Welcome to The Grasshopper’s 2025 Build Thread!
Salient Links (to be updated):
Please feel free to ask questions and leave comments, we enjoy the interactive nature of our build threads. Test requests are open as usual, as are ‘Let’s Talk About’ topic requests.
Chapter Posts
Hey all!
The last 12 months have been pretty tough on The Grasshoppers. At the start of 2024 we lost our drive space and game carpet due to school renovations (space taken over, carpet was left out during asbestos remediation). New drive space was available, but driving on a non-level floor with slippery carpet had a tremendous negative impact on our season last year (very little software tuning applied to the real field). Then we lost around half of our already small team due to graduations and people moving, 5 students instead of 11. And then we lost our classroom and assembly area due to PCB remediation efforts. We still have a room, but it is about 1/3 the area and no longer on the same floor as our fabrication resources. And then we lost some sponsorships without too much time to react. And then…
No more and then! I want to focus on our reaction to these unfortunate circumstances and not be all ‘woe is we’ about things that are largely out of our control.
This season will be a rebuilding year for us, both literal and metaphorical. We will be working on rebuilding our personnel, financial, and infrastructure resources. As such we will be down-scoping our robot ambitions. We will still wind up experimenting with some of the more advanced features of whatever this game brings, but will build a simpler robot.
Anyway, here are pictures of my two new kittens that climbed in Christmas trees as some eye bleach before kickoff tomorrow.
Initial cut at Reefscape
Scoring like:
2019/2011 - Multi-height placement
2023 - Precision placement with weird game piece
2004 - Low port recess ball
2000 - High trench recess ball
2017 - Climbing on a 6DOF dangling object
There are lessons from all of these games (and others I’m sure) that we will want to consider.
Here is a basic PPS (points per second) analysis assuming elite level play and some traffic.
We see what is perhaps obvious: you better move in auto, scoring coral in Auto will be very helpful to winning, a fast deep climb is powerful, algae is helpful but second to coral as far as total score goes.
A feature ranked decision matrix (note: not inclusive of all possible features) shows what we might want to pick, depending on what we choose for machine priorities.
My initial thoughts are:
Rank high - good coral l2-l4, good climb, process algae (genius, I know)
Picked first - great coral at l4, good climb, ignore algae?
Picked second - Super fast l1 coral and/or clean algae and/or shoot algae
I’m sure this will change over time, we are still digesting.
Team 2370 (The IBOTS) in Rutland is always here to help and we aren’t too far away! We have the VT Practice Field at the MINT in Rutland (112 Quality Ln, Rutland, VT 05701) that you can use at any point. If you want to drop into our lab at Stafford Tech we can collaborate on fabrication. The MINT also offers memberships so you can use their woodshop, metal shop or 3-D printing lab for a small monthly fee.
Further discussion and analysis have led us down a path to a robot that is not likely to rank well but is likely to be picked early.
We looked at the ranked weighted decision matrix for machine features with weighted to describe ‘the easiest to build robot that has a high point ceiling and good usefulness within an alliance.’ The results are:
Scoring L4 and L1 coral are top of the list, followed by processing algae and a shallow climb.
This featureset most closely matched the giraffe archetype in our archetype matrix.
giraffe - focused on l4 coral, shallow climb
deck scrub - cleans algae, can shoot into barge net or process
defender - go be annoying to opposing alliance
lo baller - focus on l1 coral only
l3 correlavator - score l1-l3 coral, clean algae, process algae, shallow or deep climb
do everything - do all the things
plowie - push algae to processor, push coral to partners
deep climber - focus most on endgame, some l1 and algae processing capability
We imagine the L3 corelavator archetype will be the one to rank best. It can score on enough of the Reef and cooperate to have a good chance for that RP, contributes to climb RP, can score coral in auto. Optimizing for L2/L3 probably means they can score a lot of coral solo and have a good chance at winning.
We want to build a robot that complements this robot well. That means focusing where this robot does not: L1, L4 coral. If we ignore algae for the puposes of generating points we will save ourselves a lot of effort. This looks like the giraffe archetype with some consideration for L1 coral.
Looking at how an L1 coral mechanism might package we came up with this:
It’s a basic opposed-roller claw that can accept lateral coral from the ground (8 o’clock, 4 o’clock), coral station (10 o’clock) and score in L1 near the back of the trench (3 o’clock). The outside of the rollers can even be used to collect algae from the ground and cram it into the processor when needed.
