View Full Version : Summary of responses: Ambulatory system for long-term verticalGRF measurements

12-20-2007, 01:31 AM
Dear all,

Here is a summary of the best responses to my earlier question regarding
an ambulatory system for long-term vertical GRF measurement.

Henri Hurkmans
Dept. of Physical Therapy and Rehabilitation Medicine
Erasmus MC - University Medical Center Rotterdam
The Netherlands


At Hof (University Medical Center Groningen, the Netherlands) suggested
to use an ambulatory system with accelerometers. Using the equation: F_y
= ma_y one can calculate the vertical GRF from the vertical acceleration
of the body center of mass (CoM) and the body mass. Important is that
the accelerometer has to placed near the CoM.
I attended a presentation at the American Society of Biomechanics
conference in 2006, and one of the keynote speakers was Professor
Homayoon Kazerooni from UC Berkeley. He spoke about a robot exoskeleton
he designed, and I am pretty sure he said the major factor used in the
control of the exoskeleton was VGRF. The exoskeleton could be worn
anywhere and there was good battery life for the whole system from what
I remember. Obviously you would not need the whole system, but the VGRF
part might be of good use. He might be a good source of information:

Steve Leigh
HMSC Graduate Student
3107 Bondurant Hall
Division of Physical Therapy
RSSCAN (rsscan.com) and Tekscan do insole systems but they use a local
datalogger for remote use that only logs for a few seconds. The RS Scan
logs for 8 seconds and uses a 4mb logger.
Perhaps the insole system could be configured to use radio telemetry and
log to a remote PC in the subjects home.
This would create a large data log IE 500Hz Scan / sample rate and 4
sensors per cm^2 (maybe 1 multiplexor per 4 sensors I don't know) One
foot's surface area is about 250cm^2 so that's max 2 * 62.5 * 500/sec.
That's about 2gb per 8hrs.That = 20gb data in 5days. Phew! That is a lot
of data to process. (could be more if the 4mb logger is filled in 8
Maybe you could use individual sensors on the areas of interest of the
foot. Maybe 10 sensors and then use a DAQ unit to collect data with
radio telemetry. This would cut down the data file size. Perhaps you
could configure to only record peak pressures etc.

Dave Smith
18 Radnor Pk Ave