Creating a device to measure the activity levels of slow-walking people

Hi,

I was wondering if I could get some advice about where to start with creating a device to measure activity levels of slow-walking people.

I would like to use an Arduino microcontroller (I've never used one before) along with a triaxial accelerometer (+/- 2 G would probably be good, with pretty high precision because I need to measure the slow walking accelerations accurately). In addition, I would like to have an external memory (such as an SD card) and wireless transmission of data so I can transmit the raw accelerations to a PC. I am also looking to add in a gyrosensor to sense incline of a lower/upper limb and a gps to track location. Finally, I would like to add in an LCD display to display relevant information to a user.

So far, I've heard using an Xbee would be good for wireless transmission of data, but I don't know too much about anything else. I've programmed a different microcontroller before in C along with an LCD display, so I think learning to program an Arduino would be pretty simple.

Any ideas on what kinds of components I should be using to interface with an Arduino microcontroller? Also, I was thinking of getting 2 microcontrollers, one that is very capable (I'm not sure whether to choose the Uno or the Mega or something else) and one that is miniature and a good size for a limb sensor (maybe the Nano). What would be some good choices? I've also read a bit about shields but I don't really understand too much. Some help would be great!

I'm quite new to electronics so I apologize if I'm asking some trivial questions. Thank you!

How slow is slow walking? Accelerometers are VERY poorly adapted to this type of application high precision and accuracy with them is virtually impossible especially attached to something like a human being.

About 0.6 m/s.

And what kind of accuracy do you need?

Sorry, I'm pretty new to this stuff, so I'm not quite sure yet, but I would just like to be able to distinguish an acceleration signal without too much noise. I'm not sure if that helps at all. I was thinking something like Analog Devices' ADXL335. Do you think that would work? At this point I would just like to use the accelerometer for testing purposes to get a feel for how triaxial acceleration signals should look like, using my own movements (and not impaired movements).

Any device will work for your tests as long as it's reasonably priced and you can interface with it. From that initial testing you'll be able to figure out what you can and can not do with it.

The physical construction of what houses the sensors is actually going to be very important, you need to isolate it from as much vibration as possible without causing it to oscillate in a spring system adding more noise.

Thanks for your help !