Sharing Experiences OSM Mapping in Four Countries: Armenia, Bangladesh, Fiji and the Philippines

Posted by:
RebeccaFirth
Date posted:
Aug, 23 2017

*Guest Blog: Colleen Curran GeoInformatics Center, Asian Institute of Technology (AIT)*

Throughout 2016 the GeoInformatics Center, a research center at the Asian Institute of Technology (AIT) has organized OSM remote and field mapping activities in four countries as part of a larger Asian Development Bank funded project to increase disaster resilience through the use of ICT and space-based  technology, with a focus on mobile phone applications for local government officials and community members. A lot was learned through the process, the following are highlights of the Project. 

Remote Mapping

Armenia (Pilot area - Dilijan)

Armenia’s OSM community is only in its nascent stages. Building coverage in the pilot area was good, but the imagery was slightly out of date. As a result drone imagery was acquired, tiled and added as a custom layer on OSM for mapping. The imagery was captured for a vulnerable area in Dilijan where there is a public school. Using the OSM data an evacuation plan and map was made for the school.

Image 1. Tiled drone imagery as a custom layer on OSM 

Bangladesh (Pilot areas - Barguna and Sirajganj: Hut Boyra and KawaKhola Village) 

The areas chosen as pilot sites are very remote and did not exist on OSM prior to 2016. Mapathons were organized in Dhaka, and at the local level with assistance from OSM Bangladesh. This was the first opportunity for the villagers to view such a detailed map of their communities; they enjoyed participating in mapping as well as viewing the updated maps. Black-outs, weak internet and lack of computer resources made local mapathons a challenge. Despite this dedicated volunteers added over 5,000 buildings to the OSM map in these pilot areas.   

Image 2. Local volunteers use geoMapTool for field mapping in the remote pilot area of Hut Boyra, Sirajganj

​Fiji (Pilot area: Nadi)

Nadi, Fiji, an international tourist destination was decided as the pilot site.  In the selected pilot areas, over 3,000 buildings were mapped by volunteers.  Unconventional mapping methods and unclear changeset comments created issues and caused a zero hour freeze on the accounts of some of the volunteers.  Issues were resolved with support from the OSM data working group. The mapping was accomplished without the use of HOT Tasking Manager, in the future the Tasking Manager will help to make the mapping exercise more organized.  It has also been recommended that Fiji develop an OSM wiki page for future mapping activities.

 Image 3. ​Nadi town staff and local volunteer perform field mapping in Nawaka Village, Nadi, Fiji

The Philippines (Pilot areas - Santa Josefa and Padre Burgos)

There is a strong and very supportive OSM team in the Philippines who participated in workshops and testing of tools and provided useful insights. A mapathon was held Manila and at the municipalities using HOT Tasking Manager. High-resolution images were procured, tiled and added to OSM as a custom layer for mapping. In the pilot areas 4,000 buildings were digitized.  

Image 4. [top] Acquired high resolution imagery tiled on OSM [bottom] default imagery on OSM, Padre Burgos, Philippines​

Field Mapping

Once remote mapping was completed, a mobile application called geoMapTool was introduced so trained volunteers could add attributes to the buildings. geoMapTool allows users to collect building attributes in offline mode and directly upload the information to the OSM server.  While useful, we faced a few challenges with this tool:

  1. Some of the attributes collected were not standard OSM values
  2. A feature in the mobile phone application to digitize in the field caused polygon stacking and irregular shapes of the buildings.

The OSM data working group flagged both these issues and worked with the volunteers and developer to resolve them.   

Learnings from the Field

  • We should have created a wiki page to make the OSM community aware of the project and facilitate their feedbacks on the methodology for field mapping;

  • More time should have been dedicated to Fiji mapping as there is no local OSM community to support the mappers directly. A training of trainers could of have prevented some of the mapping issues;

  • The international OSM community should have been engaged and consulted during OSM field mapping tool development;

  • OSM mapping at the local level requires additional basic trainings for communities who are not familiar with mobile phone applications and computers and do not have an email address

  • Armenian government is worried that OSM will affect national security, doesn’t fully trust crowd-sourced data

  • Bangladesh government see this as a great resource and is interested in expanding the project

The following resources were very useful in this project:

Please see our video introducing this project https://www.youtube.com/watch?v=vSN2tWZP2vE

Visit the Project website here http://www.geoinfo.ait.ac.th/resilience/

 

For more information please contact GeoInformatics Center, geoinfo@ait.ac.th.  This article was written by Colleen Curran, Program Specialist at the GeoInformatics Center, AIT.