Spine posture reminder

I've recently started a project for my electronics class:

My idea is to use angle sensors positioned on the spine to track it's position in 3D.

If the user slouches or goes into a non-optimal posture for a certain time (maybe 1min to avoid accidental triggering), the system recognizes this and reminds the user via an eletrostimulation-induced tingling sensation.

Here's a shitty paint drawing of my idea:

For now, I'm thinking of using simple potentiometers with a solid linked chain. Adhesion will be provided by one of my labmate's silicon adhesive (resists sweat, good strength and doesn't leave residue behind.)

My big problem currently is the electrostimulation system. I'm having trouble finding data on where I should start, most of what I can find is made to cause pain...

Any thoughts/ideas/criticism/etc?


  • edited September 2016
    Pain is pretty good motivation. Just saying. If not then I would just go with vibration feedback. I wouldn't think there would be any need for electrostimulation unless what you're doing is specific.

    I think this is a great idea, by the way.
  • edited September 2016
    so put very simply you want to zap people to make them sit upright?

    If i remember correctly from my electronics class tasers are very simple to make via a disposable camera. use that as a starting point. get a capacitor to hold the charge (not a very large charge like the homemade tasers) and release it.

    option 1
    lets say like this:

    1) you slouch for 30 sec.
    2) charge begins on capacitor.
    3) you remain slouched for 30sec more.
    4) NON painful zap from capacitor.

    go to CVS and get one of the tens units (nothing fancy $30 usd) rip the thing apart and use that as a means to deliver the NON painful zap. it will have everything you need to connect it to the body and it uses low power.

    bare in mind I'm not including the whole sensor part or control of the zapping just the how you can NON painfully zap a person.

    forgot to add that while sitting for me equals awful posture. it hurts for me to sit upright in a chair. standing however i will stand fully upright and shoulders back. you may need to adjust the idea of fixing posture based on what is actually considered bad for the individual.
  • edited September 2016
    I would say go with vibration do to the fact its safer in this situation. You could have a marketable product. Will be watching this one....
  • one more thing a simpler approach may be to pull the shoulders back because that will get you to stand upright. maybe some form of a brace with elastics and a sensor?
  • I actually have shoulder brace in my little project notebook so I could give you more description if your interested in it.
  • @Meanderpaul
    I thought of using a disposable camera for the eletrosystem, but since it's for a class project, I have to design it myself... So no scrounging up premade circuits, but otherwise, yeah that's a great idea.

    I might grab a TENS machine and pop it open, should provide good ideas on how to proceed.

    My idea for using electrostimulation is mostly for noise since a vibrating motor in class/office could be pretty disruptive. Also, as above, I need to make an electronics project...

    Sure, that sounds like a neat design I could use!
  • I am just going to play devil's advocate to the guys saying that you should shock yourself. If you were shocked by a disposable camera in your back, you would possibly suffer at least temporary, and at most permanent paralysis. However, the TENS machine is a much better idea. Or the vibrations.
  • I'm not sure you fully read the post I said shock your self in.....I said NON painful as in VERY LOW POWER. And no it won't result in paralysis. At most if you do it wrong you get a burn and some pain.

    If you got paralyzed from a low voltage zap then we would have quads all over the planet from taser. Oh ya and I forgot to point out again that we made these in high school and used them on each other and guess what we are still walking. A camera doesn't have enough juice to do that kind of damage.
  • edited September 2016
    Still be careful there are dangers that must be respected when you run current through the body. Mainly across the chest/head.
  • My idea is to have one of the modules, possibly in the lower back, house the whole electro stimulation system. So the electrodes would be far from anything sensitive.
  • I am behind this project 99%. I'm not entirely sold on the shocking aspect but that's just my taste.
    How do you plan to map the spine in 3D with potentiometers and chain? You can get a distance reading between potentiometers but that will only be accurate if the spine is perfectly straight and nothing touches the chain. Unless you're sitting on a stool your back will likely touch a chair's backrest while sitting.
    Perhaps two potentiometers and a rod running to an eyelet on the next potentiometer module?
    [Joystick modules]
    This would be bulky and may be cumbersome but it would give you a good idea of where each module was.
    If you wanted to go overboard, you could add a linear pot to the connecting rod and get your distance that way.
    Depending on how many inputs you have you may also be looking at an analog MUX.
  • @McStuff
    Since this is an early version and I mostly want to validate the concept, my idea is using rotary potentiometers to simply measure the angle. The chain (more like a snake really) should be extendable to allow for the back to lengthen when slouching.

