Texas Instruments Watch Claims It’s A Computer Mouse

[youtube http://www.youtube.com/watch?v=TDL3JRG_zrs%5D

[Jack Toole] and his team [Aaron King] and [Libo He] sent in their computer interface dubbed the Chronos Flying Mouse. The video above explains the concept very thoroughly, but we’ll reiterate some of the highlights here. The project uses a Chronos EZ430 with its accelerometers to wirelessly transmit delta positions of the user’s wrist. Add a little open source software and you have a regular PC mouse, a video game joystick, a game wheel, and a few other different devices in one. We just love the suave feeling of snapping to click.

54 thoughts on “Texas Instruments Watch Claims It’s A Computer Mouse

  1. yes, being able to hold a book really is an extra feature. who would’ve known.

    i can’t say i’m impressed; these guys basically did for chronos what johnny lee did for the wiimote, except that johnny’s implementation didn’t imply you can only use the setup for 15 minutes at a time before your fingers get tired from snapping.

  2. What is ridiculous is that TI couldn’t have found a better person to do the presentation or a better place to put the cue cards so the guy doesn’t have to keep his eyes off the camera looking like a tool.

  3. I like it. It’s more portable than a WII mote.

    Although I got to wonder. How this interface do with people with arthritis?

    Still I like the compact, innovative design. book as a steering wheel, broomstick as a sword, Rapid fingersnaping as gun fire?

    I am more curious, how this will work for artists, engineers and so on?


  4. Fun project, i bet the software side is interesting… And if they enjoyed it that’s what counts…
    But it is objectively useless as a mouse. It must be a nightmare to navigate, look how odd and controlled their stance is while using it..
    Now imagine that while lying on a couch as suggested! Now add you also have to use your second hand to push small unfriendly buttons. Well OK, control might somehow improve..
    What if you actually want to do something in between browsing, like tie your shoes? unstrap/strap or turn-off/turn-on each time?
    How do you even snap when holding a book ???
    I’m not even bringing up watching porn..

  5. This demonstration makes you appreciate how wonderful the classical mouse actually is.
    Yeah it ends up being sarcastic because of that but it’s actually true; the mouse is a pretty good invention actually, and sure we all feel there must be a move forwards but so far nothing convinces really.

  6. Addendum: I really like those TI gestures towards hackers, and on that count nothing but praise, but this video is just silly though.
    Anyway please continue making these things available TI, thanks :)

  7. or how about people like me who just have slightly shaky hands? – not very effective and it does leave a lot to desire but at the same time so does a mouse – most notably when sitting on the couch with a wireless mouse on the arm just trying to control the computer can be an amazing hassle – especially as the distance from the tv increases times the resolution of the screen (dist * reso)/accuracy – since the mouse pointer would be smaller the higher resolution, and harder to see at greater distance and the more sensitive the mouse the harder to click what you are after anyways – but beyond all of that it is useless for web browsing without a keyboard

  8. @DeadlyFoez +1

    The concept is ok, nothing really ground breaking IMHO. And I am still not sure if one can really use this for a long period of time… also what if one’s back is itch and he/she try to do a scratch? ;)

  9. I did the same thing, but strapped the accelerometer to a hat. My setup could just as easily be put into a glove or wrist band.

    I agree with the comments on the video. Yes, my video’s aren’t professional either, and I’m surprised no one said anything about it lol.

    Anyway I do like this post. I like things done with accelerometers. Wish I had the money for that watch, I have some ideas already.

  10. This is retarded. The software which comes with the mouse has this feature built in and I could knock something similar out in minutes using python. HAD is on the rocks.

  11. The whole idea of a watch as a control is just silly. The cool concept here is using an accelerometer as a mouse, but definitely needs some adjusting as noted above.

    Consider this:

    Grab a normal mouse (for parts and form factor), add a button for your thumb that “activates” mouse movement. That way if you want to scratch your back, you just let go of the button! Plus it’s comfortable to use, just like a regular mouse.

    Now If you wanted to get fancy and make it a joystick/remote/etc. You make a small project box with the accel/bluetooth modules that fits in the bottom of the mouse, but can be taken out and used with a book/broom to make other types of controllers.

  12. To comment on the video, they used this guy because he was the best at manipulating the mouse I’m sure.

    Also the camera crew (probably 1 person) could use lessons on white balancing and general camera operation.

  13. if you snap to click, how do you right click? press a button then snap?
    why would i want to use two hands

    when i see things like this and touchscreen computers and such, it just makes me think your arm is going to get tired

  14. Thanks for the comments!

    A regular mouse is great! But sometimes (and probably increasingly in the future) you’ll need to interact with electronics without having to find a hard surface all the time. A television already is a good example of this – no one would want a remote that you couldn’t use when you picked it up.

