custom master milepost database.
The data on occurrences are released on or soon after the 15th of each month, and date to the last day of the month preceding their release.
EGIS Data Folder: \\ncras382\Application Source Files\EGIS\Smartmap Files\
Master Milepost Database Creation:
Interpolation of Railway Network milepost: downloaded shp file from National Railway Network Canada Open Data (e.g., SK folder, only tracks and marker post, and grade crossing as well) is partially manual and automated (use FME and ArcGIS to clean data). Several factors can affect the calculation of mileposts such as:
Actual Segment length vs NRCAN length.
Double or parallel tracks.
Cross over tracks along double/parallel tracks.
Track branch off
Track direction
More…
Create Main Track Summary Report:
Create an aggregate of all track grouped by: OPERATORMK, RAILWAY, SUBD1END, SUBD1START, SUBD1NAME
Extract actual length
Calculate NRCAN length (i.e. SUBD1END-SUBD1START)
Build Continuous Track Segment:
Filter TRACKCLASS=Main from *TRACK.shp (there are also spurs, branches off main track, and railway yards, which we don’t need those to interpolate location)
Connect tracks to form a continuous segment grouped by the following attributes:
i) RAILWAY
ii) SUBDI1NAME
iii) FROM
iv) TOCompare/validate the length against NRCAN value.
Validate polyline start/end direction using Grade Crossing milepost for consistency. Reverse direction or flag in the database if necessary.
Validate NRCAN start/end values against Grade Crossing milepost for consistency. Readjust values in the database if necessary.
Perform manual cleanup such as segmented tracks, parallel tracks, split etc.
Repeat the process using the last modified dataset until all tracks are fixed.
Interpolate Milepost:
Join the “Connected Track” with “Main Track Summary” database using RAILWAY, SUBD1END, SUBD1START, SUBD1NAME as primary keys.
Extract actual length.
Divide Track into equal segments of 0.1 mile.
Extract/store the coordinates.
Verify interpolated milepost against Grade Crossing milepost for accuracy. Fix and adjust any error if necessary.
Store the newly created mileposts in the Master Milepost Database. **Note: Not all mileposts start at 0.
Disclaimer:
Interpolated Milepost is only approximate. It is not verified nor validated for accuracy, therefore should be used for reference purposes only.
Process for EGIS integration
CSV File Transformation:
Download Rail Occurrence Report CSV file from TSB.
In FME, use the Feature Merger transformer to join Occurrence data and Master Milepost database, using Company, Subdivision and Milepost as primary keys to extract the Lat/Long.
Insert a new attribute to hyperlink TSB Report for records with OccClassID=2 or 3.
A typical Report URL looks like this: http://www.tsb.gc.ca/eng/rapports-reports/rail/2018/r18w0168/r18w0168.asp
Replace r18w0168 with the OccNo and 2018 with OccYear.
Transform the Lat/Long to point.
Output to KML, SHP or GDB as required.
Import into EGIS
Display Optimization:
Sampled/thinned data (generated using data sampling technique) and Class 2 & 3 incidents are displayed at small and medium scale. With scale transition, original data are displayed at large scale thus improve display performance and reduce cluttering.
TC Smart Maps:
Not all GIS applications can consume KML file directly due to its unique data structures such as networklinks, folders, models etc.
Please contact Lawrence Chong, Digital Services Directorate for TC Smart Maps KML dataset and format translation.
TC Smart Maps KML link: \\tc.gc.ca\tcgis\GISData\Google Mashup\TSB_Investigation\Rail\TSB_Investigation-Rail.kmz
Customization:
Schema transformation: As required
Denormalization: No Required
Value coding: Not required
Attribute renaming: As required to improve readability
Attribute removal: It is highly recommended as some fields are unnecessary, this will also reduce file size and improve performance
Geoenrichment: Feature symbolization is recommended to improve visualization. OccClassID=2 or 3 are reportable incidents. It can be symbolized differently and display at smaller scale.
Format translation: As required.
Language translation: French version available.
Update frequency: As identified