How Cheap Drones Helped Create Maps After the Nepal Quake

Community mapping. Working with the hard copy map. Photo: Patrick Meier

Following a series of severe earthquakes in 2015, many communities in Nepal struggled to rebuild their homes and  lives. Aerial imagery can be helpful when planning post-disaster recovery activities, however people on the ground found that the quality of available satellite imagery was too low to be useful, especially given how densely built up Kathmandu and other cities in Nepal are.

One of the local organizations that was looking into this issue was Kathmandu Living Labs. They were interested in getting frequently updated imagery at a higher spatial resolution without being dependent on an external company that may or may not provide the satellite imagery at a discount. At the same time, Kathmandu University has been experimenting with the use of drones for agricultural purposes and other tasks such as mapping of landslide.

A pilot project led by the humanitarian UAV network UAViators, in collaboration with Kathmandu University and Kathmandu Living Labs, looked at how comparatively affordable consumer drones could be used to achieve similar results to those of more expensive professional mapping drone. The target areas was the village of Panga, which had been severely damaged by the earthquake.

UAViators had formed a partnership with drone manufacturer DJI and mapping software company Pix4D. These donated Phantom 3 Advanced multi-copter drones, software and staff time to conduct training sessions in Kathmandu. The drones were used together with Android phones running the mobile app Pix4Dcapture, that can be used to create a flight plans. Mapping the area took 1.5 days of flying six drones for a resolution of 3.4 cm. The imagery was processed by Pix4D staff on laptops.

 

Permissions

Kathmandu University and the local Community Disaster Management Committee (CDMC) filed a joint application for flight permits and received the Civil Aviation Authority’s permission within a week. Since then, regulations have significantly tightened and obtaining all necessary permissions may now take a month or more.

 

Outputs

The ortho-mosaics were printed on simple roll-up posters that were then given to the CDMC of Panga who invited community members to annotate the map directly with additional relevant features such as the location of debris, safe drinking water and health facilities. “We had displayed that map on the street of Panga for two weeks to show local people and outsiders about present conditions.” Later it was used to show the location of temporary shelters and what percentage of displaced families lived in which section of town.

According to Patrick Meier, an internationally recognized expert in humanitarian technology, the hard copy maps catalyzed conversations between community members on the strategies and priorities vis-à-vis both the recovery efforts and the preparedness activities that were ongoing in the community.

Laying out the printed map of Panga. Photo: Patrick Meier
Laying out the printed map of Panga. Photo: Patrick Meier

 

Professional vs Consumer Drones

The drones used in this pilot project are not considered professional mapping drones, but consumer models that are much more affordable than the professional versions. However, using the flight planning app made it possible to acquire imagery that could be used to produce maps of sufficient quality, as the app controls both the flight path and triggers the camera where necessary.

The resulting ortho-mosaic of Panga has a resolution of 3.4 cm. The map was displayed in public to show the location of displaced families in the respective temporary housing clusters and the percentage of the displaced population living there. Photo: CDMC Panga
The resulting ortho-mosaic of Panga has a resolution of 3.4 cm. The map was displayed in public to show the location of displaced families in the respective temporary housing clusters and the percentage of the displaced population living there. Photo: CDMC Panga

 

Major issue: short flight time

The project created very high quality maps that could not have been made via traditional methods such as satellite imagery. Some limitations also emerged, most notably that the multi-copter drones had a limited battery life of only about 20 minutes and were thus not able to map large areas. In total, mapping the area of Panga took 1.5 days of flying and data processing.

Another unforeseen problem was radio wave interference between the device that controls the photo trigger on the ground and the receiver of the drone. As a result, flights needed to be shortened, since the drone had to stay within a short radius of the controlling device on the ground.

Both Kathmandu University and Kathmandu Living Labs will continue to use the drones that had been donated. However, Elizabeth Gilmour at Kathmandu Living Labs also noted that the short range of the drones is a limiting factor. “The big focus (for Kathmandu Living Labs) is on mapping. It is going to be hard to do a lot of mapping with these UAVs because they have a battery life of only 20 minutes.”

Kathmandu University is currently actively using the drones in a number of projects, such as  building damage assessments. The university also teaches 3rd year undergraduate Geomantic Engineering students in how to use the software.

 

Over the next months, Droneblog will feature summaries of case studies that show how drones are already being used in disaster response operations worldwide. The case studies were produced under the leadership of the Swiss Foundation for Mine Action (FSD) and with funding from EU Humanitarian Aid. The goal of this research initiative is to identify use cases in which cases drones can improve the quality or increase the efficiency of humanitarian aid.

You can find more information about this and other case studies on http://drones.fsd.ch/

LEAVE A REPLY

Please enter your comment!
Please enter your name here