Re: [Orocos] Common robotics data exchange format...?

Some weeks ago, I posted to this mailinglist to suggest the beginning of
a common data representation standardisation. This post is a follow-up,
summarizing what has happened since then, and listing some concrete work in
progress.

First of all, I decided to go forward with the effort, because I received a
lot of positive feedback. So, the first drafts resulting from this effort
are posted in follow-up mails.

The Player project has kindly offered to host the detailed discussion on
their mailinglist:
  <http://sourceforge.net/mailarchive/forum.php?forum=playerstage-design>
reachable via <[email protected]>.
Please, use only that mailinglist for feedback and suggestions, such that
we have a single archive of all arguments and suggestions.

The standards will be public domain, or another open form; what exactly has
not been agreed on. My current drafts contain the following statement:
"
  This document is released in the public domain. The presented standards
  can be used without any restriction, except that the names "BROS" and
  "Basic Robotics Standards" can only be used to refer to the material in
  this and related documents, and in later versions thereof.
"
Maybe the first sentence is already enough. Because I guess we would need
some sort of trademark protection in order for the second sentence to be
valid...?

There is more or less (with much more "more" than "less") unanimity about
the need for multiple levels in the standardisation: the strategy is to
start with small-scale standards that all projects agree on and about
material that all projects are using, while higher levels of the standards
are less universal.

In analogy to the similar (and successful) approach in linear algebra (the
BLAS - Basic Linear Algebra Subprograms <http://www.netlib.org/blas/>), I
coined the term BROS - Basis Robotics Standards. The levels I will report
on in two follow-up mails are:

BROS-I: Primitive data structures (numerical representations) for motion
         (position, velocity, acceleration), force, and derived quantities.
BROS-II: Interconnected robotics objects, with meta-information.

(Higher BROS levels will standardize network transparant middleware, and
method calls. The levels I and II are only about exchange of data.)

My suggestion is to focus on BROS-I, and the BROS-I draft proposal is
worked out in detail, at this moment. I also made a BROS-II draft, with
less level of detail, but in order to make clear what kind of things are
not part of BROS-I. Both drafts are to be considered as _proposals_ only,
so be critical, but always in a _constructive_ way: if you do not like
something, don't just rant about it but explain how to do it better.

I hope we can go forward with this effort in the same constructive way that
we have experienced the last two weeks.

Finally, I invite a representative of each of the projects involved to
co-author a paper about this initiative for ICRA'05. I will go forward with
this paper as soon as there are at least three projects that want to do
this.

Best regards,

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


-------------------------------------------------------
This SF.Net email is sponsored by BEA Weblogic Workshop
FREE Java Enterprise J2EE developer tools!
Get your free copy of BEA WebLogic Workshop 8.1 today.
http://ads.osdn.com/?ad_id=5047&alloc_id=10808&op=click