No problem Nathan
I spose you could use an angle sensor/a potentiometer and two units that change angles with respect to each other, but accel or accel+gyro is ideal, i was just pointing out that if you just use an accelerometer, the program will be much simpler and the setup will be much cheaper. BUT as i mentioned earlier, it wont work if your subject is moving around.
The moment you add a gyro, you will be able to still calculate orientation as you move around, but the program will be significantly more complicated. And setup more pricey too
How good are you with programming?
On the sensory side, for the accelerometer all you will have to do is perform some basic functions to convert your accel data input into a net vector.
On the gyro you will need to read the angular rotation rate data (and maybe integrate it to get gyro estimated orientation), and write software to compare it to the accelerometer.
Im thinking you will need software to compare the two values.
eg. If the gravity vector shifts from straight down rightwards at a 30deg angle, and your gyro reads a rightward shift of abt 32 degrees, you will know your subject just turned sideways 32 degrees.
But if your gravity vector shifts forwards by 30degrees, and your gyro registers 0 rotation, it would mean your person has just accelerated forward and not changed position.
You could also take advantage of the fact that gravity is pretty much constant for you, so if you have a accel vector that is like 12m/s/s, then you will know you arent standing still, irregardless of your direction.
The accel+gyro is more difficult and pricey but if you want to take on the challenge you will learn alot on programming, and heck maybe you’ll become a bit of an expert on it and can help me with my project too! Hahaha