Sorry about disappearing for a while -- here's some content I should have posted weeks ago:
We started by testing vibration sensors – I was originally afraid they’d be too sensitive, but they ended up being probably not sensitive enough.
The test we've run so far is to connect the vibration sensor directly to an LED (and to a 3.7v battery through a resistor) on a breadboard.
We were able to get the Fast sensor to light the LED dimly when the breadboard was shaken, but the only way to get the Slow sensor to light the LED was to strike the sensor itself (with a pencil or similar).
Neither sensor would light the LED while walking normally and carrying the breadboard.
An oscilloscope showed approximately 10 activations of the sensor in a 20ms period during moderate activity that would light the LED.
It’s very possible that walking provides fewer activations, or activations farther apart, such that the LED is not visibly lit.
The next step will be to connect the sensor to a digital input and light the LED for 1 second on each activation.
This will show whether the sensor is being activated at all while walking.
If the sensor doesn’t activate normally, perhaps it can be made more sensitive by mounting it on a stalk (a feather?) which might exaggerate motion to the sensor.
Also – a visit to a local costume shop did not yield any acceptable pirate hats. I’ll probably have to order one.