Total Newbie - Wants Wireless Motion Tracking

Hello All,

Please help me to plan this project.

Goal - collect data on human motion.
Constraint1: - keep this all under $200-$300 total in expenses (not just electronics total).
Constraint2: -get data clean enough to not get kicked out of professional meetings.

Needs - be able to keep track of velocities and accelerations of arm during gestures.
Wants - can I get x,y,z locations reliably for short <1 minute gestures if I'm (you're) clever enough?

My current thoughts:

3 wireless motion sensors
each sensor will have: power supply, transmitter, motion sensor, controller as neatly packaged as I can get it.
power supply - old school batteries 2 AAAs for ~3 volts
transmitter - NRF24L01 I can see that xbees would probably be easiest, but a bit pricey, the libraries for NRF24L01 seem to have all the capability I need. I just bought the NRF24L01s and then found out about NRF24L01**+** s
sensor - this MPU 6050 module![]($(KGrHqMOKowFHkLwrv72BR+Gzl8J+w~~60_3.JPG)
controller - arduino nano (clones?) ![]($(KGrHqZHJE8FIHZFCRBgBSH(bqOG6g~~60_3.JPG)

Receiving Station
Arduino Leonardo
1 - NRF24L01 receiving from the three transmitters
serially sending data to a mac.

I'm planning on strapping the sensor packs onto the back of the hand, the forearm, and the bicep, of an individual who is creating and/or imitating gestures.

Some questions I have considered:

I'm doing this wireless so there aren't wires throwing off gestures, but power from heavy batteries might be even rougher. Are there good inexpensive non-traditional battery options given my sensor pack? NRF24L01s are 3 volt as are the 6050 module, and even the nano correct? I have little experience with hobby electronics, so power is not something I know how to think about.

Will I be happy with getting raw data from the 6050? I know how to do filtering etc. offline with software, so in my mind record the raw data and process later right? Or is there something else I should be thinking about putting in the code?

Will a newbie be able to use the NRF24L01s just fine? I'm second guessing myself because I can find only one write-up on using a network of sensors for the NRF24L01, but there are scads of tutorials for xBee networks.

What else am I missing?

Is there another solution?

Any help, encouragement, even criticism will be greatly appreciated.

You might want to look at a Kinect controller.

Thanks big red. I will look into the Kinect further. I don't think it will be for me with this project though. I want to get precise, accurate, and simple data streams (very relative term) to compare actor to imitator. I'm actually a bit of a vision guy, but doing vision to do tracking seems like you might loose all your precision at the start. But I should learn more. Thanks.

I updated this message to remove the humor in case it wasn't appreciated. I really would like any feedback on any portion. Especially power supply and consumption.

Could power to (say) a hand mounted device be fed along a thin cable along the arm from a body mounted battery? One battery could power all devices on one person?

You will probably need to organize your software so that only one device transmits at any one time. I'm using an arrangement like that to send and receive data to model trains.



I will start by making the wireless with batteries .I think I can get nano, mpu6050 / itg , and nrf24l01 to run with pretty low power and maybe use big coin cell battery. But I am thinking that a few small wires for power and data will make the sensor units so much smaller and lighter (leaving out the NRF AND the Nano) that the wired unit may be more natural than the wireless.

I can't wait to get my hands on the things. I still don't have an arduino yet, but I have watched just about every tutorial out there and I'm absolutely chomping at the bit.

If I knew the emotional impact waiting three weeks for delivery from china would have I would have paid the extra 50 or so bucks to local sources.