headers
joelmeow | 5 Dec 2004 18:08
Picon
Favicon

Quick question


Hello,

I've been looking at the site and the contest, and a question popped
up when viewing the matches:
Why are the various rounds between the same two robots not identical?
 Where is the non-determinism coming from?

If a particular robot has a non-deterministic controller, that would
explain it.  But I wouldn't expect everyone's controller to be that
way.  Looking through the documentation, I see that the GPS sensor is
noisy, so if people are using that heavily, that could result in some
differences as well.  Are all the sensors noisy?  Are the various
controls noisy as well, or are they deterministic?

If anyone could shed some light on this, I'd appreciate it.
Thanks,
-Joel

------------------------ Yahoo! Groups Sponsor --------------------~--> 
$4.98 domain names from Yahoo!. Register anything.
http://us.click.yahoo.com/Q7_YsB/neXJAA/yQLSAA/pU_rlB/TM
--------------------------------------------------------------------~-> 

 
Yahoo! Groups Links

<*> To visit your group on the web, go to:
    http://groups.yahoo.com/group/webots-contest/
(Continue reading)

Permalink | Reply | 1 comment |
headers
Olivier Michel | 6 Dec 2004 16:20

Re: Quick question


>Hello,
>
>I've been looking at the site and the contest, and a question popped
>up when viewing the matches:
>Why are the various rounds between the same two robots not identical?
>Where is the non-determinism coming from?
>  
>
Various things:

(1) The sensor measurement are noisy.
(2) The physics engine rely on random functions and is not reinitialized 
before each match.
(3) The robot controller themself may use randomness.
(4) The fact that the controller processes are running in parallel 
asynchronously with the simulator process and they sometimes miss some 
sensor information if they consume too much CPU cycles.

>If a particular robot has a non-deterministic controller, that would
>explain it.  But I wouldn't expect everyone's controller to be that
>way.  Looking through the documentation, I see that the GPS sensor is
>noisy, so if people are using that heavily, that could result in some
>differences as well.  Are all the sensors noisy?
>
Yes, except for the camera.

>Are the various
>controls noisy as well, or are they deterministic?
>
(Continue reading)

Permalink | Reply | 0 comments |

Gmane