Finger signals - hand as antennae?

[not2techy]

I have been handed a challenge that is beyond my reach.

I need a way to detect the change in skin conductance, or some other electrical signal on the surface of your hand when you lightly touch your forefinger and thumb together. In this configuration, the skin sensors have to be located on the wrist, I don't have access to the fingers directly. I can't use tendon or muscle movement as my signal, it has to be light touching of the skin. I just need a yes/no indication, touching or not touching.

Detecting a passive signal would be best, since it needs to be low power. I know the human body acts as an antennae picking up 50-60mhz a/c perhaps there is a way to play off of that phenomenon, and measure some change in the antennae configuration when the thumb and finger complete the loop? Could that work? Any other ideas?

 

[berkeman]

Wow, that's a hard one. Are you sure there is a solution to this? The impedance of dry skin and a light touch is very high...

 

[berkeman]

And do you need to be able to tell when the thumb is touching a finger, as opposed to touching some other object?

 

[not2techy]

I need to know when the thumb is touching the forefinger, and when it is not. Come on guys, there has to be a way.

 

[berkeman]

The only thing that I can think of offhand would be to use a pair of low-power ultrasound transducers, with one placed on the underside of the wrist near the thumb, and the other in the top middle of the wrist. You would have to experiment to see what kind of signatures you got with different hand positions and the thumb touching fingers, etc. I'm not sure it would be all that reliable, but maybe with enough DSP and experimentation, you could make it work.

Hey, can you use two rings instead? One on the thumb, and one on the forefinger? Very stylish, actually. Then you could do it with ultrasound, with a lot better isolation.

 

[not2techy]

Thanks Berkeman, I'll look into that idea. What is DSP? Sorry, but as the name implies, I'm not the sharpest knife in the drawer.

I like your fashionable ring suggestion, but I'm afraid that's not an option. Any other ideas out there?

 

[berkeman]

Thanks Berkeman, I'll look into that idea. What is DSP? Sorry, but as the name implies, I'm not the sharpest knife in the drawer.

I like your fashionable ring suggestion, but I'm afraid that's not an option. Any other ideas out there?

Don't sell yourself short -- that's against the PF rules, and I might have to give you some infraction points for that.

DSP = Digital Signal Processing. I was referring to the signal processing that will be necessary for you to separate your useful signal information from the ambient noise environment, and to increase the fidelity of your signal measurements.

What is the context of this assignment? What class is it for, and what year of school are you in? What is your technical background? Knowing these will help us to help you the best we can.

BTW, speaking of PF rules, homework and coursework questions generally should go in the Homework Help forums area of the PF, and not in the general technical forums like this EE forum. However, when the coursework question pertains more to an involved project, especially upper-division or graduate work, then the questions are generally allowed in the technical forums. I'll leave this thread here for now -- I think it's the best place for it at the moment. However, one of the rules of the Homework Help forums is that you show your own work and effort before we can offer you tutorial help. We don't do students' work for them, and don't offer complete solutions.

Now, having said all that, I've been thinking more about this, and poking and prodding my hand. I think I know what I would experiment with first if I were given this project assignment (for work or school or whatever).

But, instead of just giving all the details of my thought, in the spirit of the PF Homework Help forums, I'll ask a few questions and suggest some further reading and research for you.

What does the skeletal structure of the hand look like? What bones could you make contact with through the skin for the thumb and forefinger, that are located close enough to the wrist to satisfy the (very artificial?) restrictions of your assignment?

If you could pick the right bones to make gentle mechanical contact with through the skin, what mechanical mechanisms should you consider as ways to try to sense whether the finger/thumb-tip ends are in mechanical contact?

 

[not2techy]

berkeman,
Sorry, if my communication was misleading. I am not a student, at least not officially, although I am in a constant state of learning. I am an amateur inventor and this "challenge" is part of a larger project that I am working on. I really appreciate the thought you have put into this, and I wish I could answer your "homework" question, but I might need a bigger hint.

 

[berkeman]

Okay, I still have no idea what the application is, but since you are in the middle of inventing, I understand the lack of more details in your description.

So, I'll just say what I was thinking, and you can see if you can take it from there. If you really want thumb and finger specific information, I think you will need to use the appropriate metacarpal bones to localize the information. The 1st metacarpal bone is the bone associated with the thumb, the closest bone to the wrist. The 2nd metacarpal bone is the bone associated with the pointer finger, again, the bone closest to the wrist:

http://www.gwc.maricopa.edu/class/bio201/hand/anhand.htm

So you should try to couple some signal (most likely sound or ultrasound) into one of those metacarpals, and measure how much of it makes it back down the other metacarpal. You can probably ping both directions back and forth, to help improve your sensing capability. The device mechanically would be something like a wristband, with alignment marks or some other convenient way for the user to adjust it so that the sensors were making contact with the bottom ends of those two metacarpals.

Even with an appropriate mechanical arrangement like this, there will still be a fair amount of experimentation and signal processing development, before you will start to see whether this scheme is effective or not.

Hope that helps. Anybody else have any other ideas?

 

[dlgoff]

"Anybody else have any other ideas?"

I've be thinking and thinking and can't come up with anything. Your idea seems very sound to me. Would you use some kind of pizo tranducer?

 

[berkeman]

Would you use some kind of pizo tranducer?

I think so, but it would depend on what the typical resonant frequencies of the metacarpals and intervening joints were. Or else I guess you could just stay above the resonances and rely on propagation along the two metacarpal paths.

 

[not2techy]

Thanks guys, I really appreciate your input on this. I'll look into this approach and see where it takes us.