Thoughts on ‘Volunteered Drone Imagery…” (Johnon et al.)

I thought this paper was short and sweet summary of the current state of UAV/UAS acquisition tools and data processing softwares. They used OSM as a a parallel, vector-based example of what the future platform could be for aerial data and it was helpful to have some schema about this topic to build from.

One issue that they did not touch upon which immediately springs to mind when considering a database of “frequently updated, high resolution imagery” (pg. 1) is that of privacy. If they are referring to real time information about habited environments, then having an exceedingly easy way to obtain high-resolution aerial imagery comes with all kinds of implications for protecting individuals privacy. Would they blur out humans and sensitive information like license plates? At which stage would this image manipulation occur, who would be responsible for it? Even if the images are not granular enough to allow identification, there have been nefarious uses of geographic data before (like the people who used PokemonGo! data to target spaces known to contain other users and mug people. Especially since the ultimate aim seems to be for this data to be easily accessed/manipulated into third party products/services, it would be difficult (or impossible) to “opt-out”.

The authors discuss how the private sector is investing in this industry to “reduce even further the entrance costs”(pg.1) to this field. I can see why companies would want to encourage recreational use o fUAVs as a  hobbie, because the associated paraphernalia and updates presents an opportunity for endless monetization. But as they note later in the paper, the specialized data processing softwares can be expensive and complicated. So it will be interesting to see how this balance between democratization of the hardware and usability of UAVs and the high-barrier of later-stage data manipulation changes with time, investment, and public interest.

The issue of interoperability was not discussed explicitly but touched on when mentioning how the large variety in quality of sensors means that it can be difficult to host imagery on a common site and stitch images from a given area together coherently. This reminded me of the interoperability issues mentioned in the article on cyberGIS and seems like a recurrent issue in discussions of GIScience and its applications.

The example of Nature Conservancy Coastal Resilience Project as a hosting service with a concrete agenda made me think about the importance of objectivity when compiling imagery or creating a data hosting platform. I would say OSM tries to be pretty objective in their collection and representation of data (although of course complete objectivity is impossible.) But I wonder if it is more valuable to explicitly state the objectives and goals of an aerial imagery project in the hopes of solving a particular problem, or addressing a particular gap in the data. That way, users who are interested in that particular issue are more likely to participate and provide better quality data. The general public could too, but their contributions might be stronger if in pursuit of a particular feature of the landscape, or to capture specific environmental indicators. If, instead of having one platform of uniform data, a few platforms with specialized guidelines, centralized organization, and stated objectives for specific projects would be a meaningful and pragmatic first step. After assessing the success of these pilot projects,  the UAS community could reflect on the necessity of a universal, high quality aerial imagery platform.


Comments are closed.