Greetings everyone,

My name is Matt, I am a Software Engineer in Baltimore, MD. My grandmother of 87 years is suffering from Parkinson's disease. Recently her speech has become unintelligible, so not only can she barely move it is also hard for even basic communication.

During a recent visit I realized that she could still squeeze my hand which sparked an idea. I would like to build a basic yes / no indicator with a green and red LED. The trigger would be something similar to the air bulb on a blood pressure cuff, one squeeze= yes (green LED), two squeezes = no (red LED).

I found a diagram here but unfortunately it generates a random yes or no:

http://www.learningelectronics.net/circuits/yes-no-indicator-has-zero-standby_15.html

I would also need some way to change the delay time between squeezes. Since I do not know how fast she can squeeze the air bulb, I would have to adjust the timing while working with her.

I'm an old Java guy but pretty update to date with the latest web technologies, HTML5, CSS3, AngularJS. While I have built many home computers, basic electronics is a new frontier for me. I do have a Raspberry Pi that I have tinkered with, but I thought that might be a bit overkill for just a simple yes/no switch.

If anyone has time to assist me with a wiring diagram and parts list, it would be greatly appreciated. Plus any other ideas about a trigger/switch methodology would also be very helpful. Unfortunately time is of the essence, therefore I am grateful for any input. Thanks for your time.

Matt

 

A solid state pressure sensor is $13 from Digi-Key. That, a 2N7002 transistor, and a resistor and LED get you a basic squeeze bulb-to-light converter. You could add a small piezo beeper for audio feedback. Power can be 4 AA batteries and a low dropout regulator (the sensor requires 5V, not 6V), but the sensor can draw 10 mA, kinda hungry. Any 5V USB power adapter will work for the cost of one set of batteries.

What you described is a very basic frequency discriminator, easy to do with 555's and gating, CMOS logic, opamps, or discrete parts. I don't think it will give any new information, but there are lotsa ways to go if you decide you want something other than the basic sensor described above.

ak

 

So I sketched out a scheme. This scheme does not need the switch on. Power lithium battery 3 volts.View attachment 87254

 

A solid state pressure sensor is $13 from Digi-Key. That, a 2N7002 transistor, and a resistor and LED get you a basic squeeze bulb-to-light converter. You could add a small piezo beeper for audio feedback. Power can be 4 AA batteries and a low dropout regulator (the sensor requires 5V, not 6V), but the sensor can draw 10 mA, kinda hungry. Any 5V USB power adapter will work for the cost of one set of batteries.

What you described is a very basic frequency discriminator, easy to do with 555's and gating, CMOS logic, opamps, or discrete parts. I don't think it will give any new information, but there are lotsa ways to go if you decide you want something other than the basic sensor described above.

ak

ak, great information, thanks for the note on the pressure sensor, exactly what I was looking for. Appreciate the help.

 

What pressure sensor. It is good common clock button, as shown in my circuit. Go from a broken mouse. You programmer can write a program, and you use ready-made mouse.

 

So I sketched out a scheme. This scheme does not need the switch. Power lithium battery 3 volts.View attachment 87254

Bordodynov,

Thank you for taking the time to sketch a diagram for me, that software looks pretty cool. Could it by chance generate a parts list? Unfortunately the image posted is a bit pixelated and I can't make out some of the IDs of the components. Again, thank you for your time.

Matt

 

What pressure sensor. It is good common clock button, as shown in my circuit. Go from a broken mouse. You programmer can write a program, and you use ready-made mouse.

Bordodynov, yes I have plenty of old mice laying around, the reason I thought a pressure sensor would work is because of the Parkinson's disease, her movement is extremely limited. It would be difficult to near impossible for her to click a mouse button, however she was able to generally squeeze my hand, that gave me the idea for the air bladder.

 

My thinking is the "user interface" is the hard part. An old mouse may yeild a nice microswitch that could be adapted onto a small ball and covered for the squeeze switch. That part can be created apart from the controller.

