x. Community Mapping for Exposure in Indonesia


Interaction Type Government→Public→Government
Trigger event
Domain A priori disaster response.
Organisation Community Mapping for Exposure in Indonesia.
Actors Indonesian Disaster Management Agency (BNBP), Disaster Management Innovation (DMInnovation), Humanitarian OpenStreetMap Team (HOT), The Australian Community Development and Civil Society Strengthening Scheme (ACCESS), GFDRR, Pacific Disaster Center (PDC)
Data sets Satellite imagery, GPS tracks and attribute data.
Process Collecting Spatial and attribute data and tracing in OSM platform.
  • OpenStreetMap layer showing an up-to-date basemap of vulnerable areas
  • Thematic maps showing damage in case of various physical disasters
  • A standardised training curriculum for capacity building purposes
Goal Reduction of vulnerability to natural disasters.
Side effects Deemed a successful example of disaster relief preparedness that could be applied in other developing countries.
Contact point Kate Chapman, HOT, (Phase I and II)

Yantisa Akhadi, HOT, (Phase III and IV)

An example of an a priori disaster response, this Indonesian mapping project began in early 2011 and at the time of writing is still active. The main idea behind the project was to use OSM to collect previously unavailable data about buildings and their structure in both urban and agricultural environments and to use appropriate models to calculate likely damage in case of physical disaster. The combination of the impact models and the use of realistic data led to the development of an open source risk modelling software (InaSAFE) showing the affected people, infrastructure and damage if disaster were to hit a specific area. This offers a practical tool for governments to develop actionable contingency plans and fills the need for risk assessments identified by the World Bank.

Phase I: Pilot Year (Early 2011 – 2012)

The pilot phase consisted of workshops offering training on the project and building construction as well as data collection in urban and rural areas. The approach between rural and urban areas was slightly different, although the result was similar. The original data were derived from paper maps, which were edited by local people; satellite imagery, depending on availability; and GPS tracks. Data were edited using JOSM and Potlach2 web editor and then used in QGIS. Urban areas were mapped by students who took part in a mapping competition. Rural areas were mapped with ACCESS contributors and local people. The second phase lasted from July 2012 to March 2013 and focused on collecting exposure information essential for impact modeling software.

In total, 163,912 buildings were mapped during the pilot, 29,230 of them in urban areas; and 16 workshops were held with 124 people participating in rural areas and 5 universities in urban areas.

To encourage participation, Community Mapping for Exposure has a pyramid format based on leadership, with specific guidelines in data manipulation and great coordination between different contributors. The whole process is focused on workshops; participants were supervised at many stages and the procedure of data collection and manipulation was firmly defined. Motivations for participation varied, with incentives covering a spectrum from disaster protection to a mapping competition. In terms of technical support, the project was not only supported by HOT and OSM but also by open source software such as QGIS. The main innovation in data collection was the private datastore, which offered a unique ID for each object. The final output has also been a success in enabling local government to visualize where people are most in danger by combining local wisdom with scientific knowledge to produce realistic scenarios for numerous different physical disasters.

The main aspect of concern is the quality of the results, which showed great variation. According to the final report, the quality was either very bad or very good in different areas, although it was found to be acceptable generally. The attribute quality, which has a principal role in the success of the project, indicated a great number of empty or incorrect records concerning the structure of buildings. Other minor deficiencies were also noticed, such as failing to create constant mapping volunteers and the use of time-consuming technical methods in a few cases (e.g. Excel spreadsheets in data collection or manual methods of data manipulation).

