Re: axes and normalization of orientation events

Regarding the location of the origin, its good that the spec already requires the same origin for the rotation and acceleration axis.
With gyros and accelerometers in the same small package, this is easy to accomplish.

With rotation, acceleration will be different at different points in the device.
Knowing the relation between rotation and acceleration is important for dynamics.  Coincident origin is best.
For purely translational motion, acceleration will be the same everywhere and rotation will be zero everywhere.
For some motions that include rotation, linear acceleration will be zero at a point but non-zero everywhere else.

Specifying the rotation and acceleration origins to be coincident is important to the inertial navigation use case. 
It would be best to specify the rotation and acceleration axis origin to be at the center of mass of the device.
Not sure if that is common practice now.

It's recognized that NED was considered and rejected by the editors in lieu of ENU.
To avoid mistakes in use with maps, it was important to include the comment: "...alpha is in the opposite sense to a compass heading."
To avoid mistakes in inertial navigation uses, it would be important to add a comment that "the angles do not match roll-pitch-yaw."

Summary of recommendations:
- retain approach of common origin for rotation and acceleration.
- add that the origin of the device coordinate frame is to be at the center of mass of the device.
- add a comment that "the angles do not match roll-pitch-yaw."

George




On Jun 28, 2011, at 10:44 AM, George Percivall wrote:

> Steve,
> 
> As you say, location of the origin is important for location determination and not as obvious for rotation.
> 
> In 2D, rotation angles in a plane is equal for every point in the plane - so the origin is irrelevant for rotation in a 2D plane.
> 
> Not sure its that simple in 3D.  I have to think/review that case before commenting.
> 
> George
> 
> 
> 
> On Jun 28, 2011, at 5:42 AM, Steve Block wrote:
> 
>> Thanks for the comment George.
>> 
>>> Its also important to specify the location of the origin of the coordinate
>>> reference system (CRS).
>> I'm not sure why you think this is important. The DeviceOrientation
>> spec makes no mention of position - only orientation and acceleration.
>> 
>> If your comment instead relates to the Geolocation spec, can you
>> provide a concrete usecase where specifying the location of the origin
>> in the body would be an advantage?
>> 
>> Thanks,
>> Steve
>> 
>> -- 
>> Google UK Limited
>> Registered Office: Belgrave House, 76 Buckingham Palace Road, London SW1W 9TQ
>> Registered in England Number: 3977902
>> 
> 
> George Percivall
> gpercivall@opengeospatial.org
> @Percivall on Twitter
> +1-301-560-6439
> http://www.opengeospatial.org/
> OGC -- Making Location Count
> 
> 

George Percivall
gpercivall@opengeospatial.org
@Percivall on Twitter
+1-301-560-6439
http://www.opengeospatial.org/
OGC -- Making Location Count