Re: A nomenclature question

On Mon, 30 Jun 2003, Leopold Palomo Avellaneda wrote:

> some days ago, Herman wrote:
> 
> > Ok, it's true. Do you recomend some nomeclature? 
> > Just homogeneous transformation matrices. Or the "TCS" (travelling
> > coordinate system), which is however nothing else but normal common
> > sense. In general, I prefer to allow multiple interfaces for the user,
> > and internally store everything in the most singularity-free
> > representation, which are the ones I mentioned above. (or, rather, in
> > a "symbolic" form that describes the geometric relationships with more
> > than just pure coordinates).
> 
> I have thought about it, and about the document  kin&dyn API. If you want to 
> do some nice and efficient implementation, you would need some kind of 
> mathematical representation. And I think that to chose the nomenclature it's 
> a very important thing. I can understant to have several interfaces for the 
> user to intruduce the serial chain, or the kinematic structure as he/she 
> wish, but you will need some kind of representation to operate. It would be a 
> good idea to chosse ONE to implement internally and after to create all the 
> user interfaces? And which one?

Okay, there seems to be some misunderstanding :-) The ones I propose
are outlined above: 
1. homogeneous transformation matrices for the link-to-link
   transitions
2. symbolic representation that describes the kinematic family and
   the conventions used (positive direction, joint limits,
   connectivities, etc.)
3. for each specific family: an efficient and ad hoc implementation of
   the internal kinematics API.

> I have searching nomenclatures and I have found in a thesis:
> http://vrai-group.epfl.ch/projects/cinegen/thesis.pdf that choose the 
> Kleinfunger-Khalil nomenclature that the Travelling Coordinate System. I have 
> tried to find some information about the last one but I'm waiting the book in 
> the library. The implementation of the kinematics and dynamics orocos library 
> what use, or which one will use?

I don't like very much these conventions that _impose_ how you have to
choose reference frames, for the simple reason that people loose too
much time trying to adhere to the convention. While there is no real
need for such "systematic" conventions: any choice of frames on the
links and joints will do, and designers most often have very good
intuition about where they want their joint axes to lie. All the
conventions (including Kleinfinger-Khalil) work with constructions
such as the common perpendicular between two axes, etc. I think this
is not really necessary; it maybe _was_ interesting in the past, when
being able to use only four numbers instead of 6 or 16 was important,
which is not so important anymore with modern systems... When
efficiency is really important, people have always found efficient ad
hoc formulas anyway.

Herman

-- 
  K.U.Leuven, Mechanical Engineering, Robotics Research Group
<http://people.mech.kuleuven.ac.be/~bruyninc> Tel: +32 16 322480