Importing STLévis's bus routes/stops data in Lévis


I would like to propose importing Bus Stops as well as Bus routes for the STLévis service region.

The data is released under the CC-BY 4.0 license, and a waiver has been granted 1.
Data is available from the Société de transport de Lévis’s website 2

I have added the STLévis network to PTNA 3 in order to allow easy comparison of existing data in OSM and the data contained from the GTFS feed.

I propose to use the same process as was used with the import of data from the STL, with some adjustments for local data 4.

After the data has been generated, some manual conflation will be required due to some existing routes being mapped, and some routes having erroneous trip data in the GTFS as they are marked as having different directions (direction_id) when the route is a loop.

There are some stops that overlap with the RTC (Réseau de transport de la Capitale), however those are signposted seperately and should be not be conflated together

Please tell me your thoughts


L’accord écrit de la société de transport de Lévis et super ! J’essaie d’obtenir la même chose du côté de Sherbrooke depuis deux ans, mais sans succès jusqu’à maintenant.

Bonne chance pour l’import de données.

I have created a wiki page for the import, Lévis/Proposed Import of STLévis GTFS - OpenStreetMap Wiki

I have reached out to local mappers who have given positive feedback on this import.

As such, I will be moving forward with the import in the next few days, unless anyone comes and objects before then.

1 Like

Sorry, I am quite late to the party.

Reading the wiki page, I stumbled over:

In a similar fashion conflation between already merged GTFS stops and existing OSM stops is handled by checking if there are any existing stops in OSM that are with x meters (distance used was 5 meters) of each GTFS stop.

In my opinion “5 meters” is rather optimistic. Have you check the accuracy of the data regarding coordinates. I usually always find outliers up to 50 or even 100 meters in GTFS data especially if only one coordinate is used per stop (disregarding offsets per direction or multiple stop_position at a bus station with several lanes) but even with a coordinate per stop_position in the GTFS data there might be some outliers.

Considering there wasn’t much in terms of bus stops, 5m is enough to catch them.

I know there are some outliers where the coordinates don’t match the location of the actual stop, those are being corrected when I come across them.

1 Like