New rules document! (... kinda)

[Raymond Sheh]

Hi All!


First, thanks to all of you for your patience as you wait for the rules
document! Life has been rather interesting lately and as a result, I've
had way fewer spell slots to assign to this than I had hoped.


Here is a link to the working draft:

https://docs.google.com/document/d/1ta6XHinzvNRzbZgfzwVR5Y_7ua4NhY6w_eBW9HcW3d0/edit?usp=sharing

It incorporates *most* of what has been discussed, both here in the
mailing list and in the sensing/dexterity document. I'm sure there are
things that have been missed but I think I've sat on it long enough and
you all need to have visibility into this.


Now here's where I ask for all'yall to help!

I have turned on world permissions to suggest and comment. If you think
I missed something that we have discussed, please directly add what you
think the wording should be (it'll track the changes so we can see what
was added). Similarly if there's language in there that's old and should
be deleted, please 'delete' it (again it'll track changes and strike it
through). Also feel free to directly suggest improvements (again as a
tracked change, not a comment).

Of course also please feel free to add comments but please keep those to
things that don't make sense as a tracked change.


I'll check back into the document from time to time to commit changes
and/or add to the comment threads. Anything that looks particularly
noteworthy I'll crosspost here to the group. I'm hoping we can open this
up for a couple of weeks and then lock it down and consider it done!


Cheers!


- Raymond

--
https://raymondsheh.org

[Raymond Sheh]

Hi All!

Thanks to everyone who has commented on the rules document over the past
week and a half. We are coming up to the promised "couple of weeks" and
I'm working through the comments and suggested additions.

Here's the link:
https://docs.google.com/document/d/1ta6XHinzvNRzbZgfzwVR5Y_7ua4NhY6w_eBW9HcW3d0/edit?usp=sharing

There are a few things that I think are worth bringing up to the list
here, just in case you haven't been tracking the document.


Magnet:
The point of the magnet was to serve as a stand-in for any type of
sensor that requires precise proximity or sampling. Imagine a robot
carrying a gas detector looking for leaks, or a collection swab looking
to sample a suspicious substance. We picked magnets simply because they
are safe, don't require a consumable (some of you might remember CO2
being used in the past), and can be detected with cheap sensors (a basic
hall effect sensor).

The problem is figuring out how to turn this into a suitable test that
isn't too easy, isn't too hard, doesn't allow for false positives, and
has a meaningful autonomy component. The previous rulebook was very
open-ended about this, stating just a generic magnet somewhere that the
robot needs to poke at. It would be nice to both lock this down and make
this a more useful test.

How about we adjust this test slightly and have three disc magnets,
approximately 25 mm in diameter, North on one side, South on the other,
taped in a line with clear packing tape on a black background (so
they're very obvious visually, making it easier for autonomy), 50 mm
apart, in random orientation. This gives us 8 possible combinations of
North and South (2^3), the same as the number of directions for a single
Landolt-C. In the competition, we could easily prepare 6 strips of black
cardboard, each with 3 magnets on them, representing all 8 possible
combinations (we only need 6 strips because "100" is just "001"
backwards, similarly with "110" and "011"), and swap them in randomly.

Identifying the pattern would be worth 1 point (again the same as a
single Landolt-C). Just looking at the magnets visually won't tell you
what the code is of course, but a compass (a real one or a phone app
that uses the phone's magnetometer) can easily be used to "read" the code.


Autonomy bonus time:
The rulebook from 2023 included the concept of a double timeslot for
autonomy. Given the 4x autonomy multiplier and the larger number of test
lanes, I propose that we remove this double timeslot option, which also
makes scheduling easier.


Mapping:
For the Labyrinth test, we've proposed giving a 4x mapping multiplier
rather than a 4x autonomy multiplier (given that, relative to the other
tests, autonomously driving through the labyrinth is probably less of a
challenge?). Besides, we want to focus this test on mapping and I'm a
bit worried about confusion if we have two multipliers at play. What do
we think?


Please let us know what you think over the next few days, either by
replying in this thread or via comments in the document! Otherwise, I
(think I've) addressed all of the comments in the document. While I
haven't accepted everything, I've taken all of the suggestions into
account (and in most cases tried to incorporate them elsewhere in the
document).


Cheers!

- Raymond

[Gerard Elias]

Thanks Raymond,

I’ll get the team to go through the rule set tomorrow.

Talk soon,

Gerrie

Sent from my iPhone

> On 4 Jun 2025, at 2:04 pm, Raymond Sheh <raymo...@gmail.com> wrote:
>
> Hi All!

