[robocup-small] robocup-small Digest, Vol 67, Issue 5

Tue Sep 28 05:09:19 EDT 2010

Hello Dear All

I really believe that field's expansion and 
more robots can improve the challenge and I also think that adding two 
robots is a good idea because we saw in Robocup2010 that robots are 
really faster than before and they can pass the ball to each other 
accurately. As a second, According to robots velocity and acceleration, 
kick power and chip kick ability, the field should become larger. For 
Example, robots can shoot the ball from their goal area to opponent's 
goal with chip kick and also they can kick the ball more than 10 m/s 
that takes less thank one second to traverse the field.

Thank you so much

with kind regards;

Jean-Jacques 

> From: robocup-small-request at cc.gatech.edu
> Subject: robocup-small Digest, Vol 67, Issue 5
> To: robocup-small at cc.gatech.edu
> Date: Mon, 27 Sep 2010 12:00:01 -0400
> 
> Send robocup-small mailing list submissions to
> 	robocup-small at cc.gatech.edu
> 
> To subscribe or unsubscribe via the World Wide Web, visit
> 	https://lists.cc.gatech.edu/mailman/listinfo/robocup-small
> or, via email, send a message with subject or body 'help' to
> 	robocup-small-request at cc.gatech.edu
> 
> You can reach the person managing the list at
> 	robocup-small-owner at cc.gatech.edu
> 
> When replying, please edit your Subject line so it is more specific
> than "Re: Contents of robocup-small digest..."
> 
> 
> Today's Topics:
> 
>    1. Re: robocup-small Digest, Vol 67, Issue 1 (Stefan P)
> 
> 
> ----------------------------------------------------------------------
> 
> Message: 1
> Date: Sat, 25 Sep 2010 19:51:41 -0400
> From: Stefan P <stefan.posey at gatech.edu>
> Subject: Re: [robocup-small] robocup-small Digest, Vol 67, Issue 1
> To: Mustafa Talaeezadeh <brother.t at live.com>
> Cc: Omid Bakhshandeh <omidtnt2004 at yahoo.com>,
> 	robocup-small at cc.gatech.edu
> Message-ID:
> 	<AANLkTik9EeP9gzkA8ee0=trHK3HT4uUSpnYUzWE1Km5G at mail.gmail.com>
> Content-Type: text/plain; charset=ISO-8859-1
> 
> All,
> 
> Below are some of my thoughts on how to increase the field size and
> help new teams. Note my primary area of work on robocup is in
> mechanical design and integration, so my familiarization with the
> challenges imposed on software aspects is somewhat limited.
> 
> Team Size & Game Play Dynamics
> 3 robots per group and 3 groups per team. This would give 9 robots
> total and add the ability to conduct voting on game play actions. For
> example who to pass to. Each group could also veto actions involving
> their own members. For example if group 1's robot number two has a
> broken component and can't kick, but can still guard or block. Each
> group on a team could focus on one element of game play (scoring,
> defense, etc). This would mirror the way humans tend play a certain
> position / focus on certain tactics. Some teams already clearly focus
> on defense or offense and this would allow for progression of
> strategy. Game play advancement would be forced since groups would be
> have to collaborate on it to successfully compete.
> 
> Simplified Decision Communication Example
> I'm using passing as an example. This wont work for every type of
> action, and this exact method might not be the best.
> 1. Each group submits a list of who they want to receive the ball.
> 2. Each position a robot appears on a the list it gets some points
> (1st gets 9pts, 2nd 8pts, etc)
> 3. The one with the highest rank is picked to receive the pass.
> 4. Robot capability verified with who it belongs to. Fault could be
> detected with encoder, break beam performance, etc.
> -- This in a way gives each robot on the field some say in how the
> game is played.
> 5. Pass attempted or robot next in line is passed to.
> Note: Assigning points for a vote still works with 2 groups, but there
> are most likely some situations where assigning points wouldn't make
> sense. And Yes / No is best. This only works with 3 groups.
> Additionally, I'm not implying that voting be conducted on every game
