One of the things we did was a survey of N=226 AirBnB users. 86 people said that they manually searched for devices in AirBnBs, and 8 actually found undisclosed devices. As you might imagine, manually searching is hard because devices are small, but also because you don't always know if you've found everything, and because it looks socially awkward too.
People also reported on how they looked for devices, ranging from manual search (most common) to using WiFi / Bluetooth scanning tools. Here is an excerpt from our paper as to what people said they did.
11 (13%) tech-savvy participants used special tools to help search, such as using a phone camera or flashlight, scanning Bluetooth and the Wi-Fi network, and using an infrared scanner. For example, participants reported that they turned off all the lights and used their smartphone flashlight to shine around the area and look for reflections that might indicate a camera lens. Some participants reported that their smartphone cameras did not have an infrared filter, and so they could detect invisible infrared light emitted by cameras. Other participants reported using smartphone apps, such as Fing and iNet, to detect suspicious devices connecting to the network. Two participants brought RF signal detectors to scan the room for the source of RF signals (cameras). However, these methods require a strong technical background, and the devices might not be cheap (e.g., over $100 USD on Amazon for a popular RF signal detector).
The main point of the paper, though, is to look at ways for IoT device manufacturers to help physically locate devices. We called these IoT Locators. An example might be to have devices blink or beep. It's trying to address the physical layer of privacy, to help people have more awareness of what's nearby and where these devices are. The logical layer might tell you that there's a camera nearby, but we also wanted to help people find those devices and see (for example) where those cameras might be pointed. There's a big difference between a camera in an AirBnb that is pointed toward the living room versus pointed out the window, something that isn't captured just by knowing that a camera is nearby.
There's lots of limitations, for example needing lots of standards among device manufacturers and how to incentivize them to install things like these in devices. I would say the main idea, though, is that we need to really think more about the physical dimensions of privacy (which has been underlooked in research and industry), and that we need to think more about how to help hotels, AirBnBs, and device manufacturers that do want to do better on privacy w.r.t devices, but don't know what to do.