usable security

Drones on Sexton field

Students in the COS Independent Work Seminar entitled “These Aren’t The Drones You Are Looking For: Mitigating the Privacy and Security Implications of Drones” left the classroom to get out and about on Sexton field on a beautiful spring morning. The order of the day was to test out a host of Parrot AR Drones in an open flying environment.

Students also worked on testing out their semester-long individual projects involving actions such as controlling the drone with a keyboard, pinging a drone to get its identity, and tracking a drone using a lightweight Android phone mounted to the drone.

Check out pictures of the class outing here.

usable security

Drones, Privacy, and Security at CHI 2017

We describe a few highlights from our recent paper on drones that will be presented at the 2017 CHI conference.

What did we do?: To better inform policy and regulations around drones, we chose to investigate what users’ current perceptions are of privacy and security issues around drones. We conducted an experimental study with 20 laypersons in Maryland, USA with users who had never seen or interacted with a drone before.

How did we do it?: The experiment took place indoors in a laboratory at the Human Computer Interaction Laboratory at the University of Maryland, College Park. Each participant had a pre-interview asking them about their general perceptions and mental models of drones. Participants were then shown a real drone or model drone and given the opportunity to see the researchers control the drone and to fly the drone on their own. Following the demonstration and a series of tasks that included showing users the footage that drones collect and store, participants had an exit interview to discuss their final thoughts on drones, privacy, and security.

What did we find?: Compared to previous studies, 1) we found that users had many more negative perceptions of drones and how they affect their personal privacy and security. For instance, participants were worried drones could injure people and that they could be used for spying on others. Users were particularly concerned about having multiple drones in an area and keeping drones away from people and private spaces like schools, government institutions, or private residences. 2) We found that a drone’s design affected participants sense of privacy and security. For instance the sound of a drone made it seem threatening but at the same time, the noise alerted participants to the presence of the drone which was considered a privacy enhancer.

What are the implications of the work?: Our work suggested that 1) there needs to be better regulations to deal with cases where there are multiple drones in an area and to mandate geo-fencing technologies either in drones themselves, or to develop other geo-fencing mechanisms that can be implemented by those who might encounter drones using existing infrastructures such as home wireless networks. Geo-fencing could enhance privacy by keeping drones away from private spaces and making people feel more secure about being “safe” from drones in these spaces.

2) We suggested that further exploration is needed to see if it is possible to have drones operate in designated spaces or “drone highways” to keep them at a distance from people to protect privacy, be easily identifiable, and prevent injury. Finally, 3) we recommend that drone designers think about how to create designs that make drones more friendly or less so depending on whether or not people should interact with them using both visual and non-visual markers such as sound.

Read the CHI 2017 paper for more details and contact us for copies of supplementary materials such as interview guides.