Re: Extended deadline, new version of Webots & new contest

Olivier Michel <[email protected]> Fri, 02 Apr 2004 15:21:28 +0200
Newsgroups gmane.comp.cybernetics.webot.contest
Organization Cyberbotics Ltd.
Message-ID <[email protected]>
Hi,

Thank you for this interesting feedback! It took time to read but it is 
addressing very important points. Let me comment about them. Although I 
cannot satisfy everyone, I will try to make this contest evolve towards 
a general interest...

(1) New version of Webots

As Mike said, the new version will not change the physics of the robots. 
It only fixes limits bugs with joints, so it is very likely that most of 
the robots will behave exactly the same with the new version.

(2) Increase the number of sets for the top competitors / tie break.

That's a very good idea (and well accepted), so we are going to 
implement it (maybe for the end of this contest as well).

(3) two cameras

Again a very good idea, but that will be for the next contest.

(4) No limits on fall down, scoring based on the time above the fall 
threshold instead of distance from initial position.

This seems to be a good compromise. However, it has to be combined with 
another rule for eliminating robots that don't move: Robots that don't 
move away from their initial position will loose the set after 10 
seconds. This will make boring matches shorter. This will be introduced 
for the next contest.

(5) better tilt sensor

All right. For next contest.

(6) Synchronous mode, deterministic runs

As discussed ealier, this could be possible but causes a problem when a 
robot takes too much CPU time and blocks the whole simulation...
However, I am currently investigating a new idea where we could count 
the amount of CPU cycles of the JVM used by a controller process (this 
is not machine dependant) and use this information to synchronize the 
robot controllers... This looks pretty complicated and I am not yet sure 
how difficult it is to implement...

(7) leagues organization

This seems to be complicated to implement and handle and I don't see a 
clear improvement from such a change...

(8) four pressure sensors under each foot with analog return value

Yes, next contest as well.

(9) speed / acceleration control

This is actually possible with the current version! Use the 
servo_set_velocity() and servo_set_force() to perform speed control and 
torque control.

(10) self collision detection

This will be implemented in the next contest.



I will provide you with a test version of the new release of Webots so 
that you could test that the new release don't break your controller. If 
there is any major issue with this new release we will fix it before 
releasing it.

-Olivier


------------------------ Yahoo! Groups Sponsor ---------------------~-->
Buy Ink Cartridges or Refill Kits for your HP, Epson, Canon or Lexmark
Printer at MyInks.com.  Free s/h on orders $50 or more to the US & Canada.
http://www.c1tracking.com/l.asp?cid=5511
http://us.click.yahoo.com/mOAaAA/3exGAA/qnsNAA/pU_rlB/TM
---------------------------------------------------------------------~->

 
Yahoo! Groups Links

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

<*> To unsubscribe from this group, send an email to:
     [email protected]

<*> Your use of Yahoo! Groups is subject to:
     http://docs.yahoo.com/info/terms/