FRC 1923: The MidKnight Inventors 2023 Robot - OverReach

[SJaladi]

FRC 1923: The MidKnight Inventors are proud to present our 2023 robot:

OverReach was designed with the goals of being nimble, quick, efficient and accurate all while fitting into a small package to facilitate easier endgame balancing.

Robot Specs:

• 24”x24” Footprint
• ~95 Lbs without bumpers and battery
• CG located ~7” off the ground and centered on wheelbase with minimal shift even when reaching for L3 with a game piece
• Can score on all levels from both sides of the robot
• Can intake from the ground and double substation

Drivetrain:

• SDS MK4i w/ Falcon 500 - L3 Ratio
• Compact Bellypan with top and bottom mounted split E-Board
• Bottom mounted battery for lowering CG

Intake:

• Deployed and stowed via Double Jointed Arm allowing for multiple useful positions:
  • Intake
  • Stow/Handoff to Scoring Arm
  • Eject game piece Low with Intake inside Bumper Perimeter
• Vertical Rollers powered with independent Falcon 500s
• Proximity Sensor for confirming game piece acquisition
• Pneumatic Cylinders to open Intake Arms to facilitate handoff to Scoring Arm

Scoring Arm:

• Double Jointed Coaxially driven arm powered by Falcon 500s
• End Effector uses pneumatic cylinders to clamp game pieces
• Driven by HTD 8M Belts
• Carbon Fiber Tube construction with Markforge inserts to transmit torque
• All air tubes and signal cable run internal to the CF tubes
• Uses self-aligning magnets to ensure arm snaps back to and holds stow position even under high acceleration

Controls:

• State Machine manages all movements between subsystems
• LED Poles for indicating Arm and Intake States pre-enable and desired game piece post-enable
• Two Limelight 3s for tracking AprilTags and auto-aligning robot to desired scoring location
• Operator only has to choose game piece type and desired scoring location
• Autonomous trajectories generated using PathPlanner
• Motion profiling done with MotionMagic while referencing absolute encoders that are 1:1 with all arm joints on the robot

Feel free to ask any questions! 1923 is excited to debut OverReach this weekend at the FMA Mount Olive District Event!

[Akash_Rastogi]

@SJaladi

Uses self-aligning magnets to ensure arm snaps back to and holds stow position even under high acceleration

Clever stuff! Looking forward to seeing this in person sometime!

[John_Bottenberg]

Wow, I can’t wait to see this at Mount Olive this weekend! Looks very 971-esque.

[Libby_K]

@John_Bottenberg

Looks very 971-esque.

Definitely a huge inspiration for the concept!

[Skyehawk]

Holy chonky intake batman!
I like it.

[s-neff]

95 lbs

[Lil_Lavery]

Congrats to 1923 for captaining the winning alliance at Mount Olive! Arms are overrated.

@s-neff

95 lbs

Even lighter after their first qualification match

[Libby_K]

@Lil_Lavery

Even lighter after their first qualification match

Truer words have never been spoken.

Record scratch Yep, that’s me! You’re probably wondering how I got into this situation.

I think it’s important to update with this thread with the ‘why we didn’t put it back on’ components of our performance this weekend - we got a lot of questions both at the event, and from friends at home.

For clarity; a 3DP joint on the base of the arm sheared - probably just from stress during home practice - in our first QM. Even with Onyx, solid infill, we expected parts to wear out and need to be replaced, we were just hoping perhaps not so dramatically. (What can we say, we’re a little extra.)

We had plenty of replacement parts, for both the joint as well as complete duplicate arm. At first, we figured; okay let’s replace the joint and put the arm back on, or use the spare, when we have a longer break than just a few matches.

Here’s where it gets fun.

All morning, our strategic team students and mentors had been talking about this super-cool low cycler that was holding 1st seed at Midwest. Watching matches over breakfast, talking about how great the concept was… wonderful! (Hi Crowbots!) We had already talked about our gameplay strategy being ‘get points and links on the board as fast as you can, keep it low and fast if you don’t need the points margin’. We had also been working on our double-jointed intake having a position to shoot cubes at home, to help us finish links faster than doing a full handoff / arm motion. The plan for our quals, arm-included, was to run low links quickly and only use the arm when we absolutely had to.

When the arm came off, it was like a sign from the universe we were picking the right strategy. We knew we could still accomplish (most of) the things we wanted out of matches for the day, and we could take our time deciding what to do next. Refining the cube shot on the practice field, changing up autonomous to shoot the cube high, and getting comfortable with quick cycles in our qualification matches.

