balance substitution machine using accelerometer - total beginner and desperate


I want to build a helmet that has an accelerometer attached to it which finds the position in space relative to g and converts these signals into very weak electric currents which will sit on a tongue plate and give the person who wears it an alternative way to receive vestibular input (my son has no vestibular function and as a result very wobbly and cannot walk) I am totally new to this, and have no clue about electronics or programming but I am desperately eager to learn from y'all. Any input or guide into the right direction means very much to me. I dont know exactly how the accelerometer treats or processes the information about movement in space, but I am familiar with vectors and I wonder if it could be as simple as just attaching 5 little wires to the arduino board: +i, -i, +j, -j and - k. k would represent vertically upwards from the surface of the earth. there would be no positve k- axys because I dont know where to put it on the tongue. THe tongue plate would be rectangular (very small, 10mmx5mm) and the 5 electronic cables end in it like a cross arrangement, +i translates movement to the right and gives an impulse on the right side of the tongue, -i is to the left and gives in impulse on the left side of the tongue, +j is forward and send an impulse at the front of the tongue and -j is at the back of the tongue and sends an impulse when the accelerometer detects backward motion. -k is for downward motion and I would put it right in the middle of the tongue. so when the accelerometer detects downward motion the middle of the tongue tickles. The magnitude of the acceleration governs the strength of the current, sudden strong movement causing a stronger current. positional changes which are in directions which are made up of i-,j-, and k- components will result in different areas of the tongue being stimulated simultaneously by electric impulses. like a movement from (0,0,0) to (2,0,1) would mean my head goes to the right and downwards, but twice as much to the right and not forward or backwards. so 2i+k and the right side of the tongue would receive a current twice as strong as the middle, whereas front and back of tongue dont receive any impulse. Can I use the arduino Uno board to act as the intermediator between my accelerometer and the tongue plate (which will be 5 wires held together by half and inch of scotch tape on both sides) and how do I get it to break up the info from the accelerometer into currents of varying strength into these 5 wires? Thank you for reading up till here.

Total beginner --no clue about electronics ---- and you want to put "strong currents" on you sons tongue? !?

It depends on what kind of accelerator chip. Look at the specs. The last one I played with had 3 analog signals from the X/Y/Z planes, which I read with the analogRead. (I suspect from the specs that the chip actually sends pulses, but these are averaged out by a RC filter to a voltage)

My real suggestion for you is - get the Arduino, an accelarator chip on a breakout board (that probably includes the 3.3/5V s and sundry stuff the chip needs) and plug that together and then just do a lot of Serial.print with the values you read on the analogRead. This will give you an impression on what values you get - the noise, the high/low values and so on. (The chips can be set to senistive or wide range by some jumpers)

To detect "sudden shifts" you do not want an accelerometer, you want a gyroscope chip - this is better at giving you the rate of change and the accelerometer is better at getting the absolute tilt with respect to earth gravity field. You might even get a 9DOF (9 degrees of freedom) system (3 chips used together) used in flying drones, which will give you the whole sensory input you need.

I will not comment or aid in the tongue-palette. Potentially it can be lethal (small electrical currents at the wrong time may disturb heart rythm).

I don't know what "vestibular function" is but from the context I guess it's something to do with sensing the orientation or movement of the head - the sort of thing we'd normally get from our inner ears. Sending that via the tongue seems a very intrusive and on-intuitive approach as well as potentially introducing medical issues. Have you considered using some sort of touch stimulus on the outside of the head? This topic reminds me of a research team who gave their subjects a sense of direction by making a belt that applied a slight touch on whichever part of then pointed North - apparently the subjects soon stopped noticing it and just used their new sense without thinking about it. I guess that's similar to what you're trying to achieve here.

In case you haven't already don it, I'd suggest consulting your medical advisers to confirm that this is a sensible way to address the problem and see whether there is already some other known way to do it. If nothing else, that might also put you in touch with other suffers and perhaps some of them would be interested in collaborating. When the subject of medical assistive devices have come up in the past a forum member mentioned that they were part of a team who provided practical advice and support for this type of project. Depending where you are in the world and whether you're interested in getting help with your project, it would be worth reaching out to that sort of organisation. Unfortunately I didn't make a note of the name, but perhaps Google will be your friend.


I don’t know what “vestibular function” is

perhaps Google will be your friend. :slight_smile: :wink:

Such a device has been built and tested. Here is a 2007 article describing successful research on the topic: No doubt there are more recent studies.

Efficacy of electrotactile vestibular substitution in patients with peripheral and central vestibular loss

Y.P. Danilov, M.E. Tyler, [...], and P. Bach-y-Rita


Vestibular dysfunction of either central or peripheral origin can significantly affect balance, posture, and gait. We conducted a pilot study to test the effectiveness of training with the BrainPort® balance device in subjects with a balance dysfunction due to peripheral or central vestibular loss. The BrainPort® balance device transmits information about the patient’s head position via electrotactile stimulation of the tongue. Head position data is sensed by an accelerometer and displayed on the tongue as a pattern of stimulation. This pattern of stimulation moves forward, backward, and laterally on the tongue in direct response to head movements. Users of the device were trained to use this stimulation to adjust their position in order to maintain their balance.

jremington: Such a device has been built and tested.

The difference now of course, is how much cheaper the sensors - and Arduinos - are than in 2007.

The reason for the strange choice of using the tongue is that the tongue is second only - if indeed it is second - to the fingers in position discrimination, having a particularly large representation in the brain.

Accordingly, you would presumably require only a small are of the tongue to be involved, and the appliance would not be quite as cumbersome as it might seem.