View Single Post
Old 09-11-09, 09:08 AM
  #14  
rdtompki
Senior Member
 
rdtompki's Avatar
 
Join Date: Nov 2008
Location: Hollister, CA
Posts: 3,957

Bikes: Volagi, daVinci Joint Venture

Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Likes: 0
Liked 3 Times in 2 Posts
Originally Posted by WebsterBikeMan
...but it would appear that they use a MEMS accelerometer and calculated acceleration (from the wheel sensor) to compute actual acceleration and tilt. They also use two air pressure sensors, one with a port facing forward and another to the side to measure wind speed (this isn't right, but the error could probably be calibrated out). The main remaining input is cadence, measured conventionally. Optional other measurements include altitude and temperature, which appear to be used either to improve accuracy in other measurements or to fine tune things. A roll-down calibration followed by a curve fit allows a measurement of rolling resistance and air resistance. ...
Thanks for the information. You're describing what I would envision their approach. Devil is in the details in terms of how the various measurements are combined, weighted, filtered ,etc. It's surprising that their website is more of the "trust me" variety. When I get some time I'll look up the patents; might be interesting.
rdtompki is offline