    As for the shocking, I'm really not going that far. How I see it is basically just a simple tingling, probably on pulsed intervals to make sure it's easily felt. Of course, fitting a simple vibrator motor would be easily doable, so not much of a problem there.
  • edited September 2016

    Here is something that is pretty similar, at least for the posture reminding. This doesn't really help much with motion tracking though.

    Perhaps just a series of IMU's, maybe 3 or 4 to start out with, all reporting their orientation data. You could even connect to wifi and whatnot. It would also be a bit easier to calibrate that way. Plus you they could be I2C and you could daisy chain them together so you don't need a ton of analog inputs and wires.

    It may also be a tad easier to link them together to form a snake type thing.
  • IMUs are a great idea actually, it would enable the snake/chain to be super flexible since the modules wouldnt need to be mechanically connected.

    I'll talk with my supervisor if that would be okay in the context of an electronics class, he might say no since that would probably take away a lot of the pure electronics design part.
  • So I've been looking into things and I'm having trouble finding data on the frequency, current and voltage I should use.
    What I can find is either less than noticeable or too strong.

    Does anyone have an idea of the range of values I should be using?
  • @Aeris

    I have been using a tens machine for back pain and arthritis in my hands.  I bought a cheap unit off ebay and have mainly been trial and error to get the right settings and positions. 

    One thing I noticed is that the some of the patterns and frequencies feel much different.  Some of the settings, with the electrodes on my back, have a feeling almost like someone pressing on the skin under the electrodes.  You would have to figure out exactly what frequency of pulses give that effect but it felt like it could be used to input data into the body from whatever sensor you wanted.

    Another thing is that the "shocking" sensation becomes a lot more tolerable the longer it is running.  At first, the low settings seemed pretty strong but soon I was cranking the power up to the highest setting.  The low setting still could be felt though so even a low power should work as an alert signal.

    Many years ago, back in high school, I made very simple shocking devices using a simple transformer and a dry cell battery.  I think we often used a hacksaw blade and a nail to "pulse" the power going into the transformer as a kind of AC current.  With DC, you only got the shock when first connected and when disconnected. 

    Nobody ever got hurt doing these experiments which included holding one electrode in each hand or having a whole group of people join hands and send the shock through all of them but looking back, it might not have been a great idea but it was a lot of fun.

  • edited September 2016
    I have a friend who built a glove for VR using strips cut from anti-static bags static shielding bags (the ones we find electronic components in). He placed these strips along the fingers in a glove, then measured resistance to gauge the flex and then mapped that to a virtual hand.

    Couldn't the same concept be used for this?

    Three parallel strips should be able to gauge the position of the spine in 3d. This would be paper thin and super cheap.
  • edited September 2016
    Thinking further-- a static shielding bag that big may be difficult to find.

    I think any conductive film could work.
  • Didn't have much time in the past few weeks/months, but I'm happy to say the project is advancing.

    I've built this TENS board, runs up to around 80V with frequency and pulse width adjustments.


    I've been trying to figure out what the best settings would be to induce the buzz sensation, but it seems that my skin is very fickle and passes from feeling nothing to intense twitching extremely fast.

    Perhaps the electrode placement could be changed, so far I've been testing on my left forearm for safety reasons... I'd rather avoid getting electricity near my spine and heart.
  • Very cool! Do you have the schematic?
  • I dont have the schematic at the moment since its a modified version of the Silicon Chip TENS unit: http://imgur.com/gallery/0tSu2

    Basically, I dropped the intermittant output and changed the IR2155 for a IR2110 and a 7555 timer since the IR2155 is no longer in production.

    Other than that, it's mostly the same with a few tweaks to increase performance and reliability.
  • Wooo the TENS unit + control is now working and twitching the shit out of my arm!

    Control is fully analog (course requirements) and features a low cutoff to allow for slight variations from normal posture + a upper clamp to max out the output at a predetermined level.

    Now that I'm no longer constrained by course requirements, I'll rebuild it using a microcon and implement the control, MOSFET switchin and flyback converter directly in software which should reduce the footprint massively on top of having it not be a total mess of wires.
  • Bravo dude bravo
Sign In or Register to comment.