Spanish Cadastre/Buildings Import/Proposal

From OpenStreetMap Wiki
Jump to navigation Jump to search
Import guide Projects management Results Documentation

This page describes the proposal made in 2017-2018 for the ongoing import of data provided by the INSPIRE Services of Cadastral Cartography of the General Directorate of Cadastre in Spain. If you want to participate actively go to the import guide. If you want to know all the details of the proposal, read on.

Goal

This project aims to manually conflate buildings and addresses imported from the INSPIRE Services of Cadastral Cartography of the General Directorate of Cadastre.

Schedule

  1. February 2017: Preparation of the import proposal in the wiki and discussion in the talk-es list (in Spanish).
  2. November 2017: Development of the data conversion tool.
  3. January 2018: Testing and validation of the data conversion tool.
  4. February 2018: Annotation in the Import/Catalogue and Submission of the import proposal to the Import List.
  5. March 2018: Publication of the first import projects.
  6. April 2018: Announce in the weeklyOSM #402 in many languages.

Data to import

Data description

The Directorate General of Cadastre, following the European Directive INSPIRE (Directive 2007/2/CE, Infrastructure for Spatial Information in the European Community), provides the INSPIRE Cadastral Cartography Services that allow the download of predefined geographic data sets using ATOM technology.

Among the geographical data that the Directive requires to harmonize is the Addresses (AD) and Buildings (BU), object of this importation.

These services allow the complete download by municipalities of the different INSPIRE datasets, which are updated twice a year.

Data source site

The links of the ATOM services are:

License of access and use

The license of access and use of the INSPIRE Services of Cadastral Cartography (PDF 100.5 kB, English version) allows the data offered by these services to be used to improve OSM. In summary:

  1. The license authorizes any use, including commercial use, provided that the cadastral information has been subject to any prior process of transformation.
  2. Data may be outdated or incorrect.
  3. The General Directorate of Cadastre is not responsible for what we do with this data.
  4. We may not supplant the products or services offered by Catastro. The transformed data can not be identified as "cadastral cartography", "cadastral information", or equivalent terms that may lead to confusion.

Regarding the photos of facades accessible through links in the information service of buildings, the General Directorate of Cadastre authorizes its use to check house numbers citing the source.

Size of the data set

The data published in September 2017 contains 12 million buildings. With the data obtained during the validation of the conversion process, it can be estimated that the import would have a size of 177.8 million nodes, 33.2 million ways and 1.0 million relations. The OSM objects would comprise 15.5 million buildings (in 3D) and 770 thousand swimming pools. These are statistical and potential values, which would not be incorporated at once into the database, but will be divided into several phases according to the work plan.

Geographic distribution

This appendix shows the geographical distribution of the data.

Data conversion

The procedures to download the data, their full description, tagging plans for the data and the changesets, data reduction and transformation are described in this page: data conversion.

Some sample files are available in this repository. Files in .osm.gz format are the direct result of the process after executing the conversion tool. The files in .osm format are the result after having made a manual validation.

Data conflation

This is the strategy to avoid conflicts with existing data:

  • Buildings and swimming pools will be combined manually by the active participants in the import.
  • Where addresses are imported in a zone:
    • Thoroughfare (highway) names from the Cadastre will be replaced automatically with existing names from OSM. This process will be reviewed by a group of people devoted to preparing import projects in each zone.
    • House numbers will be automatically excluded if found in the OSM data. The remaining house numbers will be reviewed by the active participants in the import using other sources including facade images provided by Cadastre. The managers of the import projects will take the decision to import addresses or not for each zone, and may exclude any street in case of doubt.

Data import

As we have detected some issues that require some editing, we plan to do a manual import reviewing each object to import. The import comprises two interrelated stages with two participant profiles:

  • Import preparation: This part requires the download and transformation of the data for a zone using CatAtom2Osm, following the instructions in projects management. This is almost automatic for building but requires the intervention of people with local knowledge to review the thoroughfare names conflation and take the decision of which to exclude. Data will be transformed and published by successive projects in the Gestor de Tareas open to public participation.
  • Active import phase: Any one can participate in an active project following the import guide. This include the use of a dedicated account per user. In this phase collaborators will review each object to import, fixing the detected errors, conflating building with existing objects and reviewing addresses at a house number level.

Work plan

The amount of data is huge and is updated twice per year. The time needed for reviewing and import all the data will be considerable. For this reason, we don't plan to transform and publish all the data in a centralised way, but split the work by zones and create projects in different phases following this work plan:

  • Management by zones: To start the import in a zone, it will be necessary for one or more collaborators to take charge of the import preparation in this zone.
  • Projects in the Task Manager: To review and upload the data encouraging community participation we will use the Tasking Manager. Work in the TM is defined by projects. The conversion tool generates, for each municipality, a project definition file that allows us to create a project. Depending on the size, several municipalities can be grouped by project, or a large one can be divided into several projects.
  • Breakdown of the work in phases: We recommend publishing the projects for a zone in phases, with an increasing degree of complexity. Each phase acts as training for the next, so you must complete some municipalities before starting the next phase. Priority should be given to well known municipalities, those that are easy to check on the ground and those with less buildings mapped. The scheme to follow, adapted to each zone, will be:
    • In a initial phase give priority to municipalities with less than 1,000 buildings to import.
    • Use the next phase for municipalities with a number of buildings between 1,000 and 10,000.
    • Use the last phase for municipalities with more than 10,000 buildings.

Work team profile

  • Any volunteer could participate in the active import phase following the import guide.
  • You can also participate in the import preparation following the projects management guide.

As described in the work plan, we propose a team approach that includes the participation of all types of collaborators, with profiles ranging from the "on the ground" to the desktop mapper, from the most experienced to the newbie, promoting the community and learning of new skills.

Quality control

This import is intended to incorporate data to OSM with the highest possible quality, so we will carry checks before and after uploading the data.

  • Data reduction and error corrections using the conversion tool.
  • Review of the problems report found in the conversion process before data publication.
  • Use of JOSM/Validator before data upload.
  • Regarding addresses, the work plan includes first reviewing the existing OSM thoroughfare names, checking the real house numbers and importing only in case of no doubts.
  • We want to reduce and control the errors before import, but to be sure, each project will be followed by a review using Quality assurance tools.
  • Building uploads in Spain will be monitored to check that massive imports are not performed.

Updates

Cadastre data is updated twice a year. We will developed a feature in the data conversion tool to compare two sets of data and generate files with the differences between them. The users responsible for each zone to be imported will be in charge of keeping the Cadastre files used to make the comparison later, and of periodically updating those changes that occur.

Regression plan

If necessary, the changes will be undone using the JOSM/Plugins/Reverter plugin.

Results

Check on this page the progress of the import jobs.

References