AIV GRB building import/Background

From OpenStreetMap Wiki
Jump to navigation Jump to search

This page gives more in-depth information about the GRB import.

Main issues

The main issues we have had to solve:

  • Conversion to OSM categories: It isn’t always possible to convert the GRB categories to a corresponding OSM tag. But a good effort gets a long way. In case of doubt, we can use building=yes. The import toolset will result in warnings when trying to change existing tags. Because of how we plan to tag, QA tools will be easy to set up, to identify possible tagging mistakes. Several categories of mistakes are already identified and accounted for.
  • Existing buildings: About 1.5 million buildings exist in OSM already. The toolset will respect these objects, and their mappers. The geometry and tags will be changed but the object ID and hence their history will remain.
  • Address data: Complex addresses are flattened onto buildings in GRB. These addresses will not be imported. Instead, they can be sourced directly from CRAB, helped by local surveys.
  • Complexity of the mapper's job: To assure quality importing, the tool will be properly documented, there will be workshops and hangouts. Only people who had some training will be allowed to import into OSM with the toolset. However, we have to be careful not to be off-putting: people are already deleting-and-tracing from GRB background images or doing wild imports. The toolset will definitely improve current mapping practices.
  • Updating: Our community has the skills to keep using the GRB to detect errors once a larger amount of GRB buildings are imported. In many cases, we will be faster than the GRB, in many others we will be slower. Marc Ducobu has done a presentation at SotM about the exact same problem in Brussels, where a building import happened a lot earlier, using UrbIS data. These tools will serve as inspiration or will be extended to cover Flanders.


The Large scale Reference Database (Grootschalig ReferentieBestand) of Flanders is produced by AIV and is the data source for building outlines in this import.

This detailed database focuses on government operated infrastructure on the public domain. It contains amongst others building footprints, both road axes and road surface area outlines, roadside infrastructure, administrative parcels and water bodies. Most publicly accessible features (e.g. the front sides of the buildings) have been measured on the ground with great accuracy (typically better than 10 cm).

A detailed analysis of GRB entities and their possible translation to OSM objects is available at the WikiProject Belgium/GRB page.

Their main focus is accuracy, not recency. Changes are only made if an as-built plan is received, or when surveyors have been on the ground. Any change on the ground should become available within a year on a worst case basis. Red x.svg To do: link to details of update procedure

Only data that can be collected from public land is measured, so the backside of buildings is drawn based on high resolution aerial imagery. The building dataset contains addresses, which are periodically updated from the central address database (CRAB).

In the period 2000–2013, the GRB was established for all Flemish cities and municipalities. Everything in the public domain is checked for updates at least once a year. Since the original creation of these databases, no updates were executed at the level of the inner terrains (areas not accessible for surveyors). In 2013 an update was executed based upon terrain changes (anomalies), midscale aerial imagery on more than 5.500 km² were screened for changes at the level of the inner terrains. In total, more than 400.000 anomalies were checked and updated by means of photogrammetric mapping on the midscale imagery.

The database is maintained and updated by Flemish municipalities using administrative procedures, geodesy measurements and satellite photos. It forms the geographical basis which others can use to inoculate their data. The Flemish government expects a positive effect on the quality of the data through feedback channels. Furthermore, it expects the development of new data sets and the creation of GRB data-based applications. In December 2015, the data was made publicly available but registration was required. Red x.svg To do: and now?


This section describes all different GRB data types (entities) and their details. We also add the OSM mappings and the level of usefulness for OSM for each unique subtype by mapping objects.

Source(dutch): see

Entity list


  • Adp: Administrative parcels: Not useful due to convention that these don't belong in OSM


  • Ano: Anomalies (mistakes and new developments): should be used together with other data types, to see if a new development is happening or a mistake is reported


  • Gba: Building attachment: Partially useful for OSM.
Object description Mapping Import
afdak Overhead roof building=roof Imported
loopbrug pedestrian raised bridge, mostly covered bridge connecting two buildings (can be found between high office towers or in factories) ? Not imported
trap a steps area connected to a building (like in front of an entrance of a building) source data uses area. Problem, Proposed tag cannot be placed on area by tag convention, see highway=steps. Mapper needs to change this manually to a way. Good source data however. highway=steps Imported. Exported as an area (needed for manual mapper to correct this to way).
zichtbare onderkeldering We have no clue what this is. ? Not imported
ingezonken garagetoegang ramp of a garage entrance below ground level ? Not imported
uitbreiding AFAIK, a regular building. The meaning of the word is extension, only a few in the source data like that building=* Imported and exported
verheven garagetoegang ramp of a garage entrance above ground level ? Not imported
verdieping building that's not on the ground (so you can walk below it), but not a roof either. Need user verification. Usually in OSM a footway is already there. Connect the corners and tag accordingly building=* but add level tags and/or tunnel=building_passage on the way below Imported and exported.
  • Note: in this context, imported means "imported into our working database", not "imported into OSM"! Exported means "loaded into JOSM". Which "exported" objects are imported into OSM is then up to mapper discretion.


