Comment by transpute
1 month ago
Sensing is (sadly) part of Wi-Fi 7. If you have a recent Intel, AMD or Qualcomm device from the past few years, it's likely physically capable of detecting human presence and/or activity (e.g. breathing rate). It can also be done with $20 ESP32 devices + OSS firmware and _possibly_ with compromised radio basebands.
Was anyone asking for their network to be able to sense their breathing rate? What does this enable that actually improves people’s lives?
This is the kind of stuff that pushes me to pull a Ron Swanson and throw my technology in the dumpster.
The network already could. The standardisation is just making the feature available without hiding it.
The core of the sensing technology is about improving MU-MIMO + OFDM + all the other speed tricks. Human bodies interfere in predictable ways so you need the tech to steer around that. As a side effect, you get detection capabilities for free.
In such a setup, your laptop and router already know where you are. The question is whether or not to offer it to you so you can use that information for things like home automation. Had they not made this part of the protocol, the privacy risks were just as bad, you just wouldn't be aware of them.
Similar technology has been quietly in use for a while, with falling cost, e.g. "Inside a $1 radar motion sensor", https://news.ycombinator.com/item?id=40834349 (100 comments).
Commercialization gives consumers and regulators the opportunity to express their opinions on the sudden and unsolicited transparency of the walls, floors and ceilings of their homes and businesses.
I tried Wifi7 at my home, but most of the benefits are lost when physical walls are in the way. Therefore I think WiFi 7 is more for commercial applications.
TSA can check your heart rate / breathing rate elevating during your walk through security.
Casinos can see your heart spike before placing a bet. If the system is digital maybe that can be synced to always deal a loss hand.
The only use case I've heard of is elderly care, where no movement might mean a person has fallen and needs help. An edge, strictly opt-in scenario that would be addressed more effectively (movement+HR+body temp) by relatively cheap wearables.
Commercial use of WiFi sensing predates WiFi 7 (a notable example is Philips smart bulbs with presence detection). AFAIK WiFi 7 just includes an amendment by the 802.11bf working group to improve performance.
What's the commercial use of having this data though? Or even law enforcement use? We all have our phones on us most of the time anyways, knowing where in my house I'm at doesn't really... change anything...
There are 1000+ public research papers on machine learning + RF detection of human activity, including but not limited to breathing rate, keystrokes, body position, body motion, gestures, sleeping, biometric (identity) signals and more, https://scholar.google.com/scholar?q=device+free+wireless+se...
What's the economic value of remote collection of human behavioral signatures without consent, integrated with AI and robotics and "digital twins"? We're not there yet, but if the technology continues improving, what's the future value of "motion capture" of humans without body-worn sensors?
In theory, this will enable "Minority Report" user interfaces. 3D gestures could be combined with "AI" voice interfaces. Biometric authentication (e.g. heart rate) could replace passwords. Walk into a room and it adapts itself to your preferences. Etc.
There are lots of "cool" Jetsons sci-fi use cases, but ONLY IF the data and automation are entirely under control of the human subjects, e.g. self-hosted home server, local GPUs, local LLM, local voice recognition, etc.
[flagged]
If you had a particular idea from the LLM that you wanted to share people would be more receptive, but just dumping the whole output comes across as intellectually lazy
Please don't do this. Whether it's LLM-generated or not, we don't want big blocks of text from elsewhere pasted into comments here. Please at least try to craft original human thoughts.