Dhaka Building Import

Dhaka Buildings in Wards 67, 68, 69

About

This import is to add building footprints for the old city of Dhaka in preparation for the Open Cities Project surveying which will begin in August 2013. The initial import will include all structures from Wards 67, 68, and 69 in Dhaka, and possibly extend to include 63,64, and 66 as well. The data is provided by CEGIS, who has extended explicit permission to import it into OpenStreetMap. The ensuing data collection will add primarily structural data collection for each building in these wards, and all data will be entered into OSM.

Note that we are still awaiting final written approval for CEGIS, but expect to have it by August 16, 2013.

UPDATE (August 19, 2013): Written Approval has been given by CEGIS, and I have completed the building import. Please excuse some overlapping buildings and roads for the moment. I attempted to clean it up as much as possible but it will not be possible to assess some areas until the field surveys begin over the next two months. Exciting times!

Import Plan Outline

The import plan is:

1. Make sure there is no disapproval from local community
   1. Notifying users at imports@openstreetmap.org and talk-bd@openstreetmap.org
2. Convert CEGIS building shapefiles into OSM XML
3. Map attributes to OSM tags
4. Divide area into small chunks and upload manually through JOSM, ensuring accurate topology and quality

Goals

The goal is to get the building footprints into OSM so that mappers can easily collect attributes without the need for digitizing first.

Schedule

• Saturday, August 10, 2013: Notify community
• Wednesday, August 14, 2013: If no objections, import wards 67, 68, 69

Import Data

The data is wholly owned by CEGIS in Dhaka, who have provided written permission for importing into OSM. I will also attempt to import ward boundaries, also provided by CEGIS.

OSM Data Files

not yet created

Import Type

This is a one or multi-time import, depending on the program expanding and CEGIS providing permission for a larger set of data.

The data will be imported manually through JOSM by a single user (me).

Data Preparation

Will most likely use ogr2osm or shp2osm to convert the data, with verification in JOSM.

Data Reduction & Simplification

Will manually check for data overlap and issues in importing the data.

Tagging Plans

The building data contains the following attributes which will be mapped to OSM:

• STR_FLOOR (number of floors)
• STR_NAME (name of building if any)
• STR_TYPE (Kancha, Pucca, Semi-Pucca, or Tin-Shed - not sure what these mean)
• STR_USE (whether its residential, commercial, and so forth.

These attributes will be mapped to OSM tags, as such:

• building:levels
• name
• (not sure yet)
• building:use

And each structure will have a building=yes tag.

Team Approach

Doing this import solo in the first stage.

Changeset Size Policy

Will attempt to keep each changeset to < 1000 buildings (there are about 12000 buildings in these three wards)

Conflation

Will Manually conflate data in JOSM.

Quality Assurance

Will check data in JOSM to ensure data integrity, and once conflated with existing data will utilize the validation tool to assess any other errors.