Privacy dashboard add-on study

[past cedar]

If you’re interested in helping our research and better understanding the behavior of devices on your home network, please read on!

We are Carnegie Mellon University researchers conducting a study about a new tool that tells you about the smart home devices in your home and the data they collect and share, including the network connections your devices are making and info from the companies’ privacy policies.

In this study, you will install a Home Assistant add-on, use the add-on for 5-10 days, and then participate in a 60 minute interview with the researchers. The study is expected to take 1.5-2 hours of your time total, including installing the tool, using it, and participating in the interview. You will receive a $50 gift card to your choice of Amazon, Target, or Walmart for completing the study.

Participation in this study is limited to individuals located in the United States ages 18 and older. Participants are also expected to be able to read and write in English fluently, and have Home Assistant installed with at least two types of smart-home devices.

If you think you could be interested in participating, please click here to find out more and fill out the screening survey.

More details about the study and data collected are available here. We will invite a selection of eligible participants with a variety of smart home configurations to install the tool and participate in an interview.

Note: this post has been approved by @Missy

[warm willow]

Thats great! Sadly I dont make the cut for the demographic requirements, but I like the add-on. I am a Product person here at Open Home Foundation (Home Assistant and others) and we are building a Device Database. Being able to expose this information could be really useful for our users too. @shell spruce ping to have this on our radar

[past cedar]

Yeah unfortunately we are currently limited to US-only recruitment for logistical/bureaucratic reasons… but glad that it sounds interesting! The network connection data (what you can see in the screenshot) is captured from the devices by intercepting their traffic using ARP spoofing. We also have some data about the companies' privacy polices for some of the more popular manufacturers that we gathered using a combination of manual work and LLMs.

What info are you planning on having in the device database?

[shadow pond]

Very interested and fit the bill. However, I have questions. Is the add on sending all this data to you? How secure is it? Also, when you say interview are we talking recorded video?

[past cedar]

The network data in the screenshot is stored on device only and not sent to us, and also we only monitor network traffic for the IoT devices you choose to enable this feature with. We do collect a list of device names on the network ("Living Room LG TV" or possibly "Ben's phone") so we can validate how accurate our detection/classification is (sent to a self-hosted server).

For the interview, you don't need to share video. We record audio so we can analyze the transcripts, and also will ask you to share your screen as well which would be in the recording (and may include data like recent network connections if that is shown on screen).

The consent form and screening survey has some of these details, I also made a more detailed doc here with some of the specific data points. If there's anything specifically you are concerned about, let me know and we can see what we can work out (edit: fixed link)