CR research collaborators share how we study safety, privacy, and consumer protections in IoT devices
The sky’s the limit for the kinds of consumer electronics that can be connected to the internet; these devices use information from real-world environments and turn them into insights that help manage myriad aspects of your daily life. With these benefits, however, come potential costs – especially to your privacy and autonomy. As devices can range from smart TVs and fridges with touchscreens and cameras, to smart doorbells and streaming devices, how do researchers design experiments to learn more about how these device experiences might harm us?
One group of researchers (that’s us, at Northeastern!) collaborating with Consumer Reports took a manual approach, interacting with devices in a simulated studio apartment by hand. In an iterative, time-intensive process, we inspected IoT devices one at a time, exploring features available in companion apps or on the device itself. This helped our team get a baseline understanding of our devices’ experiences and design an interaction ‘script’ that would let us perform the same tasks in each device. From the first iteration, we also learned more about the vast differences between modality-based features across devices; to better standardize our work, we devised a voice-controlled set of questions to ask smart speakers in our lab environment. With adjustments made along the way, we finally built a process for interacting, recording, and labeling device experiences and conducted a final round for data collection. This method of data collection takes dozens of hours and can be tedious at times, but the effort is worth it to gain first-hand experience with all of the devices and their diversity. Interacting with the devices ourselves helps us understand what consumers see when setting up their home IoT products, creating accounts, and using them to improve their home life.
Emerging technologies bring consumers new benefits as well as problems. To better understand these user experiences, researchers in consumer protections and privacy harms will need to continue iterating on prior methods, rolling up their sleeves, and taking first-hand looks at these technologies as complements to automated empirical work. Read the full article here.