> action.
> 
> Common frame work for communication with in teams
> This would be needed and I think is natural / necessary since humans
> on the same team communicate (verbally, signs, and body language).
> Robot players would need something similar. We have a network at each
> field already so adding this as another use wouldn't require any
> infrastructure change. One requirement is that the opposing team
> should not be allowed to see what the other team is trying to do.
> 
> For Clarification
> - Each team would be yellow or blue
> - Each group would be RoboJackets, OMID, CMDragons, etc.
> 
> How to do Round Robin
> Increase the number of matches. We can do this as there is plenty of
> time at the competition to do it. With new vision system many teams
> now utilize laptops to run their systems and are able to move from one
> field to another. As this was done in the afternoons in Singapore for
> testing / scrimmage matches among teams. Each team gets 2 points if
> their side wins, 1 point for tie (only if scoring is involved), and
> tie breakers in ranking could be done by looking at total or average
> points scored per game. In theory if you have enough matches the best
> groups would be filtered out for eliminations as they would move
> toward the top of the pool. I think this is possible given the current
> number of teams in the league and the amount of down time experienced
> during the competition.
> 
> How to do Finals
> Teams would be formed from each of the highest ranked groups from
> their respective "home field". The top 3 from each field would form a
> team that would stay constant through eliminations. You need at least
> 4 fields to do this. Home fields would be determined the same way that
> groups are presently split up in to fields. What ever the current
> system is each field has a good sampling of teams (both in terms of
> global location and capability).
> 
> Removing Barriers for New Teams and Growing The League
> - We would be decreasing the number of robots each team is required to have
> -- This has two effects:
> --- Maintenance concerns are reduced (if you still build 6 robots you
> will be more than prepared with spares)
> --- A new team's cost to build a fleet are reduced by a 1/3 (assuming
> they build 4).
> - We are allowing teams to only have a predominantly defensive
> strategy if they choose
> -- Teams should state if they are in fact limited to only this as part
> of qualification and only a certain percentage of teams should be
> allowed in with this status.
> - As a league we should develop a common best practices document,
> wiki, or something
> -- We have received a few emails asking for help and advice
> specifically with omni wheels and I general have given each the same
> advice. (Apologies if I did not reply to your email.)
> -- This document doesn't need full design drawings and I don't feel
> that would be the best way to approach it as everyone has different
> tools. You design based on what you can build with what tools you
> have.
> -- This could save new teams months off of development time as we have
> literally spent months on such things as:
> --- Whats the best rubber to use on omni wheel rollers, on the
> dribbler, whats needed to make a safe & reliable kicker circuit, how
> big do the traces on the boards need to be to do X, etc...
> -- Many of these small details are never covered in TDP's (as that is
> not their intended purpose) and the only way to gain real insight is
> to go to the competition and talk to people with their robots (this
> something that has been very valuable to our team). This is too late
> for new teams. While email works there is only so much you can
> remember to tell people and they are only talking to one person.
> - With a larger filed a suggested reference/test field should be
> standardized to give new teams a defined environment.
> -- For our team getting a larger space at our school to do testing is
> both a political and monetary impossibility.
> -- New teams will have a harder time finding space to test than any of