We knew our ideal partners needed to be excellent with the cones to make up for it, but it didn’t feel like a great use of our time instead to reinstall something we were doing fine without. Instead, we could focus on doing the stuff we’d planned really right, and develop the next iteration between now and our Week 5 event. (This is why we LOVE leaving big gaps between our two district events.)

More than anything - We got a lot of questions about ‘will you put the arm back on’, ‘when are you going to replace it’ - we had the two arms sitting in the back of the pit all day ready to go. The answer at this point is - maybe? We’ll see.

TL;DR - OverReach has taught us quite a few lessons in the first event alone. And we still have a whole season to iterate, learn, and refine our performance. That’s what happens when you push your technical limits to try new things, and that’s ok as long as you learn from it. We’re not sure what’s next, but we promise it’ll be in this thread. Our Mt. Olive bot got a temporary name - Lemonade - because we sure as heck squeezed a lot out the lemons we got in our first match of the weekend. No arm? No problem.

Many thanks to 125 and 484 for being excellent partners. We had so, so much fun.

[Revision0]

Arms are indeed overrated

[mfbrechting]

Awesome story, lesson, and showcase of ability by a cool team! Congrats on the win. Love the irony of the original name and the new name. Best of luck!

[Libby_K]

Oopsie woopsie, we had a whole thing typed up and then we got busy prepping for our event.

I know we promised to talk about ‘what’s next’, so here’s a quick recap even though we already did the thing.

Sometimes, when you OverReach… you have to:

We decided to be intentional about the cubes-only, super-fast meta. With only 3 weeks between events, and then a back-to-back-to-(hopefully)-back gauntlet of Montgomery, FMA DCMP, and (hopefully) Worlds, we knew we needed to keep it simple, develop better autonomous modes, and get really consistent.

Quick run-down:

• Changed to a single-jointed arm on a pivot (heh) point for mechanical simplicity
• Added a top-roller intake that actuates to (down) pick up cubes more efficiently and (up) stow out of the way for shots
• Redistributed weight even lower - 98lbs at Montgomery inspection
• Created 3, 4, and 5-cube autos (3 & 4 with balance)
• More automation for robot state sensing (though admittedly with fewer states)
• Practiced, practiced, practiced.

We knew this was going to give us dependence on cone bots for partners, but we felt confident in our partners reliably getting 2-4 cones as the level of play deepened throughout the season.

We’re very proud of this Funny Little Guy^TM and were overjoyed to win Montgomery alongside our friends @ 1403 Cougar Robotics and 8714 RoboGriffins (who we mentored last year in their rookie season and continue to support.)

Onward to Mid-Atlantic DCMP!

[John_Bottenberg]

What an RP machine! Hard to outscore as well, especially with quals partners that can complement the cube scoring with a few cones. With Statbotics’ EPA breakdown saying that you do the most cycles in the district by a substantial margin, it’s going to be hard for the best high cyclers in the district to compete with you in the link RP department. They’d probably need to get significantly more wins to seed ahead of you. If I had to guess a team to seed 1st overall, it would probably be 1923. Or 2539, just to hedge a little.

@Libby_K

We knew this was going to give us dependence on cone bots for partners, but we felt confident in our partners reliably getting 2-4 cones as the level of play deepened throughout the season.

This robot paired with 2 DCMP playoff-level conebots seems almost unstoppable. I’m sure you guys have practiced the defense you’ll have to switch to after filling all of the nodes…

[1493kd]

I’m really excited to see this robot compete at high levels. Your low trajectory shots into the low nodes for auto are amazing!

Do #teamcubeshoot proud

[Brandon_L]

Watching this thing zip around the field was insane. It’s just controlled chaos.

Loved working with you guys this weekend, we’ll see you at Lehigh!

[jedgar6]

[Michael_Corsetto]

[Drost]

The 1923 students and other mentors when I suggest scoring cones mid with the cube shooter flag:

✂️ pivot

[Gdeaver]

As you explore the use of composites and 3d printing on First robots, remember sometimes you just got to go with “heavy metal”.

[beter]

Glad to see we’re not the only team with a 5-cube auto now. Super cool robot with a super cool story. I can’t wait to see y’all at Champs. Have you had the chance to run your 5 cube auton yet? I’d love to see video.

[SJaladi]

Here’s a video from earlier in development when we were pushing for the balance at the end too. That may not pan out but we do have a better tuned 5-cube and sprint to mid-field auto ready to go.

Probably going to work on a variety of more accurate 3 cube cable protector side and balance autos that are more partner compatible as well as dialing in shot accuracy on the 4 cube and balance next.