    @The constructive comments: Thanks!

    Before going through the other points, some of the questions can be answered by reading the wiki link (http://processors.wiki.ti.com/index.php/Chronos_Flying_Mouse)

    @the negative video comments: Sure, the video could use work. It was done late at night on the last day, because we spent quite a lot more effort perfecting this for the competition. This wasn’t a professional, team-developed project over several months. We shot and edited the video in the timeframe of a couple hours and without professional equipment. Sure, it leaves a lot to be desired, but if you’re interested in a specific question, you can check out the wiki or download it.

    @stol24: If you *must* have the cursor remain in place while tying your shoes, you can press the transmit button before and after. Although usually the mouse wandering around the screen a bit while tying shoes isn’t a big deal

    @shinjikun34: You can adjust the movement deadzone to the shakiness of your hands

    @sariel: It’s not a professional video. The program is freely available to download though, nothing’s faked on the video.

    @Tech B: Cool!

    @Alexander Rossie: If you try both the built in “mouse” feature and our project, and can still say that they’re the same, or even similar, the go ahead. But don’t say that before you’ve tried both. The original is definitely something you could hash out in an hour. Our project could definitely not.

    @Spork: We used the presenter we did not because of skill, as it’s not difficult to pick up, but based on who knew the most about the project. As to the actual control, we’ve had several other people try it and say it’s a lot more intuitive than they expected it to be.

    @zool: To right click, press the upper right watch button (similarly, the left perform up and down scroll wheel). It’s not ideal, but most people don’t right click nearly as often as they left click. Also, some tired arms could probably do the U.S. some good, as the latest Wii commercials point out.

  15. Pretty cool and I love that he’s seriously geeking out on it.

    Now if I could only get TI’s fulfillment house (Harte-Hanks or some such) to ship mine! I got the 433MHz. Yes, I have my amateur radio license and I can get it.

    But they can’t seem to get their shit together at both TI and HH.

  16. @Jack Toole
    It is nice you are taking the time to read through comments! My point is the following..
    If you:
    -tie your shoes
    -go to the fridge and open a soda
    -use your hands in general
    you WILL move the cursor and I assume you will accidentally trigger one or more clicks. Some random clicks can range from harmless to disastrous. Turning it on/off by clicking takes away the fun, since not-clicking is the whole point… not to mention clicking every half a minute is tedious.
    Anyway, you know best.
    Good luck with the proj!

  17. @stol24
    I’m excited to have this on here and share our project a bit, rather than having it sit on the wiki unseen, so it’s nice to read everyone’s comments

    You’re right, it is a pain, and we were thinking about having a disengage mechanism, although with time constraints and the use of every dimension already we weren’t yet able to implement one.
    If there’s a nice motion that makes sense for disengage, and doesn’t too easily trigger on/off, I’d be interested in adding it.

    We do perform a bit of sensing to try to identify if a click was intended or not, based on mouse position, but not at the level where you could just go about your business while it’s on.

    You’re right. It’s used for the options reading rather than the com port – we need to save and load the user’s settings when the application is closed and opened. It shouldn’t be too hard to take that part out of the source (it’s all in options.cpp) and just load the option defaults by mimiking the save code in reverse. I’m hoping to get rid of that and write or find a better INI reader eventually, but with project time constraints, there wasn’t a point doing something that voters wouldn’t see the benefits of before the deadline.

    It also might be possible to download a prebuilt boost::program_options.lib and header files and just use those, or when I get some time upload those too.

  18. Great application – if you’ve used the chronos control center you know this was needed. Unlike the other submissions, this can be used for more than one narrow purpose. I might buy the watch just for this app.

  19. Great stuff and has turned me onto these watches at this price. Can anyone tell me if you can use 2 simultaniously and differentiate between the signals?

    Onto the use of it as a HID…use of this as a pure pointer isnt going to be as good as a real mouse of course, any one here REALLY want to use a wii remote as a substitue for their mouse? No thought not.

    But in an environement where precision isnt necessarily demanded AND hands free control is needed (e.g. think as an interface to a smart phone) these could be great.

    As for the video and presentation skills of these chaps, BRILLIANT! Best way to get good at something youre bad at is practice so I really hope the crappy comments here dont put these chaps off.

  20. @Lee Jackson
    The wireless protocall in the default watch software (SimpliciTI) can differentiate between channels, so you can have multiple watches attached to multiple computers just fine, although making sure each links with the right computer is a tad difficult, but doable.

    It’s possible to connect multiple watches to the same computer, but I’m not sure if the Chronos Control Center or the Chronos Flying Mouse software supports it.

    And thanks for the support!