> the current teams.
> -- I think that the current field is a good candidate for this or
> maybe a field that is missing 1/2 or 1/3 of the length.
> 
> Ways This Advances SSL and Other Improvement Thoughts
> - Reinforce the necessity for low latency system performance
> -- When we started, just handling vision in a timely matter was hard
> for our team. Powerful multicore CPUs and GPUs now exist in a laptop
> form factor.
> - Force real mixed teams, and multiple decision agents.
> - The ability of teams to focus on specific aspects of game play.
> - Move to 4 1080p cameras with a move to a larger field?
> - Develop standard field construction documents. The ones in Singapore
> were pretty nice. The foam backed goals were interesting. Do drawings
> and spec's exist for the fields if so could they be published?
> - If vision could handle 4 cameras we could test in more places as the
> cameras could be lower to the field. The high ceilings that 2 cameras
> require at present is hard to find. I've been told that this
> functionality is somewhat present / easy to add. I think this
> functionality being fully implemented would add some much needed
> flexibility for everyone.
> 
> Thanks,
> 
> PS - If requested I will post this on our wiki in a more readable format.
> 
> --
> Stefan Posey
> RoboCup Mechanical Lead, RoboJackets
> http://www.robojackets.org
> 
> On Sat, Sep 25, 2010 at 6:01 AM, Mustafa Talaeezadeh <brother.t at live.com> wrote:
> > Dear all,
> >
> > ????I think we all agree on the expansion of the field size. The thing is,
> > the purpose of the SSL is teamwork, with realistic physiques involved, so,
> > Dear Omid, I think the more preparation time for "joining" (instead of
> > mixing) is in contrast with the main purpose; the robots should play their
> > tactics IN COMBINATION with others, not alongside, nor dominated by.
> > ?? The latter matter, the number of robots; Though increasing the number of
> > robots helps integrity in tactics, the matter of maintenance is still
> > causing issues, both financial and technical, specially "newbies". So I
> > think we should put our heads together to find AN OPTIMUM number of
> > robots?of each team to be shared in?a mixed-team,?and that can be?tested in
> > local events' Technical Challenges and feedbacks sent to the Community and
> > the Committee.
> > ?? I myself, regardless of my respectful fellows, suggest that each mixed
> > team be formed as a 12-robot combination of 2 discrete teams,?with
> > some?FLEXIBLE?limitation for each team; by that I mean if a team have 5
> > robots to share instead of 6 (= 12 / 2), the other team is allowed to put 7
> > robots in the mixed team; but all teams should AT LEAST have 3 robots in the
> > field, that means?NO team can have more than 9 robots, and the mixed team
> > that have less than 6 robots in the field PERMANENTLY (being fixed alongside
> > the field and the game,?and timeouts are excepted),?will be OUT, 10 to 0?in
> > the favor of the opponent, if both mixed teams?hit the criteria, the TIE,
> > 0-0 will be put as the result.?By these, the matches?keep their toughness
> > and still?welcome?newbies.
> >
> > PS: Practically, less than 3 robots?mans a?2-2 junior soccer?team with
> > vision that is provided for them dedicated!, and joint AI's!
> >
> > Peace and Love
> > Mustafa Talaeezadeh Khousani
> >
> > From: Omid Bakhshandeh
> >
> > Sent: Saturday, September 25, 2010 10:30 AM
> > To: Mani Monajjemi
> > Cc: robocup-small at cc.gatech.edu
> > Subject: Re: [robocup-small] robocup-small Digest, Vol 67, Issue 1
> > Hi,
> >
> > Dear Mani I think increasing robots up to 7 or 8 is possible, now teams in
> > SSL have at least 1 robot as backup.Mixed team is a good idea if teams have
> > time for preparation with each other.
> >
> > Best Wishes
> >
> > ________________________________
> > From: Mani Monajjemi <mani.monajjemi at gmail.com>
> > To: Omid Bakhshandeh <omidtnt2004 at yahoo.com>
> > Cc: robocup-small at cc.gatech.edu
> > Sent: Fri, September 24, 2010 10:03:25 AM
