Written by: Rebecca Bicksler
Open Data Kit (ODK) is a free open-source software that can be used for collecting and managing data. Participants of the UR Field Lab used ODK along with KoboToolbox to aid in data collection for a flood vulnerability study in the Nong Hoi Neighborhood of Chiang Mai, Thailand. The data collected with ODK included:
• Occupancy type
• Construction material
• Lot elevation
• Access to street drainage
• Number of building floors
• Building elevation (height of the building with respect to the property)
• Ground cover
The survey was created to be as simple as possible, with check one or check multiple options (Figure 1). There was also an “other” option, with the ability to swipe and enter specific information. Overall, the ODK software was easy to create and to use, and it met our needs for data collection.
Figure 1. Screenshot of ODK on an Android smartphone
Figure 2. Building Footprints with Unique Identifiers
Figure 3. Comparing Building Footprints and Aerial Photos
Pre-Survey
The only setback in our process was that we were not able to use the geolocation function. As we first set up the survey, we expected to use geolocation to mark each property; however, the geolocation from cell phones was not accurate enough to pin the location for each property in such a dense neighbourhood. Instead, we had to create a GIS layer of the building footprints and create original ID numbers for each building (Figure 2). Then we took printed maps (Figure 3) to the field so that we could enter each ID number into the survey. This was a labor-intensive workaround that proved to be a bit difficult to manage as the building footprints were not completely accurate (though it was still more accurate than using the geolocation function). Had there been more time, we would have first surveyed the neighborhood to ensure accuracy of the footprints before starting our survey. If lot lines were available for the city, we would have used unique IDs for each lot, rather than each building footprint so that we could include information about the elevation and ground cover of open space and empty lots in addition to structures. Ideally, the geolocation would be accurate enough that paper maps and ID numbers would not be necessary and the work time significantly would be significantly shorter.
Figure 4. Field survey Day-1