[Tinyos-help] Analysis of Accelerometer reading

[Michael Schippling]

One of my brilliant QnD analyses indicates that you might be able to resolve
in the neighborhood of .005g acceleration (the specs say .002 and it's
always prudent to double). I don't know the G-forces of which your typical
mother is capable however. I'd guess that you're in the ballpark but only
experiment can tell really. There might be too much noise in the signal
to make a high confidence determination.

You'd probably want two sensors to get a 3-d signal.
The analysis software could be an interesting adaptive application...

Even though xbow says the accel is available on the sensor boards, it's
not standard and I've only seen minor mention of them on this help-list.
So you may be mostly on your own in getting something to work. Got any
bio-css students who need thesis topics? It's more robotics than bio-
informatics, but you might trick someone into a career divergence.

You might also think about video image analysis as an alternative...
MS

Kathy Stansbury wrote:
> Hi – If I put this sensor on a baby, would your software be able to help 
> determine whether a crying baby remained in the bed crying or was picked 
> up and/or rocked?
> 
>  
> 
> Thanks,
> 
>  
> 
> Kathy
> 
>  
> 
> Kathy Stansbury, Ph.D.
> 
> Associate Professor
> 
> Department of Psychology
> 
> Morehouse College
> 
> 830 Westview Drive SW
> 
> Atlanta, GA 30314
> 
> kstansbu at morehouse.edu
> 
> 404.681.2800 x2796
> 
> 
> ------------------------------------------------------------------------
> 
> _______________________________________________
> Tinyos-help mailing list
> Tinyos-help at Millennium.Berkeley.EDU
> https://mail.millennium.berkeley.edu/cgi-bin/mailman/listinfo/tinyos-help