@roataway/infrastructure-data

Overview

This repository is related to https://github.com/roataway/web-ui and it contains information about the transport infrastructure of a city:

• details about vehicles
• route-related data (stations, route segments) extracted from OpenStreetMap

The process of obtaining the data

Route segments

These are obtained via the Overpass API.

1. Go to

https://overpass-api.de/api/interpreter?data=[out:json];relation%20(xxxxxxx)%3B>>%3Bway._%3Bout%20geom%3B

2. Replace xxxxxxx with the relation id, e.g. 8649765
3. Save the resulting JSON to data/route_<upstream_id>_segments.json, replacing upstream_id with the route's internal identifier in the upstream GPS tracking system.

// URL query as Overpass query
[out:json];
relation (8649765);
>>;
way._;
out geom;

Route stations

The process is similar to route segments, but the Overpass API query is different:

1. Go to

https://overpass-api.de/api/interpreter?data=[out:json];relation%20(xxxxxxx)%3B>>%3Bnode._%20[public_transport%3Dplatform]%3Bout%3B

2. Replace xxxxxxx with the relation id, e.g. 8649765
3. Save the resulting JSON to data/route_<upstream_id>_stations.json, replacing upstream_id with the route's internal identifier in the RTEC system.

// URL query as Overpass query
[out:json];
relation (8649765);
>>;
node._[public_transport=platform];
out;

Overpass queries can be tested at https://overpass-turbo.eu/

CSV files

Some of the information is stored in CSV files, to make it easier for stakeholders from the public transport industry directly contribute to the repository, supplying fresh data. It is assumed that they are familiar with how to use spreadsheets, so CSV is appropriate because it allows them to edit it as a table.

These details are obtained through personal observations and through interactions with RTEC staff.

Route information

id_upstream	name_concise	name_long	osm_relation
1	30	Aeroport - str. 31 August 1989	7390177
2	32	Stăuceni - Chișinău	8649765
	37	Bubuieci - Gara Feroviară	9478330

• id_upstream - the internal identifier of the route in the originating GPS tracking system that we get our data from
• name_concise - a short name for a route, usually it is a number, but it can also contain letters, e.g. 3A
• name_long - a verbose route name
• osm_relation - the relation of this route in OpenStreetMaps

Note that not all trolleybuses have GPS trackers yet, therefore not all routes are provisioned in the tracking system, and we don't know their id_upstream. In this case we simply omit it. These routes will not be shown in the interface.

Vehicles

The fields are:

• tracker_id - string identifier of tracking device installed in this vehicle
• organization - string identifier of organization that owns the vehicle, so far all of them are from PT-x, which stands for "Parcul de troleibuze nr. x"
• board_number - number shown on the vehicle itself
• vehicle_type - string enum, one of {trolleybus, bus, minibus, tram}
• model - compound string, that has the format <maker>/<model>, e.g., Skoda/14Tr.M
• door_count - number of doors in the vehicle
• release_date - date in the form DD/MM/YYYY, this is an optional field
• articulated - stringified boolean, yes if the vehicle is articulated ("гармошка")
• accessibility - stringified boolean, yes if the vehicle is designed to make it easier for disabled people to get in and out.

tracker_id	organization	board	vehicle_type	model	door_count	release_date	articulated	accessibility
000004	PT-1	1232	trolleybus	ZiU/9	3	01/11/1989
	PT-1	1278	trolleybus	BKM/AKSM-213	4	15/11/2005	yes
	PT-1	1285	trolleybus	BKM/AKSM-321	3	27/04/2011		yes
	PT-1	1286	trolleybus	BKM/AKSM-321	3	27/04/2011		yes

Note that at the moment not all vehicles are equipped with GPS trackers, hence the tracker_id field is empty for them. This information will be updated once more trackers are deployed.

Helper tools

getRoute - automatically retrieve routes metadata from OpenStreetMap.

This tool retrieves information about a given route, by taking all routes from a CSV file that was fed into it, or by explicitly retrieving one particular route specified by its OpenStreetMap relation id.

The data are stored in GeoJSON, in 2 files:

• route_<upstream_id>_segments.json
• route_<upstream_id>_stations.json

Installation:

• virtualenv -p python3 venv-getRoute
• source venv-getRoute/bin/activate
• pip install osm2geojson

Examples of use (activate the virtualenv via source venv-getRoute/bin/activate first):

• python getRoute --csv ../routes.csv --dst data - Get all the routes specified in ../routes.csv and save the resulting files to out/ in the current directory.
• python getRoute -r 9478330 - Retrieve data about relation 9478330; in this case the file name will use the relation_id, rather than the upstream_id.
• python getRoute --help - See what command line args are available

If all is well, you will see something like this:

 INFO - Processing all routes from ../src/data/routes.csv
 INFO - Processing route 30, `Aeroport - Piața Marii Adunări Naționale`
 INFO - Processing route 32, `str. 31 August 1989 - com. Stăuceni`
 ...
 INFO - Processing route 1, `or. Durlești - str. Sarmizegetusa`
 INFO - Done

And the metadata will be saved in the directory you've indicated.

Note that the Overpass API of OSM returns a JSON which does not conform to the GeoJSON schema (though they look the same to the untrained eye)! Thus, the tool relies on a third-party library, osm2geojson to do the conversion.

What happens in principle:

• retrieve XML data from Overpass
• use osm2geojson to convert that XML to GeoJSON
• save GeoJSON to file