Thanks to our teammates for the special 2025 run!
*I’ll see the mods can join the reveal thread and we are happy to answer any questions.
179 Likes
At 28"x28" is this considered your smallest robot?
10 Likes
Every single year, I think my mind can’t be blown again from the amazing work you guys do, and every year you prove me wrong. Congrats on the well deserved win!
I didn’t see a lot about the automation of your bot in the tech binder, can you go into how much of your intake to scoring process for algae and coral is automated?
3 Likes
Amazing robot. Congrats on winning worlds!
4 Likes
Our 2017 was the smallest frame size in most recent years. ~107" frame perimeter vs 112" this year
8 Likes
What material did you use to print the bevel gears?
2 Likes
I’ll try to be as detailed as possible:
Coral → Scoring off the ground was done via the ground intake in conjunction with a diffuse Banner sensor + candi. Once the robot saw the coral it auto lifted, handed off and got into a scoring state. From there the co driver would pick the level and the driver would pick left/right pole to score onto. The software handled the rest including front and back side scoring.
Algae → Off the ground was done via a similar setup. There is a candi + retrfo reflective banner sensor in the claw that detects coral or algae, once it triggers the robot auto hands off the ball from the ground intake to the claw and goes into the scoring position.
Algae → Off the reef, same as above but purely tag based, this motion is completely automatic.
The programmers did an amazing job dealing with all 6 DOF’s and most of the robot was automatic.
15 Likes
Would it be possible to share a little more what’s going on here? I gather that that’s supporting the A-frame tube, but I can’t tell what’s going on with what appears to be a bent piece in there? Or is that 4 pieces and the internal one is just a nut plate basically?
If those are inside the a-frame tubes, did y’all do anything special to hold it in place? Or did you assemble the a-frame tube to the supports and then attach to the base rails?
Is the blue brace bent? It’s hard to gauge the thickness, but from the tech binder, all those plates look 1/4" to me
2 Likes
The practice robot was fully printed in PETG HF on the bambu printers. The competition robot is fully onyx on the markforged.
2 Likes
Hoping this image helps the black large plate is 3/16" AL with custom made nut plates that go inside of the 2x2" structure box tube. The outside blue bent plate is .090" AL.
We’ve copied 4414 and now use the top portion of the swerve plate for structure. The 4 highlighted holes were bolted to a nut strip that was bolted to the swerve plate. Also the team recently picked up at a Titan 25T brake press after following along Spectrum’s journey. But the better place to get bent parts IMHO is always fabworks.com
This image should give a better picture of the thing assembled.
15 Likes
How is the x44 on the end effector wired and are there any sensor to detect where the coral is in the end effector?
2 Likes
The end effector on the wrist is in PWM mode due to our fear of breaking the CAN chain and not wanting to stub that long. In CAN mode, we updated firmware on it and set stator/current limits then wired it up as PWM.
There is a retro reflective banner + candi setup that detects both the algae and the coral.
Putting sensors on the candi helped us a lot with the speed of detects, we’d always have poor results plugging directly into the rio in the past.
8 Likes
Beautiful robot and well-deserved win!
Did you notice a difference in PETG vs Onyx? Did the practice robot see more or less time than the competition one?
2 Likes
Answering from the other thread:
The team used both thinner gauge 12/14 awg along with alpha wire to help reduce weight/size across the border. This allowed us to drop from a 15mm x 30mm echain to a 15mm x 15mm echain.
4 Likes
Yeah the practice robot definitely saw more time and probably deserved to be retired a long time ago. We do PETG on the practice robot for two reasons:
-
Affordable and super fast on the bambu. It doesn’t hurt when students/mentors wanna re rev parts.
-
The PETG parts fail faster/earlier than MF parts, it really allows us to see which parts or areas need improvements early vs failing late into an elimination run.
Competition robot only sees a few auto tunes + matches only. Our goal is to put the least amount of run time on it.
21 Likes
What is your Reef Alignment automation/method? Do any of the LL4’s have Hailo’s being utilized for April Tags? Also, which did you prefer more for cooling the LL4’s, the milled heatsink on the ‘center’ cameras, or the fans on the other cameras?
1 Like
How hard was it building within the new 115 weight limit?
We had trouble staying within for a while, but y’all do build bigger robots than us.
3 Likes
Out of curiosity, from a strategic standpoint, what lead you to your current design, and how does that differ from the team’s initial strategic beliefs?
1 Like
Let me double check on the exact alignment method but we used 3 LL4’s. The two for april tags did not use the hailo tag as we noticed performance was better without it? The other LL4 was simply for auto only and object detection.
Every camera had a milled heat sink and we were only able to fit 2 fans out of 3 on the LL4’s. The third wouldn’t let a battery come out… 
The LL4 has to be the best camera limelight has put out and it worked great for us all year.
5 Likes
Honest answer, it sucked.
Our practice robot weighed 130 lbs on the scale week 5 of build season. Things from the other thread we did:
- Switch to WCP SRPP
- Pocket more… A lot more.
- Switch out steel bolts for plastic
- Switch out steel bolts for Ti
- Pocket inside robot faces of tubing, leaving the outside pointing faces alone
- Remove unnecessary holes and add more pocketing
- 3DP stuff that was metal or solid plastic
- Make aluminum plates thinner
- Drill out shafts
- Go to thinner wire
- Use bumper weight more effectively
- Stub roller more of our shafts
You know its tough when we are looking for .01 to .05 lbs at a time.
31 Likes