Object description Mapping Import
hoofdgebouw A building, almost always has associated address information building=* default: building=house. When merging with existing buildings the building key is often cause of conflict, which is good because OSM data is often better categorised that GRB. GRB only makes 2 major distinctions, this one and bijgebouw. Toolset also checks if the building has address data or not to decide over the type of building.
bijgebouw secondary/adjacent building. Never has associated address data. building=shed/hangar/... default: building=shed Watch out for existing hangars/greenhouse/garage. You need to resolve any conflict when merging with an existing OSM building.
gebouw afgezoomd met virtuele gevels Doesn't belong in OSM. This building is roughly estimated from street-side measurements without aerial data (so the depth of the building is always 5m). N/A This building is imported, not exported (since it shares nodes sometimes with other objects). Postprocessing will remove the way from Postgis DB that provides GRB vector layer for the toolkit.


  • Gvl : The building sides split up by measurement method used (measured from the street, from imagery, estimated, ...) can be useful to deduce the quality of the buildings, but not wanted in OSM.


  • Gvp : point of a building with annotated measurement method. Not useful for OSM.


  • Knw : Different man-made objects . Are imported and easily remapped due to detailed distinctions
Object description Mapping Import
overbrugging A bridge man_made=bridge Imported
waterbouwkundige constructie In the extract I used: bogus data Not wanted in OSM. Not imported
cultuurhistorisch monument Small elements, most likely chapels, but there's no subdivision. Not interesting now Skipped for import.
hoogspanningsmast / openbare TV-mast Tall construction sporting wires or antennas power=tower / man_made=mast Imported and exported: man_made=mast only. TV mast seems less accurate.
pijler ??? Skipped for import.
rooster ??? Skipped for import.
schoorsteen chimney man_made=chimney Imported and exported.
koeltoren Cooling tower man_made=tower + tower:type=cooling Imported and exported.
silo / opslagtank storage tank man_made=storage_tank or man_made=silo Imported and exported: man_made=storage_tank, man_made=silo
cabine Electrical, gas or other cabine N/A Skipped for now.
watertoren water tower man_made=water_tower Imported and exported: man_made=water_tower
tunnelmond tunnel exit probably doesn't belong in OSM Not imported in DB.
chemische installatie man_made=works Chemical installation, piping, long structures commonly found in petrochemical and heavy industries. Imported and exported: man_made=works
golfbreker (strandhoofd en lage havendam) Breakwater installation, a bit vague on subtypes. man_made=groyne (perhaps also man_made=breakwater)| Not imported (remapping a bit unclear)
havendam Another type of breakwater installation. man_made=breakwater Not imported (remapping a bit unclear)
staketsel man_made=pier N/A Not imported


  • Lbz : zones used for internal management of GRB, no use in OSM


  • Sbn : railway area
Object description Mapping Import
railway area The bedding (or street) where rails are located. (split per type of usage: train, tram and metro) landuse=? Not imported.


  • Trn : public terrain that doesn't fall under highway area (mostly unregulated parking space). Very sparse data, so not useful for an import. Import Toolset defaults: Not imported.


  • Wbn : highway area landuse=* every crossing is a different area, so it uses the "plumbing" principle, but it overlaps other landuses where separation isn't clear (in rural areas). Slightly over-noded, possibly not suited for import. Import Toolset defaults: Not imported.