> --
> You received this message because you are subscribed to the Google Groups "The Open Academic Robot Kit" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to oarkit+un...@googlegroups.com.
> To view this discussion, visit https://groups.google.com/d/msgid/oarkit/f60462c0-3915-400b-9a1c-e8e984bfa2c8%40gmail.com.

[Philipp Hock]

Hi All, Mr. Sheh, 

We still have a few further questions regarding the rulebook you prepared: 

• In the courses "Linear Rail Inspect" and "Linear Rail Pick," you wrote that "The robot must be completely on the dexterity field to score any points (i.e. no contact with the ground outside the dexterity field)." However, from our perspective, this could be problematic, as it may be difficult for our robot (and likely others) to stand entirely on the wooden board you specified. Would it be possible to amend the rule so that the robot may, for example, stand halfway on the platform or something similar? 

• Additionally, we noticed a potential loophole in the stair challenge rules that you defined, which could lead to unintended scoring. It seems that one could possibly just hang from the edge to gain points. We would appreciate an improved and more clearly defined rule on this point. 

• Since you are placing great emphasis on autonomy this year, we would also like to propose the introduction of an award for "Best in Class in Autonomy." 

• Initially, you mentioned that the three smallest Landolt Cs must be detected. Over the past few weeks, we invested considerable time and resources into making this possible. However, your current rules state that only the 3rd and 4th rings need to be detected. This change is confusing to us, especially since you had previously insisted that the three smallest must be recognized. We therefore kindly ask you to uphold your original statement and allow the three smallest Cs to be counted, as previously communicated. 

• For the upcoming world championship, we would like to build and bring a second, identical robot to have as a backup incase of a mechanical failure or any kind of serious problem. This would allow us to resume operation after a reset if necessary. Since this is not explicitly addressed in the rulebook, we are reaching out to clarify in advance to avoid building and bringing a robot that may not be allowed and risk disqualification. A second identical robot is essential to ensure operational readiness under all circumstances and to guarantee a smooth competition throughout the championship. Anything less would pose an avoidable risk. Furthermore, since other teams can also read this message, they too would have the opportunity to prepare accordingly — for example, by building a second robot as well. 

We look forward to seeing you in Salvador. 

Best Regards
Philipp Hock
CJTec

[Raymond Sheh]

Hi Philip, 

Thanks for your email and for going through the rules so thorougly! Please see comments inline.

On 6/15/2025 4:42 PM, Philipp Hock wrote:

Hi All, Mr. Sheh, 

We still have a few further questions regarding the rulebook you prepared: 

• In the courses "Linear Rail Inspect" and "Linear Rail Pick," you wrote that "The robot must be completely on the dexterity field to score any points (i.e. no contact with the ground outside the dexterity field)." However, from our perspective, this could be problematic, as it may be difficult for our robot (and likely others) to stand entirely on the wooden board you specified. Would it be possible to amend the rule so that the robot may, for example, stand halfway on the platform or something similar? 
  

I'm wondering if we have the same definition of "dexterity field". The dexterity field is the entirety of the wooden dexterity apparatus, including the corners. (In the keypad test, we refer to the center portion of the dexterity field, that doesn't apply in this test.)

Below are some examples of permitted (green robot) and disallowed (red robot) positions. Note that while the lower green robot's outline does extend past the dexterity field, the ground contacts (for this robot, the center of the wheels) are still on the dexterity field.

Could you give some examples of positions that you would like to have allowed that don't match this? Perhaps is there some confusion because in some cases the linear rail is shown away from the dexterity field a bit? In the current design it will be right up against the dexterity field as shown here.

• Additionally, we noticed a potential loophole in the stair challenge rules that you defined, which could lead to unintended scoring. It seems that one could possibly just hang from the edge to gain points. We would appreciate an improved and more clearly defined rule on this point. 
  

I'm not quite sure how the rule allows the robot to just hang from the edge. As written: 

"A robot is considered to have moved from one level to another if it moves from only being in contact with one level, to only being in contact with the next level, without returning to the previous level for at least 5 seconds (to exclude situations where the robot “bounces” up and down as it attempts to grip the edge of the step as it climbs, momentarily losing contact with the previous level)."

To me this very clearly excludes "just hanging from the edge" (presumably while still in contact with a previous level) as they need to go from only touching one level, to only touching another level, for at least 5 seconds.

What do you mean by "just hang from the edge"?

• Since you are placing great emphasis on autonomy this year, we would also like to propose the introduction of an award for "Best in Class in Autonomy." 
  

