P10010: Motion Tracking Sensor

Sensor Research - Add links

Table of Contents


Nike Sport Accelerometer-

Vendors/Pre-Existing Systems

Found this blurb.. biomechanics, focuses on accuracy.. but no links

Looks like UNB does research on biomechanics/tracking, would be a good contact

Additionally looks like they are licensing out the technology. POC- Chris Mathis from UNB

PDF for a project- wearable wireless motion sensor for use in medical care

Inertia Cube Sensors Inertia Sensors


This one uses motion captures..not our interest but has some good point we can use for our paper, etc. http://www.scribd.com/doc/12393531/Tracking-Human-Motion-by-using-Motion-Capture-Data



Related papers (mostly for my reference): http://www.vuse.vanderbilt.edu/~bobbyb/pubs/dealingdata97.pdf.


I think flex sensors/ potientometers are the way to go. We can design a simple ciruit ourselves. The sensor costs about $10, small, highly light and portable. We have to figure out how to convert the resistances into angles. Example: http://www.google.com/patents?hl=en&lr=&vid=USPAT5295704&id=WsseAAAAEBAJ&oi=fnd&dq=%22Ski+binding+with+knee+flex+sensor%22&printsec=abstract#v=onepage&q=&f=false

My Favorite Concept: FLEX SENSOR http://mech207.engr.scu.edu/SensorPresentations/Jan%20-%20Flex%20Sensor%20Combined.pdf

Rotary potentiometers.

Rotary Accelerometers.

Example: Apls RDC50 Series Rotary Position Sensors. http://www.alps.com/WebObjects/catalog.woa/E/HTML/Sensor/Position/RDC50/RDC50_list.html


XSENS/Xbus Kit

XSENS / Inertial Motion Capture

InterSense NavChip - Also other products available


Several Motion Tracking Systems Existing in Market Today (Several that were mentioned by Sara, i.e. WASP, Flock of Birds)

Research Papers Specific to Measuring Lumbar Regions

System Used in the Above Paper

(Both use the Polyhemus System that I showed everyone last friday. Both systems are electromagnetic, which Sara said would not work, might need to look into the reason why, also I may call Polyhemus over break to discuss how their systems calculate angles and discuss our application)

3DM-GX1 Sensor

All About Accelerometers

All About Gyroscopes

More Results Found For "3 axis Sensor", Rather Then "Motion Tracking Sensor", More Discussion Needed.


LIS302DL from STM (the iPhone accelerometer):

Here is a link about accelerometer applications and it talks about the ipod and wii accelerometers

This is a site know for cheap and high quality custom elecronics (sparkfun), and the sensors category:

This is the KinetaMap GPS and Accelerometer logger with 1Gb flash memory (since the package is pretty big, we could just use the sensor inside - big price tag though):

Atomic IMU 6 Degrees of Freedom - XBee Ready:

Ardupilot Sensor Board - Six Degrees of Freedom:


I'm not sure if maybe our definitions of sensors are different, but I was thinking something along the lines of just hopping on Mouser or Digi-Key or whatver and buying a few different Accelerometers/Gyroscopes/etc. that would work with most microcontrollers out there. Of course, there are different kinds of sensors -- some with analog output, others with digital output. Obviously the analog output sensors are going to be more accurate, but would require processing through either an external ADC or an ADC internal to the MCU. I think we would then need to consider the number of sensors we would require and purchase an MCU based on the number of digital/analog ports we need. If we don't care about the MCU and the goal is to just come up with a list of sensors that would work for _________ function then a generic MCU with a handful of digital IO pins and a handful of analog IO pins should be sufficient -- all we're showing is proof of concept.

That said, here is an example sensor: http://www.mouser.com/ProductDetail/STMicroelectronics/LIS344ALH/?qs=sGAEpiMZZMsB9HsreUc%252biQ5WkouSnZMF

I just clicked on a random sensor, so don't worry about actual VDD/GND/output voltages. At any rate, take a look at the datasheet. This is what a standard analog accelerometer will look like -- three output voltages, one for each axis. Interfacing these with an MCU is the easy part. What requires a bit more work is sampling each accelerometer's XYZ axis data at whatever sampling rate we choose and performing various simple geometric calculations based on the known sensor locations relative to each other. This data can then be written to the SDcard or whatever storage medium available on the MCU.

I am not saying it will be as clear-cut and simple as it sounds above, but that was my impression of what we were going to do. Feel free to correct me if I am wrong, though!