Bringing drones to the people: Development of a low-cost fixed-wing UAV
and multispectral camera for custom application in earth science field
work, education, and outreach
Abstract
Incorporation of Uncrewed Aerial Vehicles (UAVs) has greatly enriched
earth science field data collection but the cost of acquiring UAVs,
particularly those with customizable payloads, can be high, thus
creating a barrier to using these valuable systems. We set out to design
and build a low-cost ‘student budget’ drone and multispectral camera
that could be customized for a variety of field, education, and outreach
applications. UAV development followed an iterative design process
guided by open-source plans resulting in a progressive series of 5
design iterations each incorporating lessons learned from previous
versions. The final aircraft incorporates a flight controller for
autonomous flight and a removable/customizable payload pod. Cheap
components allow for easy and inexpensive repairs when damaged in the
field. The final design cost is $625, but this is a max cost as a
simpler system could be built based on the design for <$300.
The multispectral camera was built with the popular Raspberry Pi 2
computer, standard and infrared cameras, and low-cost/well-characterized
filter material. The resulting multispectral camera collects imagery in
the visible and near-infrared spectrum, with a total cost of
~$230. The UAV and camera cost ~1/5 &
~1/10 that of commercial systems, respectively.
Unfortunately, the camera did not yield research-grade results due to
image inconsistencies. Despite the lower cost, there are additional
considerations when choosing a UAV and imaging platform including data
needs, data quality and repeatability, ease of data collection and
processing, required UAV pilot skill, and time investment for UAV
construction. Given the ease of use and minimal pilot training time,
commercial systems (e.g. DJI quadcopter) provide the best fit for many
research applications like aerial photography or 3D outcrop modeling.
However, for more complicated data needs (e.g., multispectral imaging)
and/or projects with small budgets, a low-cost UAV with a customizable
payload can open up new data collection avenues and scientific inquiry
that was previously unavailable. Additionally, this approach has
applications in STEM education to teach engineering processes,
aeronautics, remote sensing, etc., and is useful as an outreach tool to
educate general audiences about UAVs and their responsible use.