  21. I’m surprised nobody has brought this up yet, but I was thinking with 2 of those watches, you could, at least in theory, create a Minority Report style interface.

    I noticed a trend in smartphones recently, where you zoom in images, google earth, etc. by moving 2 fingers apart or towards each other on the screen, and you scroll or move to the next image by dragging the image on the screen.

    These watches could be used for a non-touch version of such interfaces, I’d think.

  22. @Davo1111

    If you have the mouse, then you can download and try our project before dismissing it (or read the other comments here discussing that issue). And if you had done that, you would already know there’s really no comparison.

    If you don’t want to read the rest, it basically comes down to this:
    The Chronos Control Center is not reasonably usable
    The Chronos Flying Mouse is.

    I have been asked by the Chronos Control Center group how difficult it would be to change the Chronos Control Center to use the Flying Mouse code, and this is why:

    The “standard” mode is velocity based, so you have to balance the watch to even try to stay in one spot, can’t click with one hand, can’t act as a joystick, is raw-accelerometer data based rather than angle-based, so if you flip it past 90 degrees it will get confused, and motion will be reduced at angles significantly different than 0 degrees, has near no customization options, and won’t remember your calibration, the only thing it stores, between runs. A full features list (all of which are lacking from the Control Center) can be found on the wiki: http://processors.wiki.ti.com/index.php/Chronos_Flying_Mouse#Chronos_Flying_Mouse_Features

    I’d appreciate it if people who have the watch actually try it before dismissing it :). If not, I can assure you that the “standard” feature is quite basic, and wasn’t ever intended to be more than a demonstration of reading the accelerometer data.

  23. @Sparky : Part of the reason I was asking about the use of 2 of these. You can get away with 1 but 2 increases the range of gestures that would be possible (or at least, the range of natural gestures). Dunno, I’m looking for something to base my masters thesis on (HCI) and had been thinking of homebrew skinput but these might be fun. Course, still have the dead reckoning problem although if the accellerometers can be used to workout the angle of the watch (Is this possible Jack?) theres the possibility to have a good stab as to its current position.

    Anyways, tempting to look through the source on this one and buy a couple to play with :)

  24. @Lee Jordan
    I would quite enjoy seeing what could be done with two.

    But for gestures you’re going to run into the same issue we did:

    Accelerometers can distinguish orientation or acceleration, assuming you know the other, but not both together.
    (In practice it’s easier to measure orientation by assuming all acceleration is due to the normal force, which is close to correct most of the time)

    With cheap accelerometers (as in the watch), it’s close to impossible to distinguish movement from rotation, especially very small rotations, since the normal force the watches are feeling is so much greater than any acceleration the watch feels from movement.

    Various tricks with assumptions about not moving and twisting at the same time could work to get usable input on a higher end accelerometer, but with the noisy accelerometer in the watch, distinguishing motion and slight twisting back and forth is difficult even by a person, although larger twists are possible to identify.

    If it’s for something important like a thesis, I’d recomment going with something that has both an accelerometer and a gyroscope, which would let you measure both the orientation and objective acceleration.

    I’m not sure if a gyroscope could be attached to the watch somehow, since I’m more of a software developer, but that’s another option.

    And good luck! I’d be interested in seeing how it turns out.

  25. Nice, that’s awesome, a lot more user friendly then the stock software. I had a chance to play with one of the stock ones at IEEE SouthEastCon and I was impressed. Been thinking about picking one up, its a pretty amazing little gadget.

  26. @Jack : Thanks for the input; I had thought that would be the case but hoped you guys might have worked around it. Still might have to treat myself to a couple of these *grin*

  27. (PC, TV or other) Putting the watch on, taking it off, going out with it still on wrist and inconveniencing other users. As a mouse = as a remote = fail. For the PC, sure still use mouse/ pad as Jack admits in comments. TV, add a trackball to existing remote (cheap) or simply design the software interface properly so as not to require mouse style input. Tired arms is not so good for the elderly. Yes, this _is_ a cool and fun project, but does seem like it has yet to find the problem it really solves.

  28. @d1ggitydan
    Several people here and on the wiki have downloaded and tried this. I assure you it’s not a fake :).

    When I get fed up with snapping, a sharp wrist flick, for lack of a better name, works fine.

  29. Looks like a neat toy.

    I’ll have to pick one up to try it out, hopefully along with one of TI’s new $4.30 USD uC development kits.

    TI has come out with some neat stuff lately.

  30. I unfortunately don’t have Linux support for the Chronos Flying Mouse software – porting a GUI like that would be time consuming and not worth it for a competition where that’s not a judging criterion.

    TI however does have linux support for the Chronos Control Center and all of it’s compiler tools, and the free mspgcc is linux only.

    You can find Code Composer Studio beta for linux here:

Leave a 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.