> > Subject: Re: [robocup-small] robocup-small Digest, Vol 67, Issue 1
> >
> > Hi all,
> >
> > From my point of view, increasing the field size is a good idea,
> > however increasing the number of robots per team will make the SSL
> > harder for new-comers to join. In addition, extending the number of
> > robots for many currently active teams might be so expensive, besides,
> > maintenance of? for example 10 robots is much more challenging and
> > expensive than maintenance of 5 or 6 robots.
> >
> > I think it would be better if we consider mixed-team competitions as a
> > solution to put more robots on a expanded field.
> >
> > Regards,
> > Mani Monajjemi
> > Parsian Member
> >
> > On 9/23/10, Omid Bakhshandeh <omidtnt2004 at yahoo.com> wrote:
> >> Hi dear all,
> >> What is your idea about changing the field size after 3 years and add more
> >> robots? to small size game to decrease the distance between real? soccer
> >> game
> >> and small size robots cooperation? I think SSL-Vision will support the
> >> size
> >> ,
> >> all team can test and inform others about the size of field.
> >>
> >>
> >> Best Wishes
> >>
> >> Omid Bakhshandeh
> >>
> >> MRL Member
> >>
> >>
> >>
> >>
> >> ________________________________
> >> From: "robocup-small-request at cc.gatech.edu"
> >> <robocup-small-request at cc.gatech.edu>
> >> To: robocup-small at cc.gatech.edu
> >> Sent: Tue, September 7, 2010 8:30:01 PM
> >> Subject: robocup-small Digest, Vol 67, Issue 1
> >>
> >> Send robocup-small mailing list submissions to
> >>? ? robocup-small at cc.gatech.edu
> >>
> >> To subscribe or unsubscribe via the World Wide Web, visit
> >>? ? https://lists.cc.gatech.edu/mailman/listinfo/robocup-small
> >> or, via email, send a message with subject or body 'help' to
> >>? ? robocup-small-request at cc.gatech.edu
> >>
> >> You can reach the person managing the list at
> >>? ? robocup-small-owner at cc.gatech.edu
> >>
> >> When replying, please edit your Subject line so it is more specific
> >> than "Re: Contents of robocup-small digest..."
> >>
> >>
> >> Today's Topics:
> >>
> >>? ? 1. Dear all (KANJANAPAN SUKVICHAI)
> >>
> >>
> >> ----------------------------------------------------------------------
> >>
> >> Message: 1
> >> Date: Tue, 07 Sep 2010 21:20:04 +0700
> >> From: KANJANAPAN SUKVICHAI <fengkpsc at ku.ac.th>
> >> Subject: [robocup-small] Dear all
> >> To: <robocup-small at cc.gatech.edu>
> >> Message-ID: <d5f16dde301f1558d41dce9b9e6ef055 at ku.ac.th>
> >> Content-Type: text/plain; charset=UTF-8
> >>
> >>
> >> Dear all,
> >>? ? It has been 2 months after robocup 2010, Singapore.? Before TCs start
> >> to
> >> discuss about Robocup 2011 rule, we would like to welcome all of you send
> >> any comments or suggestions to us in order to make next Robocup better.
> >>
> >>
> >> Cheer!
> >> KANJA
> >>
> >>
> >> ------------------------------
> >>
> >> _______________________________________________
> >> robocup-small mailing list
> >> robocup-small at cc.gatech.edu
> >> https://lists.cc.gatech.edu/mailman/listinfo/robocup-small
> >>
> >>
> >> End of robocup-small Digest, Vol 67, Issue 1
> >> ********************************************
> >>
> >>
> >>
> >>
> >
> >
> > --
> > Mani Monajjemi
> >
> >
> > ________________________________
> >
> > _______________________________________________
> > robocup-small mailing list
> > robocup-small at cc.gatech.edu
> > https://lists.cc.gatech.edu/mailman/listinfo/robocup-small
> >
> > _______________________________________________
> > robocup-small mailing list
> > robocup-small at cc.gatech.edu
> > https://lists.cc.gatech.edu/mailman/listinfo/robocup-small
> >
> >
> 
> 
> ------------------------------
> 
> _______________________________________________
> robocup-small mailing list
> robocup-small at cc.gatech.edu
> https://lists.cc.gatech.edu/mailman/listinfo/robocup-small
> 
> 
> End of robocup-small Digest, Vol 67, Issue 5
> ********************************************

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.cc.gatech.edu/pipermail/robocup-small/attachments/20100928/05021178/attachment-0001.html 