Project idea seeking comments - golf swing

[pine 140]

Friend of mine is a golf lover, and he mentioned some devices that seems to record the swing for review and study, so as to improve the technique.

 

Wondered if this can be done from a hobbyist. 3-axis accelerometer, 430 MCU, and possibly a bluetooth module and 3v coin cell battery.. If it is doable at all, what kind of precision do i have to look at for the accelerometer? Like those controlling the quadcopter? Needless to say the swing completes in seconds and i think the sampling rate, processing rate (by the mcu), and transmission rate (to some host device for recording) are very critical?

 

The thing i have in mind is a small board with minimal components, and can be stick to the golf club. A host device, maybe a cell phone app, listening and control the recording of 3-axis data.

 

Any suggestion is much appreciated

[chibiace 46]

how are you supposed to compare your data when you dont know what a good swing looks like?

 

perhaps one of those wii motion plus thingies and the socket adapter, i think they are i2c, might not be fast enough. but might be handy for prototyping.

bluetooth data could be sent after a swing?  how to initialize the capture or if the values look like a swing it captures and sends it?

would you need accurate timing?

 

i can easily imagine a small stick board with usb rechargeable battery and a case that clips tightly on to the club.

and an android and iphone app with different coloured line graphs. lol

 

kickstarter?

[abecedarian 330]

You'll need more than 3 axes to calculate a golf swing. Besides the obvious swing itself in 3 dimensions, you also have to consider the physical orientation of the club head throughout the swing, at least at the point where the club head impacts the ball.

 

In my opinion, you would need 3 gyroscopes and 3 accelerometers to get a close to accurate representation of the swing. The gyro's would measure the orientation of the club head throughout the swing and the accelerometers would measure the forces applied in 3 dimensions during the swing. I might also suggest some sort of proximity sensor on the club head so you could measure the distance from the club head to the ground when the club is in the all-so-important position where it smacks the ball. ... and maybe some pressure sensor to measure how much energy is actually transferred to the ball.

 

And then you have to correlate the gyro's with the accelerometers... and as mentioned, have some way to compare whatever data you capture to a "perfect" swing, or at least good knowledge of golf-swing physics, in order to give advice.

[cde 334]

Wouldn't the 3-axis accelerometer/gyro give you a relative height information?

 

Have a small push button. When the club is at rest, at the golf ball before the swing, the person either holds or presses the button, which zero's out the positions, and then starts graphing as they take their swing. When the club and ball collide, the relative position tells you how high it is from the start position.

 

As for the energy transferred into the ball, the same sensor graph would tell you that (given the weight of the club head and the graphed acceleration, you could figure out energy)

[mbeals 74]

It depends on the scope of what you are trying to do. If you want to do full 3-space tracking, then this becomes equivalent to an inertial navigation system (INS) similar to what they use in airplanes as a backup for GPS.  Since you are integrating the accelerometer data, errors add up quickly and your solution 'drifts'...which is why commercial units are so expensive.  This is also why biomechanics researchers use stereoscopic cameras to do point cloud tracking for this sort of thing instead of trying to use inertial sensors.

 

If you can constrain the geometry the problem becomes a little simpler.  You might be able to use a gyro to track the orientation and rate of change of the club's shaft...which you can use to model the stroke given you make some assumptions about the geometry of the pivots.  You will need to estimate the angular velocities involved to figure out a decent sample rate, then match that to your sensor and micro

[igor 163]

Could consider adding a couple of LEDs to the club strap on module (or paint the club dayglow orange), then video the swing with the command module (cell phone/tablet/whatever).

Then can maybe use video data to help limit the error in accelerometer data.

 

[EngIP 31]

a hobbyist. 3-axis accelerometer, 430 MCU, and possibly a bluetooth module and 3v coin cell battery.. 

 

Someone attempted this with the chronos.

 

http://processors.wiki.ti.com/index.php/Pro_Swing_Analyzer_-_Chronos_Project_Design