There are other ways to mitigate trap use and still get the RP and still only need 2 robots (I.e. have a robot that’s good at harmonizing (getting on 2nd at a higher point on the chain) with a climbed partner and a human player that is really good at throwing one of their 3 rings on the same peg every time). I personally think teams are going to figure this out, like they did with climbing the tiny bar in 2018 and really fast tripling last year.
This is of course assuming the thresholds don’t change for DCMP/Worlds, which they could.
We had another snow day so no prototypes today either. We ended up discussing the pros and cons of 5 different concepts:
There are hundreds of ways to combine these elements with miscellaneous tradeoffs. In the end, after discussion, some voting, some discussion on how this isn’t a democracy and actually @brendanb is in charge, the team settled on(the voted for concept) Concept 2. What’s Concept 2 about?
4B26
There’s two key tradeoffs that most of these other designs make that 4B26 does not. Firstly, the design can go under the stage without extending the intake. This should make it easier to cycle without having to constantly manage the state of the intake. Secondly, 4B26 has a good concept for human loading for both cycles to the amp and to the speaker.
The tradeoff for those functionalities is a little more complexity in having a separate amp mechanism and having to hand off the note between the intake and the two scoring subsystems.
We like having different subsystems separate when possible - it makes it easier to work on one system at a time, get more students designing and prototyping, and make mid-season upgrades.
What’s Next?
The big milestone we’re pushing toward next is constructing an Alpha Robot. Our goals for Alpha this year are to give our young software team time to get their legs under them with some of our key concepts. With no inverse kinematics to do this year (whew!), that’s mostly about getting some universal items ready to go - PID, path-planning and vision targeting, and to get that all working with an intake and shooter on the robot.
So the plan for Alpha is as follows:
Fixed angle shooter
Floor intake
26”x26” swerve
With the goal of starting manufacturing this weekend, we’ve started sketching our alpha mechanisms.
We’re still working out the details of this and whether we can work with the geometry of the intake/4B directly to the shooter or if we need a handoff. Should have an update over the next couple of days.
Still not a lot in the way of exciting videos. We had some minor intake prototyping to confirm our geometry for the alpha intake but no videos were taken. Hopefully we’ll have alpha together within the next week and this blog will start heating up.
Alpha CAD has been updated with many tweaks:
We’ve also been cooking up some concepts in CAD for the next iteration.
An amp scoring mechanism:
A rig to test different angles of the non-parallel shafts top/bottom shooter that 95 showed off in their blog:
If you haven’t noticed, we do a lot of ideating in CAD, and most of it doesn’t end up actually happening, so we’ll see if these get made or end up in a further iteration of the design!
We put together our Alpha robot this weekend and… we weren’t super happy with the results.
Due to a number of manufacturing issues, this took longer than expected to get together. And once we did, we found a number of issues.
The first issue was the intake failing to consistently intake and center the game piece. This isn’t a big deal, as this can be tweaked and iterated on, but it wasn’t promising.
The second being that the handoff was very sensitive to changes in shooter angle - not great when our priority list includes 3 different shooting angles.
We also discovered we’d mis-read the rules about shooting from the podium. G422 gives protection to robots whose BUMPERS are touching the podium, so a position where our intake is touching the podium does NOT provide protection.
The question then becomes - how do we proceed? Are these failures on our first mechanisms indicative of how working on future novel mechanisms and concepts will go?
We’ve mostly settled on option 3. Our current concept isn’t going to do the things we want to do well, and it’s important to recognize that early and move forward in a different direction.
So What Now?
Well we’re looking hard at everything that’s been made publicly available. We love the “inside-outtakes” made by 95, 4522, and others. We think 3255’s shooter is top-notch. We want to keep it simple, especially since we’ve lost a week from a misguided alpha build (and three snow days already!!!), so we’re looking to find some combination of ingredients that does the trick.
We haven’t seen a lot of good amp mechanisms out there other than 4481 and the everybot amp mechanism, so there’s still a lot to learn there.
We’re leaning toward a combination of an under-bumper intake and the 3255 shooter, pending prototyping.
If all goes well we’ll be putting together the pieces for a “Beta” robot by this time next week.
I think it’s important and respectable for teams to post things like this. One of the guidelines of OA is posting the bad with the good; I think it’s easy to not want to post the bad, because, well, it’s “bad.” But bad happens to everyone (I think…), and one thing that good teams have is their ability to learn from the bad and pivot/adjust.