This is the plan - we're working on finalizing awards now.

• Initially, you mentioned that the three smallest Landolt Cs must be detected. Over the past few weeks, we invested considerable time and resources into making this possible. However, your current rules state that only the 3rd and 4th rings need to be detected. This change is confusing to us, especially since you had previously insisted that the three smallest must be recognized. We therefore kindly ask you to uphold your original statement and allow the three smallest Cs to be counted, as previously communicated. 
  

I think there is some confusion as the rule you refer to only applies in the finals, where your team-mate, Eli Schr, raised the issue of points inflation. Having the Linear Rail only worth 10 points seemed like a suitable compromise. Note that my insistence on retaining all 5 levels related to the preliminaries (and as currently written that is still the case). While I had hoped to keep the Landolt-C scoring to be the same everywhere, Eli's comment convinced me that this wasn't possible. 

Actually I would have preferred to have the finals be for the smallest and second-smallest Landolt-C but the only feedback I've received from any team regarding this was Eli's comment that even within your team there is debate as to if even the second smallest is too small so we decided that the smallest wouldn't be needed in the finals (especially as there is so much else to do).

Here is the Landolt-C scoring system for the finals:

Preliminaries: 
Linear Rail Inspect: all 5 Landolt-C sizes count. Arguably this is where being able to see right to the smallest is most important.
Sensor Crate: Only the 3rd ring counts. There's lots to do here so each sensor test is only worth 1 point.

Finals: 
Only the 3rd and 4th largest count (for both Linear Rail and Sensor Crate) to prevent points inflation. I'm assuming this is the part of the rulebook you're referring to.

• For the upcoming world championship, we would like to build and bring a second, identical robot to have as a backup incase of a mechanical failure or any kind of serious problem. This would allow us to resume operation after a reset if necessary. Since this is not explicitly addressed in the rulebook, we are reaching out to clarify in advance to avoid building and bringing a robot that may not be allowed and risk disqualification. A second identical robot is essential to ensure operational readiness under all circumstances and to guarantee a smooth competition throughout the championship. Anything less would pose an avoidable risk. Furthermore, since other teams can also read this message, they too would have the opportunity to prepare accordingly — for example, by building a second robot as well. 
  

I believe this is covered under "Robot Configuration" under "Repairs" (where swapping the entire robot is perhaps carrying the idea of swapping parts to its logical conclusion). As long as the requirements there are satisfied, then this is fine. 

The general requirement is to make sure that the repaired robot (or replaced robot), going forward, would have no advantage compared to the old robot.

I've added a clarifying statement to make it clear that a replacement of a robot is covered under this rule.

I look forward to hearing of how your team, and everyone else, performs in Salvador! Sadly I won't be there to witness it first-hand

Cheers!

- Raymond

--
You received this message because you are subscribed to the Google Groups "The Open Academic Robot Kit" group.
To unsubscribe from this group and stop receiving emails from it, send an email to oarkit+un...@googlegroups.com.

To view this discussion, visit https://groups.google.com/d/msgid/oarkit/006b4977-565e-4931-b8ad-3e2c3b4a16c2n%40googlegroups.com.

-- 
https://raymondsheh.org

[Raymond Sheh]

Hi All! 

Thanks again to everyone who contributed to reviewing and commenting on the rules document! Especially as the substance has been stable for a while now and all of the updates have been clarifications, I think it's well and truly time to say we have come to a conclusion and announce that this is the "final" rules document. Same links as in the discussion: 

Rulebook: https://docs.google.com/document/d/1ta6XHinzvNRzbZgfzwVR5Y_7ua4NhY6w_eBW9HcW3d0/edit?usp=sharing

Construction guide: https://docs.google.com/document/d/1bj2Z3S5dQBwtVHq4OXep9l7vTiFjvXwiyYi6rSo3q7o/edit?usp=sharing

I *am* going to add some resources (e.g., more printable scoresheets for the tests that don't work so well with the existing ones, printable sensor test sheets to match the renderings, etc.) and will email around PDFs when that happens. 

In the end, this rulebook is mostly an evolution of the previous one, clarifying some of the judgement calls that used to be made on-site and adjusting some of the tests to better reflect how the competition has advanced. 

Of course, if anyone has any more suggestions, please do let us know and we can keep it in reserve for next year! If someone does find a glaring error that we need to fix for this year, also do let us know and we can issue a revision but I do want to only do that for things that are otherwise going to cause major problems as we've basically had the substance of this rulebook stable for a while now. 

Cheers!

- Raymond