Instead of timing her squeezes (and I assume her speed will vary on good days and bad days) how about a simple "one flash for no, two flashes for yes." LOTS of flashes could mean something else (HELP).

Matt, I am very sorry about your grandmother. Welcome to the forums.

 

B's comment opens up a question. What other kinds of switches or devices could she handle besides a squeeze bulb? A pushbutton switch mounted on the end of a hand grip or tube is called a pickle. Hold it in your hand with fingers wrapped around the tube, push down with the thumb.

As for the yes/no indicator, two things to consider. First, in order to get a single indication for each response, the circuit has to wait through the "yes" period to make sure there isn't a second press coming within the "no" timing window before it can indicate a "yes". The longer the time window for the 2nd press, the longer the delay before getting a green light from the first press. Second, how does the circuit reset? A fixed-time indication, like 1-2 seconds, with a shorter timed beep underneath?

I think 1 second is too short a time window to discriminate between answers, so lets start with 2 seconds for discussion. First press, 2 second monostable (timer circuit) fires. If at any time during this time a 2nd switch closure happens, you get an immediate red LED for 1 second (plus the optional short beep). If the first monostable times out, you get a green LED for 1 second. Or something like that?

If you eliminate all of the logic and gating required for that, and go with a single indicator that blinks/beeps whenever the button is pushed, people sitting with her will pick up on her yes/no timing after only 1 or two responses, and automatically adapt to changes in her timing without even noticing. Having a circuit do that reliably probably *would* take a Pi. IMO, better to let the electronics handle the blinking and beeping, and people's brains handle the discrimination.

ak

 

I may be jumping way out of the box here, but haven't people used a device that can follow the eyes? This allows them to type onto a screen by just moving their eyes from letter to letter on a display. Combine it with the type-ahead used on a smartphone to reduce the effort.

I'm a fan of not reinventing wheels, and I think your path is already well trodden.

 

I presented your question to a professional OT/SLP (Occupational Therapist/Speech Language Pathologist). The response is that the patient needs to be assessed by an Augmentative Communication Clinic or Technology Access Clinic in order to determine an appropriate access mechanism.

From the perspective of a simple Yes/No device, this would be very limiting. A device that can provide a broader range of responses is one that is based on "scanning". This might consist of a number of LEDs with one lit LED that cycles through a number of possible responses. Squeezing the bulb would allow the lit LED to stop at the desired response. The speed of the LED would be adjusted to suit the response time of the patient.

Hope this helps.

 

Nice to see you @MrChips.

If you elect to make something yourself, how about using an Arduino? They are cheap, already assembled, easy to program and have all the peripheral support you are likely to need plus the ability to reprogram delays etc. You would have no trouble with the programming and the hardware is all done for you. I'd think that would be a more expeditious way to a solution than a wired hardware solution.

We've been having a spirited discussion about Arduino in the new scheme of AAC but for this one it seems a perfect fit. Overkill maybe, but since you're not quite sure what the final requirements are, more horsepower than you need is a good thing. I've had many technical challenges over the years - buying too much computer for a one-off job has never been one of them. Even one of the small UNOs would do it.

Any help you might need getting it going is right here on AAC.

 

I may be jumping way out of the box here, but haven't people used a device that can follow the eyes? This allows them to type onto a screen by just moving their eyes from letter to letter on a display. Combine it with the type-ahead used on a smartphone to reduce the effort.

A friend of mine with ALS had one, and got to be a surprisingly fast typist with it. Not cheap even with insurance, and there is a significant learning curve even for someone with extensive computer skills. But it is off-the-shelf, well regarded and supported, etc.

ak

 

Greetings everyone and thank you all for the great suggestions. I believe the scanning eye-tracking typing is very cool and would probably be a great fit, however they are very expensive and I'm not sure how our family could afford the device along with her care. That is whey I was trying to find something I could build for around $50.