Phase II: Scenario Development for Contingency Planning (August 2012 – December 2013)

The main idea of this project was bringing the OSM to fill in data gaps and QGIS and InaSAFE for spatial analysis under a curriculum program which comprises a series of training: beginner, intermediate, and training of trainers (ToT). SD4CP has three core activities:

  1. Training disaster managers, NGOs, and universities in six Indonesian priority provinces to use these tools and through this training disaster managers have begun to be able to utilise the new tools in their contingency planning process.
  2. Developing a training curriculum which comprises of OSM, QGIS, and InaSAFE materials. The final training curriculum was developed, tested, and standardised by following the Indonesian Disaster Management Agency (BNPB) requirement.
  3. Providing ToT workshop to chosen individuals from the six provinces where the
    beginner and intermediate training has previously been provided. These individuals will serve as local support and leaders for the SD4CP program in their provinces.

According to the final report, OSM reached 1.4 million buildings in Indonesia; and 18 workshops were held with more than 350 participants. Additionally, a robust training team of 9 was developed.

Phase III: Supporting the Expansion of InaSAFE and OpenStreetMap in Indonesia (December 2013 – August 2016)

Based on recommendations from previous phases, HOT in coordination with AIFDR proposes strategies to further expand the use of InaSAFE and OSM in Indonesia. The proposed activities are focused on harmonizing training materials with the needs of disaster management agencies while keeping it up to date with the software version. HOT continue to support local disaster management agencies in multiple provinces for the use of OSM, QGIS and InaSAFE in the contingency planning development. Additional training materials on data validation are developed to improve OSM data quality in Indonesia.

As identified in the previous phases, there are success stories through engagement between HOT and universities. To replicate these success stories, HOT developed OSM University program, targeted at universities in disaster prone areas. The main activity of this program is University roadshow, which are consists of three main events: one socialisation and two trainings, each will last for one to three days, making up a total of five to seven days activities with several weeks of tasking in the middle of the two trainings. The tasking was used as a way to certify university students and later these certified students will provide support to local disaster management agencies. In total, there are more than 51 university students from 8 universities that completed the certification process.

This phase introduced and promoted the use of mobile data collection using smartphone to reduce the use of pen and paper during survey. GeoDataCollect, the in-house mobile application, have been installed in more than 5,000 devices and used by Jakarta Disaster Management agency as their official app for disaster reporting.

By the end of this phase, more than 4,2 millions buildings throughout Indonesia have been mapped in OSM.

Phase IV: Disaster Management Innovation through Open Geospatial Data (September 2016 – April 2018)

On phase IV, the main focus is on transferring the knowledge from HOT to BNPB. This would enables BNPB to have its own capacity to map disaster prone areas, train local partners, and building collaboration with local universities. This sets of skills are important as BNPB already listed 136 priority districts/cities having high risk of disaster and they would like to map these areas before disaster strikes. HOT also working closely with DMI to develop guidelines on how to use OSM, QGIS and InaSAFE for all phases of disaster lifecycle, such as for risk assesment and situational awareness. This would be relevant to the goal of institutionalizing InaSAFE in the BNPB.

The second focus of this phase is to build and maintain links between people and institutions in disaster management. The strategy is to leverage existing partnership with universities and bridge the connection with BNPB. HOT also expected to start building links with underrepresented groups, such as people with disabilities.

As part of its sustainability strategy, HOT will support the establishment of local organization in Indonesia to institutionalize the acquired skills, knowledge and experience in capacity development.

Main lessons:

  • The collaboration of multiple stakeholders in disaster management is important to address the challenges in disaster management.
  • Capacity development is essential to be able to take full advantage of technical innovations in the area of disaster management.
  • University is a strategic partner to collaborate with local disaster management agency as well as retaining knowledge and skills on mapping and data collection
  • An a priori disaster response can be focused on appropriate models and parameters and can calculate damage in case of a physical disaster by using OSM.
  • Interaction between official providers and OSM is a parameter of success not only for the beginning of the project but also for its continuity.
  • Open source data can be reliable for scenario building but its quality can vary, especially in terms of attribute data.
  • Risk managers, local communities and the public can combine local wisdom with scientific knowledge to produce realistic scenarios for numerous different physical disasters that may occur in an area of interest.