Hi guys,

I want to detect the point, during a hand swing motion, when the arm aligns or comes to the level of shoulder like the 1st, 3rd and 5th ones in the figure.
Could using an accelerometer work? If yes how would I position the sensor as the arm could be rotated in any way about its own axis and not necessarily be like in the figure.
A design constraint is that I can place sensors on the arm only nothing on the shoulder.

Thanks in advance.

 

Do you want to know
- when the arm slows to a stop or
- when it is horizontal to the shoulder
- What if the arm swings higher than the shoulder?

Be a little more specific.

 

@Lestraveled When it becomes horizontal to the shoulder the arm might swing higher then the shoulder doesn't matter.

 

@Johnny1010

So, an accelerometer will not tell you position. It will tell when something is accelerating or not.

You need something to tell you position of the arm, relative to the torso. Correct?

 

@Johnny1010

So, an accelerometer will not tell you position. It will tell when something is accelerating or not.

You need something to tell you position of the arm, relative to the torso. Correct?

Yes exactly. What I had in mind was that the arm also becomes parallel to the earth's surface (when horizontal to the shoulder) so may be something could be done using an accelerometer's tilt like in mobile phones.

 

@Johnny1010
The trouble is, is that the arm is moving. This motion will overwhelm an accelerometers ability to detect "down".

 

@Lestraveled
Hmm, then what else could be done

 

Be more specific about what can and can not be connected to the person.

Could an low power IR laser be attached to the arm and a detector be on a stand at the right height in front of the person?

Fishing

 

An accelerometer will tell you when it is level, not necessarily moving, as one cannot escape the acceleration of gravity. That is why another name for them is "inclinometer." I suggest it either be on the wrist, if the arm is always stiff, or on the humerus at the elbow (e.g., one of the epicondyles).

Of course, level is relative to the Earth. With 2 accelerometers, one could calculate the relative position without respect to Earth level.

An indicator for rotation at the shoulder (e.g., goniometer) could also work, but might be more noticeable and/or restrictive by or on the wearer.

John

 

Does the sensor have to be on the body? Place an orange wrist band and elbow band and use video tracking libraries.

 

Really good suggestion but I am looking for a wearable and portable sort of a thing i.e. something only on the arm capable of detecting it. So, placing a stand or using a camera would make the thing a bit immobile.

 

Tell us about what this is going to be used for. Give us an idea of the motions and movements expected.

 

I am studying the mechanics of a person while skipping so the motion would be a repetitive one with the hands circling around.

 

@jpanhalt
I am a bit confused with using the "inclinometer" because if I place it at the wrist it would work only in cases when any one of the three axes aligns with the earth, but when the arm becomes horizontal to the shoulder, the arm could be rotated a bit and there is a possibility that none of the axes aligns with the earth so it might miss that case.

 

@Johnny1010
Can something small, be mounted on the back, in the area of the shoulder blade? Mount a single turn potentiometer on the shoulder blade. Connect a flexible shaft to the pot. Route the flex shaft around the shoulder to a slip connection on the outside of the forearm. This would allow full mobility and still report the position of the arm relative to the torso.

Edit: Delete the slip connection on the arm. The flex shaft itself will provide full mobility of the arm.

 

@jpanhalt
I am a bit confused with using the "inclinometer" because if I place it at the wrist it would work only in cases when any one of the three axes aligns with the earth, but when the arm becomes horizontal to the shoulder, the arm could be rotated a bit and there is a possibility that none of the axes aligns with the earth so it might miss that case.

That is why I suggested using two (2). One on the shoulder and one on the humerus or wrist, if the arm is always stiff. The person could be doing flips, in theory. Of course, constant motion doesn't affect an accelerometer, so if the person is moving at a constant rate and moves one element relative to the other you will be able to calculate the change based on relative orientation to Earth. However, if the person accelerates, it will mess things up.

A goniometer is probably simpler from your perspective. Fix the center of rotation at the shoulder and the motion detecting element (e.g, lever) at the elbow/wrist.

John

 

@Lestraveled
I was thinking of fitting in all the circuitry inside an arm band so I would have to go for a sort of t-shirt wearable for what you are suggesting. If somehow it could be done by placing things on just the fore arm or the upper arm.
I was looking at these
https://www.sparkfun.com/products/10289
May be it could be placed on the upper arm and the sensor could indicate the leveling of the arm relative to torso, but I think these might not perform very well in case of fast motion like that of skipping as it just has a mercury ball and it would be susceptible to the fast motion giving false indications as well.

 

Do you know what a potentiometer is? It is a small cylindrical device that has a shaft coming out of it. Turn the shaft and it changes resistance, which can be connected to an analog port.

The pot body could be secured to the shoulder blade area so that the shaft is horizontal. Some kind of linkage could be used to connect the shaft to the forearm. As the arm moves up and down the shaft will turn and the position recorded. If the arm moves horizontally the shaft will not move.

This would not be affected by any other movement (accelerations)

What do you think?

 

@Lestraveled
You have just described a goniometer. Thank you.

John

 

You are Welcome