Came here to say what Justin said - really good to show all teams reading this that even a high performer from 2023 is having a tough go of things not going right, and that things don’t “just work” the first time around even with all the planning in the world.
Glad to see these updates, hopefully we’ll have more intake posts of our own tomorrow night.
Our 2022 intake, one we were pretty happy with, was on its 5th or 6th major revision by the time competition rolled around and under another 2 or 3 during the season. I expect this year to be no different.
The main issue with this specific version is… the bumper. The bumper rules here are both annoyingly specific and not specific enough.
R101 says we must have a frame perimeter of fixed, non-articulated, structural elements of the robot within the bumper zone (<7.5"). It does not state a minimum height or thickness of that frame perimeter, which seems to imply the frame perimeter could be made of very narrow, thin material and would count, per the rules, as a frame perimeter.
Then R410 discusses how the bumpers must be supported by the frame perimeter of the robot, must not have large gaps in it, and the blue box mentions that
Flexible ROBOT elements, such as thin plastic, do not accomplish this intent and are not considered “structure/frame” of the ROBOT.
So what does this mean?
I think it means that as long as whatever member we have in front of the robot is stationary, rigid, and attached to rigidly by the bumpers (per R408G), it legally counts as frame perimeter.
By that definition, a single 3/8" hex axle across the front would count!
I think you’re misunderstanding R408G - the bumper is not required to attach to the FRAME PERIMETER on every side of a ROBOT (e.g. for a square robot, you don’t have to attach the bumper on the front or back sides, so long as the bumper is rigidly attached to the frame in general). Your hex shaft does not need to attach to the bumper, provided it backs the bumper.
To the rest of your question…
I think the hex shaft is fine in my interpretation of the rules, although your inspector may vary. A piece of aluminum hex is a rigid frame member (at least in my reading of the rules), and provided that the intake plates are also not thin plastic (since they also back the bumper on the sides), then I would say that this is a valid interpretation of the rules.
YMMV - every inspector and LRI will have a different reading of the rules, although I imagine you would make a strong case for a “pass without correction”.
As you mention, the rule that most applies here is R410. I would guess that this will be interpreted quite differently by different inspectors. Having seen a variety of different calls from different LRIs on similar issues; I would probably try to beef this up.
I personally think your solution is totally acceptable per the rules as they are written as well as complying with the intent of the rules. If you have trouble with an RI at your event, ask to speak with an LRI, if you have trouble with the LRI ask them to contact HQ for clarification. This is a totally reasonable request and is well within your rights.
In addition, I think it would be smart to manufacture a metal bumper frame for additional reinforcement. Some notable examples to check out include 5406 and 118.
We’ve finally found a concept we really, really like.
This manages to combine the amp, shooter, shooter angle adjustment, and potentially climber into one decently simple arm (It’s fairly similar to Unqualified Quokkas robot, but with an under-bumper intake (we didn’t realize this similarity until we were reading the build thread of our friends on 5459 who came to a similar concept!). And luckily, we have some recent experience with arms!
Tonight we also finally got some shooter testing done:
The 3255-style shooter got assembled:
And the inside-outtake got assembled!
Look forward to some testing videos of these other two prototypes this weekend!
Have you had a chance to test the consistency of this shooter? And is there a reason you went with single wheels (8 total) instead of putting them in groups of 2 (16 total)?
We finally got this shooter working well. Ended up with the ThriftyBot 4” Solid Urethane Wheels in pairs all running off their own falcon.
We’ve still got a lot to work on to see if we can get the desired spin-up time with just 2 falcons rather than 4, and how well this works when it’s sitting on top of a big arm, but we feel much better about the shooter than we did going into the weekend.
On the other hand, we’re still NOT happy with our intake testing. It seems that having two sets of mecanum rollers trying to center the ring is not working. We beefed up the ratio on the intake and it still stalled out on the sides. We’ve got plans now for a much more active centering intake, also like the ones 3255 prototyped (good blog, guys!), initial CAD of it below the video.
Most of the arm details have been worked out. We debated how much of last year’s arm design to keep and how much to change. After much deliberation (and @burbozo3467 having to change the arm joint in cad 4 or 5 times), we think we’ve settled on a live thunderhex axle on top with the Rev Thru-Bore encoder on that live axle.
The biggest pain on last year’s arm was having to remove and shim the 3d-printed gears from the arm at every competition to keep the slop between the arm and the encoder at a minimum. This was exacerbated by those gears being inside of the arm uprights.