So, two degrees of freedom to get forward and rear ground collection, station collection, L1 coral scoring, and algae processing? Deal!
So, about L4… I think we can do it with one DOF if we are clever.
This shows how a single shoulder joint with virtual 4-bar linkage could move some TBD coral scoring device (tall rectangle) into position to drop coral on L4. We hope to make the coral release mechanism completely passive. You drive the arm into position and a lever gets hit by the L4 tine, releasing the coral automatically. Time will tell if we can make it work well enough. This device only receives coral from the coral station, but ought to be super fast to score in L4 positions.
The L1 and L4 mechanisms ought to play mostly nice with each other, assuming each one is a single-sided arm type of device.
Why is L1 so valuable if the point value is the lowest of all levels? It seems counterintuitive to me.
A reasonable question.
This type of ground collection scares me in terms of necessary precision to successfully acquire a gamepiece and non-collapsability. Maybe some kind of resettable breakaway mechanism in the arm could help mitigate the second issue? Such a device would add a lot of complexity, but that arm geometry looks like a perfect target for defenders/partners/walls to destroy.
Something my team learned from our rack and pinion intake last year is that if there’s something outside of your robot, it will be hit. A lot.
We were just discussing how to make a mechanism like this compliant enough or strong enough to survive at our last meeting.
Working through the ‘utility arm’ geometry a little more I have a crayola bot to share! A crayola bot is just general chunks of reserved space, vague mechanism shapes, and correctly-mated linkages. This lets me drive the robot around a little in CAD and see how different conditions might work out (or not) without investing too much time in details.
Important features:
-Single roller intake
-Rear ground intake only (more robustly supported)
-Single arm/cantilevered rollers, a little wild even for me, but it does some really fun stuff
Reaches over the L4 mechanism to collect from the coral station
Can score in L1 over coral
Unexpectedly it can probably extract L2-3 algae
In ground-collection position it is close to the bumper, which will support it in the event of an impact. I think the single arm is an advantage here: we can make it flexy.
We may even be able to do a handoff to the L4 mechanism… not counting on it, but it would be nice.
Things will also be able to stow out of the way of a centered-on-the-cg climbing mechanism.
It can still collect an algae on the floor too
Two degrees of freedom (one PIDF, one on/off roller) get us:
Outstanding concerns are still:
All solvable.
A wildly useful robot would have just the L1 mechanism/utility arm and a climber. Hmmm…
Side note: in order to manage burnout and keep meetings as efficient as we can we are only meeting 3 days/week for now. I have to say I am very much enjoying the extra time with my family.
I am always excited for this thread as it is well documented, brings on some great insights and you can always learn cool stuff
I realized I didn’t have the scratch work sketches shared in anyway, so I’ve done so here. OP has been updated with the link too.
It’s a work in progress.
Whipped up a detailed model of the utility arm. It was a good exercise for a new design student to figure out the CoM and mass of the arm, punch it into ReCalc. A few more bits and bobs to add of course, but enough done to start a prototype.
The blue shovel and light blue/gray axle mounts will be plastic. The red baseplate is also likely to be plastic and our main source of mechanism compliance. This arm does not need super precise position capability, so mounting the whole thing on something flexy ought to buy us a lot of impact forgiveness.
We started testing a passive intake for our planned L4-only mechanism. Results so far are quite promising. The objective of the mechanism is to accept and orient a coral from the coral station (which seems to work) and then release the coral via passive mechanism when it is in scoring position (design TBD).
(videos are uploading while I make this post, it might not work yet)
Climbing sight lines will be important this year, and the reef is right in the way. Flying around in OnShape it looks like these will be the best sight lines/climb positions in order for all teams to have a mostly-clear view.
The last one is not super clean, but it is the best view to any of the cages.
Isn’t this perspective from the Blue alliance’s driver station? Also algae being removed would help a bit with the view.
You got me. Maybe crafting a post way past my bedtime isn’t the best options.
So just mirror the arrows?
I agree with this set of arrows and the concern that the algae can block things up a bit. Alongside the CAD, the images from the the Field Photo Album on the playing field page (Playing Field | FIRST) were educational for me when trying to identify how much of concern this may be. Driver station 3 leaves a bit to be desired without removing at least one algae or leaning in towards Driver Station 2.
I mean… Yeah visibility is bad, but its not against the rules to lean over to your buds in alliance station 2 to do a deep climb on cage number 1, right?