  • Wga : highway-related feature
Object description Mapping Import
bushok shelter at a bus stop amenity=shelter + PT tags Skipped for import.
telefooncabine Public telephone cabin. There should be none left in Flanders as the last one was removed on Oct 1 2015. N/A Not imported
overdekte fietsstalling Covered bycicle parking amenity=bicycle_parking + covered=yes Skipped for import.
bergplaats storage shed (not really compairable with OSM shed) no idea how to map Not imported


  • Wgo : highway border per usage type (no areas, but the borders can be combined to areas) Import Toolset defaults: Skipped for now
    • grens circulatiezone zwakke weggebruiker (Wcz) : boundary of the pedestrian area (not every way has this, cannot always be close). Import Toolset defaults: Skipped for now
    • rand van de rijbaan (Wrb) : boundary of the motorised area (most inner boundary, can always be closed). Import Toolset defaults: Skipped for now
    • grens onverharde zone (Woz) : boundary of the paved zone (cannot be closed, is cut at every place where a private paved area touches a highway - like a service way or path), probably not suited for import. Import Toolset defaults: Skipped for now


  • Wgr : ditch waterway=drain (not oriented in a meaningful way). Import Toolset defaults: Skipped for now. But it can be used to draw manually. In JOSM validation will actually check the waterflow and warn about direction so you can fix the orientation problem. But it needs to be connected to other waterways. But it is easy addition when editing.


  • Wkn : way knode (endpoint of a way), only one case is useful. Import Toolset defaults: Skipped for now


  • Wni : lenghtwise highway element. Import Toolset defaults: Skipped for now
    • verhoogde boord- of kantsteen : kerb=yes (looks like this data isn't very consistent, so perhaps not suited for import)
    • muur, stootband : small wall, enough for barrier=wall ?
    • vangrail : guard rail barrier=guard_rail
    • niet-afgeboorde verhoging : rest-category, not mappable


  • Wpi : pointshaped highway element. Import Toolset defaults: Skipped for now but good candidate to add when test phase show good results. Seems quite accurate and detailed in cities area's.


  • Wri : manhole covers manhole=* ? Import Toolset defaults: Skipped for now


  • Wrl : railway when crossing the highway: too restricted for OSM. Import Toolset defaults: Not imported.


  • Wti : transversal highway feature (lower edges of a traffic_calming=table/bump/hump or upper edge of a road lowering (inversed table)). Import Toolset defaults: Skipped for now


  • Wtz : water area natural=water, these are tagged a lot wider than the average upper water level, and often include the grassy riverbanks. So perhaps not suited for import due to differing

definitions. Import Toolset defaults: Not imported.


  • Wvb : highway=motorway/trunk/ ... /service/track Follows the center line, and contains the following info about the road. Import Toolset defaults: Skipped for now.
    • Left and right name (equal in most cases) : name=*, name:left=* and name:right=* (left and right aren't needed in OSM when they are equal)
    • National reference ("ng" means no reference) : ref=*
    • Left and right municipality name (equal in most cases, unequal on boundaries) : not needed in OSM
    • surface (paved or unpaved): surface=* (though more difersity is sometimes wanted in OSM)
    • road class, choice among:


  • Wlas : streams and rivers (everything wider than 3m) with name and name of watershed. Import Toolset defaults: Skipped for now

Quality of GRB data

GRB building data has two quality levels: measured and drawn from aerial imagery.

  • The measured data is always very precise (more precise than anything we can achieve), but may be outdated a few months.
  • The data derived from aerial imagery is about as good as we can achieve in OSM with mapping from correctly aligned pictures.

In general, everything that can be measured from public roads is also measured. So the front sides of the buildings are measured, the back sides aren't.

Data recency

The data is continuously updated by the government. Every change on the terrain that requires permits, is communicated to the GRB team. Once or twice a year, surveyors visit all places that have changed. Everyone can make an "anomaly" to communicate an error in the data. Often, building data will be updated even before it is visible in yearly updated aerial photography, though not always.

Overnoding problem

The GRB buildings data seems to have around 10% useless nodes. This mostly comes from their measurement process. If they can't measure a back corner of a house, they measure a point in the side wall. That measurement point is then used to map the rest from imagery. But that measurement point is always on a straight line, so useless to OSM.


The Central Reference Address File (Centraal ReferentieAdressenBestand) is AIV's database with addresses and their positions. The import procedure for this database was started but it has stalled now that we're planning to import the building outlines. The CRAB's addresses are copied to GRB almost manually, so CRAB is more up-to-date and correct. Importers use CRAB to do a check-up round after GRB's buildings are added.


The Agency for Information Flanders (Agentschap Informatie Vlaanderen or Informatie Vlaanderen) is the Flemish governmental agency tasked with producing geo information and digital data (official website).

AGIV (Agency for Geographic Information Flanders, Agentschap voor Geografische Informatie Vlaanderen) is the name of the former agency tasked with geo data. It merged with other entities to form AIV. The name is deprecated but is still in wide use, even on portals of the agency itself and in data attribution requirements.


UrbIS is the Urban Information System of the Brussels-Capital Region region. Buildings from there are imported independently from this import.