Gesture Control Uses WiFi Doppler Shift

wifi-gesture-control

We’ve said it before: in the future simple interfaces will use nothing but your body. At least at first glance that’s the case with this WiFi-based gesture control system. If you have Internet at home you probably have a WiFi access point. That’s the first portion of the equation. The remainder is a way of measuring how the radio waves bounce off of your body. So far this is being done with Software-Define Radio (SDR) but researchers at University of Washington think it may be possible to build the technique into future WiFi devices.

The demo video shows this man waving his arm to adjust the volume of his home entertainment system. Intuition tells us that this would be impossible if your arm wasn’t the only thing in motion at the time. But that issue is quickly addressed. Multiple antennas can track multiple people at the same time. There is also consideration for false-positives. The system requires a moderately complex wake-up gesture sequence to prevent you from, say, accidentally turning on the stereo when you roll over in bed.

If you’re having trouble wraping your mind around this, consider this ultrasonic music player. The WiFi version does the same thing, but processing changes in the returning radio waves is much more complex.

[Thanks Mateusz via UW Today and Y Combinator]

29 thoughts on “Gesture Control Uses WiFi Doppler Shift

  1. And nothing could possibly go wrong…

    The website says that you need a “startup sequence” before activating the actual gesture, which can be seen as a “personal password”. However, eavesdropping somewhere in the neighborhood would be enough to copy the plaintext. A replay-attack is inevitable. Hell, brute force is easy done from a remote location.

    This type of technology, seemingly useful, is security-wise fundamentally flawed. Let us all hope it will never ever catch on.

    1. That’s only a concern if you live in an apartment. Even then, you should only be concerned about your direct neighbors.

      You most defiently don’t have to worry about being snooped by someone “somewhere in the neighborhood”.

      And if you did have these concerns, the bigger concern is that someone is tracking your movements in real time in your house. That would worry me much more than someone figuring out how to change the volume on my stereo.

      1. You have to fully see the implications of this technology. It can tell wether you’re home or not, and how many persons are at your “party”. And what windows are near you. And if there are any dogs in the house. And if you see someone else while your wife is away. And that information could be hacked and accessed from anywhere in the world. Sounds like a lot of things to consider besides just the volume of your stereo.

        1. What you fail to realize is, that regardless of this application for WiFi signals, those signals are already almost everywhere. Even if you don’t have any wireless devices turned on, your neighbours will. So your every move could already be watched. In other words there’s nothing left to consider at this point, except the security of this particular application… Everything else is already lost (that is of course, unless you consider isolating all your walls and windows against RF and throw away your cell phone).

          1. There could be a business opportunity there for Faraday cages built rooms/houses into for people who care more about their privacy than the U.S. Peeping Tom government, voyeur of the world.

    1. I don’t know about Batman, but didn’t Zaphod Beeblebrox have a radio in the Heart of Gold that changed stations this way? And he had to stay still to stay on the station he wanted?

  2. The problem with radar is that a transmitter must be on. WiFi density is already a problem, and at least for the moment I would say that having a could connection for downloading content is a more valuable feature then gesturing.

    Propellers are not natural, but they are more efficient than any natural analogues for moving fluid. Large hand movements might be natural, but are they really better than remote controls? Probably not: pushing a button on a remote is tremendously efficient and well-understood.

    1. But what happens when I lose the remote?

      If my hand is the remote, it will hopefully remain un-lost.

      And in the event that I do happen to lose it – well, I’ll unplug the stereo as I wrap my bloody stump in a rag and drive myself to the hospital.

  3. That’s pretty amazing!

    A couple of potential issues. First, doppler only detects if something is getting closer or further away from the antenna. So a single radio in the horizontal plane of the user wouldn’t be able to detect a rising hand movement, for example. You’d need one above (or below) the user for this. A quick scan of the paper seems to show they need at least four antennas to capture movement in 3D space with fair accuracy. Which I assume must be carefully placed in strategic locations around the capture area, not just all on the back of a single radio separated by mere inches like a normal AP.

    Sounds a bit tricky. And what if you live in a building where you don’t have (or don’t have access to) an attic, basement, or multiple floors? You could put one antenna on the ceiling, but sensitivity/accuracy of vertical movement detection would fall off pretty quick the further you got from that antenna; as once you exceed a 45° angle from it, it would be detecting horizontal movements better than vertical. In your average apartment, that’s only a few feet! Would you then need sets of antennas in each room? Additional radios to avoid having to run long antenna cables all over the place?

    Which leads into the second issue. I agree with [jpnorair] that WiFi density is a problem, and would prefer not to have another device (or possibly many devices in this case) transmitting frequently. I’m also not sure how tolerant a device such as this, which uses such small RF nuances, would be to interference. Either false detections, or failure to detect a valid motion, would get aggravating *very* quickly.

    So while I like the idea, I think I would trust an ultrasonic version that does the same thing more. Sure, it would be limited to a single room. But it would be cheaper, less prone to interference, possibly DIYable, and have no privacy issues for the paranoid. Plus, such a device could inherently detect audio, which allows the option of combining audio triggers with motion triggers. I think simply doing a double-clap to “wake up” the device would be faster and nicer than some wild hand-waving. The device could verify both the clapping sound and the appropriate hand motion occurred, providing good rejection of false positives. Or if someone’s sleeping and you don’t want to disturb them, you could use a more complex, strictly gesture-based “wake up”.

    It would be very cool if accuracy and sensitivity could be improved to the point where it’s possible to somehow discreetly trigger a call to the police, should you be threatened in your home, Or call paramedics if you become incapacitated and cannot reach a phone. How this could be done with near zero possibility of false positives or negatives I don’t know. I think I recall WiFi can also be used to detect breathing and heart rate, so perhaps a suitably elevated heart rate, plus deliberately breathing in a specific pattern, could serve as part of the trigger. Would be interested to hear if others have some ideas on this.

  4. “The machine was rather difficult to operate. For years radios had been operated by means of pressing buttons and turning dials; then as the technology became more sophisticated the controls were made touch-sensitive–you merely had to brush the panels with your fingers; now all you had to do was wave your hand in the general direction of the components and hope. It saved a lot of muscular expenditure, of course, but meant that you had to sit infuriatingly still if you wanted to keep listening to the same program.”

    Douglas Adams

  5. “The machine was rather difficult to operate. For years radios had been operated by means of pressing buttons and turning dials; then as the technology became more sophisticated the controls were made touch-sensitive–you merely had to brush the panels with your fingers; now all you had to do was wave your hand in the general direction of the components and hope. It saved a lot of muscular expenditure, of course, but meant that you had to sit infuriatingly still if you wanted to keep listening to the same program.”
    Douglas Adams

  6. Dear Hackaday. I absolutely love how you have changed the length of the articles in the primary homepage, i hated having to click on every article that i found interesting (which is most of them) and then having to wait for the thing to load…

    I greatly appreciate the change to be able to read everything on the front page and having all supporting documents (like links and youtube) in the “read more” link

Leave a Reply to MrshkoCancel reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.