A funny story, when I tried to give them a computer back in 2001, my grandmother had never seen a mouse. As I showed her how to move, she would watch her hand move the mouse, look at the screen, go back to watching her hand move the mouse. It was quite comical but very touching at the same time. I will go back and look again for prices on the eye-tracking technology.

As for switches, some very good points. I like the idea of the 'pickle', I could take and old mouse apart and put the switch on the end of an old handle and see if she has movement enough in her thumb to press the switch. If that is the case, definitely the way to go over the pressure sensor and air bladder method.

As for the circuit diagram, I think JohnInTx said it best, you can never have too much computing power. With the mouse idea, I might just give it a shot with creating a small java program on the Pi with an old monitor that would respond to one or two mouse clicks. Might be a bit overkill, but like you said, i could improve it over time to display other responses depending on clicks.

I think I'm off! Thank you all for the direction and 'talking out ideas', greatly appreciate everyone's time.

Matt

 

Greetings everyone and thank you all for the great suggestions. I believe the scanning eye-tracking typing is very cool and would probably be a great fit, however they are very expensive and I'm not sure how our family could afford the device along with her care. That is whey I was trying to find something I could build for around $50.

A funny story, when I tried to give them a computer back in 2001, my grandmother had never seen a mouse. As I showed her how to move, she would watch her hand move the mouse, look at the screen, go back to watching her hand move the mouse. It was quite comical but very touching at the same time. I will go back and look again for prices on the eye-tracking technology.

As for switches, some very good points. I like the idea of the 'pickle', I could take and old mouse apart and put the switch on the end of an old handle and see if she has movement enough in her thumb to press the switch. If that is the case, definitely the way to go over the pressure sensor and air bladder method.

As for the circuit diagram, I think JohnInTx said it best, you can never have too much computing power. With the mouse idea, I might just give it a shot with creating a small java program on the Pi with an old monitor that would respond to one or two mouse clicks. Might be a bit overkill, but like you said, i could improve it over time to display other responses depending on clicks.

I think I'm off! Thank you all for the direction and 'talking out ideas', greatly appreciate everyone's time.

Matt

Hello, Matt. Nice to make your acquaintance.
I am the father of a a most-loving, beautiful 16 year-old girl who was born with cerebral palsy and spends her days either on her wheelchair, or playing with her iPad on the floor (that is, when neither her brother, mother nor me are pouring our full attention on her, of course).... so I have some experience regarding communication issues. I only want to make you aware that, regarding motor-limiting skills, each case is unique and has to be handled in a particular, case by case basis.
My first advice to you is this: Make sure that you can adjust whatever device you choose to build in such way that hysteresis can be accounted for. That is, include some sort of adjustable mechanism (or electronic component) that could make it easier for you to differentiate between accidental and voluntary interactions.
I'll be watching this thread, and I promise you I'll try to help you in any way I can...

PS: That story about the mouse and your granny was heart melting...

 

As I showed her how to move, she would watch her hand move the mouse, look at the screen, go back to watching her hand move the mouse. It was quite comical but very touching at the same time.

Same experience with my mom, now 89. She never did master the little move to lift and reposition the mouse, to allow longer travel in the same direction. To her, parts of the screen were off limits because her mouse was always getting trapped at the edge of the pad.

Good luck with the project. Be sure to let us know how it goes.

 

Here are some nice pre-made options, either to buy or copy. I've used the $30 one, back when it was a lot less than $30.

http://www.digikey.com/product-sear...e=0&rohs=0&quantity=&ptm=0&fid=0&pageSize=500

Another variation:
http://www.digikey.com/product-detail/en/40-3352-02/528PB-ND/611177

ak

 

any other ideas about a trigger/switch methodology

How about a large (i.e. easy to grip) lever-actuated microswitch? Not only does it give an audible click-clack when operated, but the switch is easy to interface to whatever LED-driver circuit you choose to use.