> Imagine a bottle of laundry detergent that can sense when you’re running low on soap — and automatically connect to the internet to place an order for more.<p>Why. Why why why do all IoT-related articles always use awful consumer goods examples like this. Nobody, or next-to-nobody, would want to have that. There are so many good ideas and existing uses of IoT tech outside of the consumer goods sector, and pretty much every application <i>in</i> the consumer goods sector is hot garbage, and consumers know it.
"connect to WiFi" is not remotely accurate. It should be: "Reflect wifi signals in a predictable pattern." This technique, while impressive, requires custom hardware/software to detect and interpret (presumably via filtering and a fourier transform) the back-scatter signal.
Anyone that finds this interest should look at how “The Thing” (1945) worked:
<a href="https://en.wikipedia.org/wiki/The_Thing_(listening_device)" rel="nofollow">https://en.wikipedia.org/wiki/The_Thing_(listening_device)</a><p>EDIT: Here are some diagrams of how it works:<p><a href="https://hackaday.com/2015/12/08/theremins-bug/" rel="nofollow">https://hackaday.com/2015/12/08/theremins-bug/</a>
This is a truly awesome technology. But ITT an abundance of nit-pickiness on how this isn't <i>exactly</i> what the title says it is or how it could be presented better.<p>Yes, the nitpicks are correct. But wouldn't it be more inspiring to talk about how it could be used?
"Without electronics" except for the computer devices that have to be dedicated to monitoring for the back-scatter created by the 3-D printed objects and translate it into something meaningful.
This is a surprisingly simple solution to the problem: Make the working "device" (3D printed object) do as little work as possible - just enough to be detectable by the surrounding devices that are "real" computers anyway.
Actually, I wonder how much those devices would interfere with the normal wifi, like "when the wind starts to move at higher speeds, my movie stops playing" ;-)
It's a clever idea, but without a 1:1 connection between an object and a network, it seems too dumb to be useful.<p>How is my WiFi network supposed to know that the interference in the signal came from my detergent bottle - not my neighbor's, and not from some other random object passing by? Do I have to do some sort of pairing every time I buy a new detergent brand to teach my network about it?
Curious to see the impact this would have on wifi performance. Wouldn't this cause all of the really fun stuff making WiFi fast (beamforming, channel hopping, etc.) a lot less effective? Would be curious to see more information.<p>I imagine this would be really good for things like weather sensors or even security systems... but I have serious doubts of the efficacy of what they are doing.
So my WiFi should easily be able to detect if my (all metal) garage door is open or closed? I would like this. I run dd-wrt. Is there an all-software solution to this? Or does this research rely on fairly specialized hardware? (sorry, didn't have time to read the whole article)
While rather impressive it sounds like a huge security problem to me.<p>With machine learning someone could probably train models that know what you are doing in your home.
I am reminded of <a href="https://en.wikipedia.org/wiki/The_Thing_(listening_device)" rel="nofollow">https://en.wikipedia.org/wiki/The_Thing_(listening_device)</a><p>This development will probably lead to a new combination of something like RFID, WiFi and electromagnetic power delivery (AM radio).
Ignore the soap example and just focus on the underlying Wi-Fi backscatter technology. There are many applications in extreme low power or no power devices. Most are missing the point.<p>More details here. <a href="http://iotwifi.cs.washington.edu/" rel="nofollow">http://iotwifi.cs.washington.edu/</a>
How can this _possibly_ work? You can't make a connection to Wifi or implement any of the protocols that run over by passively reflecting some signals. All of them require some processing and two way communication. The article just makes no sense at all. It is 1st April?