[robocup-humanoid] [Fwd: About the rule]

[Shibatani Naoki]

Dear Mr. Oskar von Stryk, Dear all

Hello, thank you for sending reply.

>> - 19.1 Obstacle Avoidance:
>> there is possibility that the open-loop walking robot attain the
>> gaol(achieve the task).
> I do not share your concern, because I think it is very unlikely.
> But I would also agree to some modifications to increase
> the difficulty for open-loop programmed robots.

I think the open-loop programmed robot
that dash across the field has a great chance to get high score.
And the robot that uses vision sensor would take a certain time to achieve 
this task,
so this robot has disadvantage in the current scoring system.

>> c.Make this TC harder
>>  The goal means the robot touches the pole in the goal area.
>>  (Now in the rule, the goal means the robot enter the goal area.)
>
> A pole in the goal area which is to be touched may be very difficult to
> distinguish from a pole that must not be touched.
> The poles are supposed to imitate other robots.

How about the goal means the robot touches the ball in the goal area or
kicks the ball to the goal?
I think it makes Obstacle Avoidance more difficult for the open-loop 
programmed robot.
And it seems like more soccer.

Thank you very much for your attention.

Sincerely yours,
 Naoki Shibatani (TeamOSAKA)


----- Original Message ----- 
From: <stryk at sim.tu-darmstadt.de>
To: <robocup-humanoid at cc.gatech.edu>
Cc: <shibatani0120 at hotmail.com>
Sent: Friday, January 05, 2007 8:15 PM
Subject: Re: [robocup-humanoid] [Fwd: About the rule]


> Dear Naoki Shibatani, Dear Teams,
>
>> - 19.1 Obstacle Avoidance:
>> there is possibility that the open-loop walking robot attain the
>> gaol(achieve the task).
>> (It might be no vision sensor needed for achieving this task.)
>
> The aim of the obstacle avoidance challenge is to develop
> skills for obstacle avoidance using vision.
> The setup should be in a way that it is very unlikely to solve the
> obstacle avoidance challenge with offline programmed robot.
>
> I do not share your concern, because I think it is very unlikely.
> But I would also agree to some modifications to increase
> the difficulty for open-loop programmed robots.
>
>> For countermeasure,
>> a.The referee puts the robot at the start position.
>
> The start position could be modified from the center mark
> to any position within a radius of 20cm (as the ball location in penalty
> kick) around the center mark.
>
>> b.Limitting the number of retire.
>>    -> When the team call "retire", the referee change the position of the
>> obstacle poles.
>>         So the team might call and call until the pole positions are as
>> they
>> want.
>
> The total time for the three technical challenges is limited to 25 
> minutes.
> Before a team can call for a retry of a challenge they must first make
> serious attempt to solve the challenge before it can be retried.
> And this takes time. So there cannot be too many retries, I think.
>
>> c.Make this TC harder
>>   -> The goal means the robot touches the pole in the goal area.
>>        (Now in the rule, the goal means the robot enter the goal area.)
>
> A pole in the goal area which is to be touched may be very difficult to
> distinguish from a pole that must not be touched.
> The poles are supposed to imitate other robots.
> So the robots should learn to avoid them by walking around them using 
> vision.
>
> But referee training for this challenge will be important to make
> it work well. (A similar challenge has been used successfully in the
> middle size league previously.)
>
> With best regards,
> Oskar von Stryk
> www.dribblers.de
> _______________________________________________
> robocup-humanoid mailing list
> robocup-humanoid at mailman.cc.gatech.edu
> https://mailman.cc.gatech.edu/mailman/listinfo/robocup-humanoid
>