Re: SmartSoft / CORBA next release

On Tue, 3 Dec 2002 [email protected] wrote:

[...]
> > That is similar to our real-time core design: the sequencing is
> > "outsourced" to a component, such that the other components need not be
> > coupled at all. The coupling is exactly and only through the sequencing
> > engine. One single point of decision making!
>
> Yes, at least to have a hierarchical system of decision making where
> the parent node configures the space of possible configurations for
> its successors. Since we have bounded resources in robotics, it is
> quite natural to have a hierarchical approach which however does not
> mean that we can have different masters in different situations !

Yes, indeed. I am still puzzled however by how we should/could formally
describe such interfaces, and leave open the desired flexibility to
program dynamically reconfigurable systems, switching "masters",
adding/removing sensor or planning or control modules,...
And not end up with chaos :-)

> CIDL is a step into the right direction, but does not contain anything
> describing timing behaviors e.g.

_Every_ sequencing language becomes a pain as soon as one tries to
specify timing .... Therefore, I like events/state machines a lot more,
because there time is not absolutely present.

Herman

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