Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Storing Groceries #520

Merged
merged 6 commits into from
Feb 19, 2019
Merged

Storing Groceries #520

merged 6 commits into from
Feb 19, 2019

Conversation

johaq
Copy link
Member

@johaq johaq commented Jan 18, 2019

-Remove specific numbers
-Moved hint to DEM rules
@kyordhel kyordhel requested a review from a team January 18, 2019 11:46
a heavy object,
a tiny object, and
an amorphous object.
\item \textbf{Clear area:} The robot may assume that the working area is clear, (i.e. can move slightly backwards for its task).
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I strongly disagree with this change. It is necessary to keep fairness (even distribution among teams) and for logistical reasons. Further, the heavy and tiny objects are necessary due to bonus points (see scoresheet).

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I agree that distribution has to be equal for all teams but I think that is something for the referees to decide on location and is not something that should or needs to be in the rulebook. Why not allow the referees the flexibility to decide: "Hey at this competition no team is really strong in manipulation let's use easier objects for everyone." And also if teams get stronger we do not need to update single numbers from year to year and keep more consistency in the rulebook over time

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Because scoring needs to be consistent among tests. This tests makes pick and place an object automatically worth ~100pts, so it can be taken as measure for assigning scores to other tests. Further, we are implying that each manipulation should not take more than one minute.

Besides, we need to be consistent in the three leagues (or at least between OPL and DSPL). If you leave the decision to the referees, the difficulty will vary between leagues and you have no consistent progression over the years.

Teams know:

  • There are [5, 10] objects to deal with
  • We must move 5
  • 3 are easy to recognize (and grasp?)
  • The rest are a PITA. (← all decision making enters here)

So yes, a state machine can be used, but a more robust decision making will produce better results to maximize scoring. IMHO, that setup is way too complex for a state machine.

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I don't see how this affects consistency among other tests.
Consistency among leagues I understand but don't the referees switch leagues anyway during a competition and talks are with referees of all leagues.
Consistency over the years I agree is lost a little here but definitely not entirely.
I prefer teams know:
-There are several objects on a table
-We get points for moving up to five of them
-Our robot should know which of the objects are easy to grasp for it

Copy link
Contributor

@kyordhel kyordhel Jan 18, 2019

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

up to five

Or maybe at least.

Let's hear other opinions from the @RoboCupAtHome/technical-commmittee

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Copy link
Contributor

@kyordhel kyordhel left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I agree on most corrections and removing some numbers, but we need to keep consistency among tests and, from the referee and OC perspective, the object distribution is important.

Please check the Scoresheet

Copy link
Contributor

@kyordhel kyordhel left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Move an objectsEach object moved or Move an object

@@ -3,7 +3,7 @@

\begin{scorelist}
\scoreheading{Main Goal}
\scoreitem{500}{Move 5 objects next to their peers in the shelf}
\scoreitem[5]{100}{Move an objects next to their peers in the shelf}
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Careful!

\scoreitem{500}{Move 5 means that team won't score unless 5 objects are moved.
\scoreitem[5]{100}{Move an means one object is enough to start scoring bonus points.

I agree on the 2nd one. HOWEVER: move an objects?

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Yeah should be "an object"
I think second option is more in line with scoring in clean up task which seems similar in some ways.

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

:+1

@kyordhel
Copy link
Contributor

I'm OK with this so far, but lets hear other TC members opinion on the removal of the object distribution before merging.

@RoboCupAtHome/technical-commmittee @moriarty @justinhart @awesomebytes your comments, please

Copy link
Contributor

@kyordhel kyordhel left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

We need to resolve the removal of object distribution before merging.

@kyordhel
Copy link
Contributor

@johaq We discussed yesterday the object numbering in the meeting. We came out with the following decision:

  • To keep consistency between local (open) tournaments and the Worldcup, we need to specify a range of objects (e.g. 5-10).
  • It's up to robots to interact with available objects or not.
  • The number of objects available can never be smaller than the number of objects to be scored.
  • The number of objects to be scored is already in the scoresheet and so tests are balanced. There might be variation in the number of present objects but all teams in all leagues shall be evaluated the same. Hence this number remains fixed.

Please do the pertinent corrections so we can merge.

@balkce balkce mentioned this pull request Feb 5, 2019
@kyordhel kyordhel merged commit 4c37f5b into RoboCupAtHome:2019/restructure Feb 19, 2019
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants