Sensor crate, CO2, and magnets
Raymond Sheh
We've just received a question regarding CO2 and sensing.
TL;DR: Since the pandemic, CO2 has been replaced with sensing the
polarity of magnet(s) as the proximity sensing test so please don't get
confused when you're reading past years' TDMs!
I realize that there might be some confusion because between 2022 (the
first competition after the pandemic) until the most recent rule update,
the RMRC rules have pointed to the Major rules for sensing and the
sensor crate. While the Major rules have not included CO2 sensing since
2019, I can also understand that teams who were familiar with the
pre-pandemic rules (or had read TDMs from teams prior to the pandemic,
or from teams who also didn't realize this had changed) might not have
caught the change. This is one of the reasons I pulled those
specifications into our own documents (and simplified them) in the
latest refresh.
The original goal of CO2 sensing was as a proxy for sensing tasks that
required the robot to place a sensor in proximity to an object of
interest. In a real application, instead of CO2, this might be some kind
of dangerous gas or a radiation source. The problem with CO2 in
competition was that it was hard to maintain a consistent concentration
in an open area. CO2 canisters vary in their output, air currents move
gas around, having people breath into the robot becomes inconsistent
(and wasn't a great idea coming out of the pandemic).
As the underlying idea was to simulate a proximity task, we replaced the
object of interest with a magnet, and the sensor with anything that can
detect the presence and polarity of a magnet. This can be randomized and
doesn't require a "recharge".
As discussed in the current rulebook, the current test involves 3 disk
magnets, around 25mm in diameter, of the kind you would use to hang
things on a refrigerator, with North (N) on one face and South (S) on
the other. They are placed 50mm apart in a random, orientation. The goal
is to read the orientation pattern. For instance, "N S S" would mean the
left magnet is North up, the center and right magnets South up. With 3
magnets, 2 states each, you get 8 combinations (the same as the number
of options in a Landolt-C).
An automated electronic way of doing this would be via a Hall effect
sensor. For this year, we would accept something as simple as a compass
attached to the gripper that the operator reads with their camera. (This
sounds weird but there are plenty of cases of deployed robots with
various sensors or chemical test strips attached to them that are read
via their cameras.) If you do use a compass, make sure you invert the
reading - if the "N" on the compass is pointing at the magnet, that
means that the South pole of the magnet is up. Note that we *might*
change this in future to require something more integrated so please pay
attention to the rulebook and rules discussions in this mailing list.
Please let us know if you have any questions!
Cheers!
- Raymond
--
https://raymondsheh.org
Raymond Sheh
rules have not included CO2 sensing since *right after* 2019 (2019 being
the final year where we ran with CO2 ... 2020 was cancelled, 2021 was
remote, 2022 used magnets).
Cheers!
- Raymond
Axel Pérez
We hope you are well.
We have several questions that we would like to clarify and share with the other teams.
The 2019 rules mention the motion detection sensor, and in previous competitions this sensor has been evaluated in both the preliminary rounds and the final. We would like to know if this sensor will be evaluated in the 2025 competition or if it will be discarded.
Additionally, the current rules do not mention penalties for leaving the dexterity track. If a robot leaves the track, can it return to try to regain points, and if so, would those points be valid?
Thank you in advance!
Raymond Sheh
Hi Axel,
Thanks for the question!
We will not be evaluating motion detection this year (and are
open to suggestions on how to do it usefully in future years). See
discussion on this list regarding sensing from around May 3rd.
For the dexterity field, as written the robot could theoretically
drive anywhere (within other rules such as safety) between scoring
points, it's just the scoring of points that is limited to the
dexterity field (so for a pick and place, they need to stay
completely on the field from pick to place). Of course open to
suggestions as to if this needs to be tightened up but I guess
I've been assuming that there really isn't much incentive for
robots to stray too far from the dexterity field and this is more
of a dexterity and situational awareness test, rather than a
terrain test.
Do note that, as situational awareness is part of the test, it is
up to the operator to know if they are on the dexterity field when
they declare that they've scored a point. The judge, audience,
team-mates, safety person, tether wrangler, etc., should *not* be
telling the operator that they haven't done so and to try again,
for instance. See the definition of "Operator Station" on pages
12/13 for further details.
On a separate note, I accidentally put in a physical
double-negative in my description of the magnet test. My statement
that "If you do use a compass, make sure you invert the reading -
if the "N" on the compass is pointing at the magnet, that means
that the South pole of the magnet is up." is wrong. It would be
correct if one were using something sold as a magnet, such as a
bar magnet which would have a North pole marked "N". If allowed to
swing freely it would point to a South pole.
Of course, magnetic compasses are also magnets - but as sold they
already have reversed markings. The end of the needle or half of
the disk of a magnetic compass that is marked "N" is actually a
South pole. That way, it's attracted to (and points at) the
Earth's North pole. So if you were testing with (a magnet sold as)
a magnetic compass, you don't need to invert the reading to get
the correct answer.
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/ca457a4a-cc5b-4eb1-b0f8-0ee774373683n%40googlegroups.com.
-- https://raymondsheh.org