As a new parent who did all kinds of research on SIDs and wearables this is awesome. I’d love to have a product like this to put in baby’s room to detect breathing. It could also make an amazing device for analyzing sleep quality for adults as well.

Sadly, this is likely just going to be applied to advertising.

The ResMed S+ is a commercial product with these capabilities (Disclaimer: I was on the team that built the data collection backend in Azure)

The algorithm team was working on an advanced version that could be used to monitor seniors and other immobile patients throughout the day, but they had to refocus on another product. One of the algorithm guys told me that they could get heart rate with the sensor, but not cleanly enough that it could be a medical product.

The article is about a research group in Mass, but you might be interested in keeping an eye on Advanced Telesensors. They are a traditional startup in Austin that does exactly what you want and their sensors work well to boot.

http://www.advancedtelesensors.com/

Current Sleep Number beds can track breathing and heart rate through the air mattress. It is reasonably accurate. Their app lets you review restfulness levels over time. The mattress will also self adjust if you switch to side sleeping.

Can it alert on sleep apnea episodes?

You can use Eulerian image magnification for that. Just needs a normal camera: http://people.csail.mit.edu/mrub/evm/

Yes very sad - this would be a great idea if we wouldn't have to live in a f*ed up world like this.

Surveillance and Advertising are the #1 sectors to use this for.

Realtime monitoring of breathing and heart rate during A/B tests sounds like an adtech engineer's wet dream.

"Surveillance" is exactly what the GP wants to use it for.

I think most would agree that ‘surveillance’ for increasing odds of healthy outcomes is significantly different than surveillance used to make a buck.

Would you object to blood pressure and heart surveillance if you were in the hospital?

It's still surveillance, which is the point. Making a broad statement about surveillance is ridiculous, particularly when it is made in the same breath as a statement about advertising (which is a much less nuanced situation).

Here’s a sensor that’s already used by the local police to monitor the breathing of those locked up over night - to be able to respond asap if someone stops breathing after excessive drinking or drugs. Seems benevolent to me. https://www.xethru.com/x4m200-respiration-sensor.html

I think we're going to have to learn to live with some kind of either:

A. Radical transparency, where everyone has access to everything about everyone else and privacy exists as "rice paper" privacy[1]; or

B. Techno-totalitarianism with people divided into "Morlocks" and "Eloi", technocrats and techno-serfs.

As technology progresses the amount of available data grows geometrically, and I don't think there's any realistic way to put that genie back in the bottle (without postulating some sort of massive civil/ecological collapse), so I don't think the argument about "privacy" makes sense. The FANG companies et. al. are already in a god-like position in re: the hoi polloi.

I think we should assume "total information awareness" and then try to build a worthwhile civilization on that basis.

(FWIW, I hope for A but it seems like B is pretty much inevitable at this point. YMMV)

[1] In Japan walls were made of literal rice paper stretched on frames, and as an obvious corollary people generally knew their neighbors' business. "Rice paper privacy" refers to social conventions to, er, "pretend" that there was actual privacy.

> [1] In Japan walls were made of literal rice paper stretched on frames, and as an obvious corollary people generally knew their neighbors' business. "Rice paper privacy" refers to social conventions to, er, "pretend" that there was actual privacy.

Where can I read more about this?

Dunno, sorry, I heard about it from a Japanese person.

I did a lot of research into the state of Ultra Wideband (UWB) during the early 00s. It utilized time of flight and a wide band of frequencies to do similar things. There were 2-3 major, well funded companies pursuing the tech at the time.

Interestingly those companies never made it commercially, I suppose because the cost of computation was prohibitive and they required special antenna.

It is very exciting that the technology can now be done in a cost effective manner.

Most interesting though is how easily and effective ML was in adding functionality to this implementation. Very cool.

No sure how these guys (https://www.xethru.com/) are doing financially, but they sell development kits for respiration sensors etc.

Been following these guys for a while. Fascinating product.

To me though the most obvious but sadly simple use for this is sensor lights in damn toilets. Sit on the toilet for more than a few minutes in some places and the damn lights go out!

Figure there's malicious wi-fi capable of doing the same, since it's possible to read lips and detect keystrokes via passive snooping of wi-fi reflections.

Thie is nice,but laws that clearly define proper use of such tech are missing.