Controlling A Stopclock Via Accelerometers

[jmbelectrical]

Hello all,

My colleagues and I are trying to design a system in which a stopclock can be started and stopped by hand gestures. Using accelerometers fixed to the person's body, lateral (x-y) hand movement would be used to stop the clock and vertical (z-axis) hand movement would be used to start the clock. My questions are as follows:

1. Does this system, or something even remotely similar to it, already exist on the market?

2. Would it be easier or more practical to design and assemble a stopclock of my own, or purchase one that's already built and somehow tie it in to the accelerometers?

3. What type of components would I need between the accelerometers and the stopclock? A microcontroller of some sort?

4. Are there some other type of sensors that could be employed other than accelerometers?

Thanks for your help. I appreciate any input offered.

 

[itsmoked]

Is this for school?

Keith Cress
kcress -

http://www.flaminsystems.com

 

[MikeHalloran]

It could already exist as an iPhone app.

Mike Halloran
Pembroke Pines, FL, USA

 

[IRstuff]

It sounds consistent with apps that use shaking as a means of signaling a scroll.

However, such an application will be horrifically inaccurate and unrepeatable, and, prone to erroneous false alarms. When races are decided by 10 ms differences, a 60ms activation time is a bit long. Activation of a typical stopwatch is probably repeatable to within 20 ms.

If you're still insistent on doing it, then using an Android pad or phone would be the most plausible approach, since you probably already have one. You would need to download the Android SDK, and look at any open source app that accesses the accels.

TTFN
faq731-376

 

[VEBill]

OP didn't specify any start/stop timing precision criteria. As noted, triggering in this arm-waving manner is far less clearly defined than activating a snap action switch. It'll be ± about a second.

The interfacing of the complicated arm-waving acceleration measurement system to a stopwatch is the final (extremely simple) 1% of the project.

Activation of a typical stopwatch is probably repeatable to within 20 ms.

Reminds me of a contest the kidiots and I played with a stopwatch: capturing the shortest elapsed time. After much practice, I won with 0.04 seconds (40ms).

 

[jmbelectrical]

itsmoked: Nope. It's an independent project.

IRstuff/VE1BLL: The intended application is limited to football, basketball, or some other sport where such high level of precision is not necessary. I hadn't even thought about looking at Android applications. I'll be certain to check that out, though.

 

[IRstuff]

I think that the only sports where imprecise timing is acceptable are lawn darts, curling, and the like.

Offhand, I see this as a solution looking for a problem. Stopwatches are absurdly cheap, and there are an absurdly large number of free stopwatch apps in the Android store (1654 to be precise), so where's the fire?

TTFN
faq731-376

 

[itsmoked]

If an Android or Wii can not be made to work, then yes, you'd likely want a processor so you can tailor the motion detection to whatever you're doing. There aren't any other sensors that would be as effective as a multi-axis accelerometer for what you've described.

Don't know what you're doing but you might look into Kinect. They're pretty amazing and would deduce your motions from a distance.

Keith Cress
kcress -

http://www.flaminsystems.com