Mobility Cooperative Switzerland Data Semi-Automated Import Action
>> Status: Anyone can now validate using cf_audit (see below).<<
This page documents the semi-automated import, i.e. the update and integration of Mobility Cooperative data of Switzerland (website www.mobility.ch). Mobility Cooperative Switzerland gives OSM openly licensed data upon periodical request. They currently own around 1'500 stations.
Current tasks are described in the following section. The semi-automated import action consists of a workflow of tasks: see below. This workflow can potentially be repeated sporadically in the future as Mobility updates it's dataset monthly.
For the reporting and discussion of issues see tab Discussion of this page!
- Setup this page; documented pre-processing steps.
- Unnecessary notes:de/fr/en deleted. Changed tag "amenity=car_rental" to "amenity=car_sharing" for a half dozen nodes who had Mobility in name or operator. See discussion.
- Pre-processed data using OSM Conflator and setup validataion task.
- Informed talk-ch and email@example.com
- Validate the <415 Locations!
- Create dedicated Mobility-Import-Account (any volunteer?).
Validate with this the cf_audit web app (March 2018, Step5, All): https://cfaudit.nicj.ch/project/Mobility%20Update
- It's real data but nothing is changed yet in the OSM database. The output of your validation goes to a JSOM dataset which will be uploaded at once later through JOSM with a dedicated OSM account.
- There are mainly two cases when you validate (~200 nodes each): A. "Create new node" and B. "Update tags on a point".
- In case B. "Update tags on a point" you have to click on the (small) tag change to "Mobility Cooperative" and can't move the node. In case you recognize that the location of the stations needs to be checked, pls. fill in fixme tag value e.g. with "Check location".
- In case A. "Create new node" the exact location is often unclear too given the imagery and the remarks (i.e. the description coming from the original Mobility dataset). So you can also fill in the fixme tag value with "Check location".
- Click "Good" - or if unsure "Skip".
- Currently, there's "ref:mobility:station_id" in the validation, but in the current OSM data there's "mobility:station_id" (without "ref:"). See issues in Discussion page.
A Mobility Station
This is a typical Mobility station entry (an OSM Node):
Mandatory tags: * amenity=car_sharing * mobility:station_id=* * name=* * operator=Mobility * website=http://www.mobility.ch/ Optional tags: * opening_hours=24/7 * capacity=* * wheelchair=limited Deprecated tags: * source=http://www.mobility.ch/ => belongs to Changeset text. * addr:city=* * addr:country=* * addr:postcode=* * addr:street=*
Workflow of Action
Main Action Tasks:
- Pre-processing tasks using OSM_Conflator/Audit, see "Dataset Pre-processing Documentation" below (done currently by user Geonick).
- Public validation tasl using cf_audit (currently hosted by nicj.ch).
- JOSM upload with dedicated account (to be defined).
Checklist according to the OSM Import/Guidelines:
- Step 1 Prerequisites - Status: Done.
- Step 2 Community Buy-in - Status: Done. IRC and local meeting; mail to talk-ch and firstname.lastname@example.org mailing lists.
- Step 3 License approval - Status: Done, obtained.
- Step 4 Documentation - See also this mailinglist. Status: Written a wiki page (this). Done: Add line to Import.
- Step 5 Import Review - Note that this is rather an update not an import. Status: In process: See above.
- Step 6 Uploading - Progress is tracked here. Status: Dedicated user tbd. No action date yet.
Mobility is a car sharing cooperative, a market leader in Switzerland for car sharing. One of it's first partners was Swiss railways SBB.
The original dataset contains attributes like station_id, name, capacity, as well as postal address (street, postcode, city, country) and remarks.
Last years an import into OSM, then a large update happened, which both have had some issues like a note which implied to not touch the node, which was inadequate and led to misunderstandings.
In fact the accurracy of the original coordinates is constantly inaccurrate. It seems like Mobility is geocoding it. This means that mappers have to manually check the location which currently implies almost all existing entries, an in future also all new entris from an eventual import.
Dataset Pre-processing Documentation
- NEW: Add amenity=car_sharing, name=*, operator=Mobility Cooperative, website=http://www.mobility.ch/
- UPDATE: mobility_id- / 0.001m-match and name change => Update tag name.
- UPDATE: mobility_id- / 30m-match and name change => Update tag name
- source=Mobility Cooperative is part of changeset.
- Neue Nodes => Geometrie überprüfen
- Als bestehend erkannte (über Radius 100m) erhalten
- operator=Mobility Cooperative
- Bestehende "city, postcode, country" stehen lassen
- Falls nicht vorhanden, neu "city, postcode", ev. street dazu tun.
- Tag source löschen, und später (im JOSM Upload) source auf changeset tun mit "Mobility Cooperative"
- source ersetzen, website-value "https://www.mobility.ch/" neu mit https