As a corollary to the Mistakes thread, how about a thread for the key decisions that led to success in the 2023 season?
I’ll start off for 857:
- Swerve. This was one of our fastest and most maneuverable robots ever and the sideways movement was so helpful for lining up cone placement.
- Being able to herd game pieces with our bumpers. Our bumpers were two C pieces with a gap in between which was very helpful for herding game pieces on the ground. In qual 23 at MSC we were able to score two knocked over cones from midfield in the last 25 seconds.
- Focusing on balancing Charge Station in auto. Our first competition was never getting links rank point, so our super consistent auto Charge Station got us 3 RP almost every match landing us at rank 1.
- Scoring both game pieces. In 2015 and 2023 we scored all the game piece types and were very successful, and regret not doing both in 2019.
- Scoring high. Especially at our first event where scoring about 4 game pieces per robot was a lot, getting every point out of those game pieces was super helpful to the match score.
- Low CG. Not only did we almost never fall over, we seemed to help bring the Charge Station into balance a lot better than a lot of other robots. The two times we did fall over was from accidentally extending our arm while moving, but luckily we also found our wrist had enough torque to flip us back onto our wheels.
- Our arm geometry let us get right up to the grid before extending our arm, unlike some other designs that had to sometimes even back up, extend arm, and then move forward again to score.
- Floor pick up. Necessary for our multi game piece autos, and we had some matches where we able to score more of the midfield pre-setup game pieces than I would have expected, like Match 9 of MSC playoffs where our robot scored 6 of the 8 staged game pieces over the course of the match.
This is a great idea I will write a long post here when I can.
Heavy steel belly pan on a tiny frame, nearly untippable and our frame was nigh unbendable.
Our gripper was having problems getting stuck to the human player station so we added wheels to the end of our intake to help it move on the plexi
We put a counter weight on the bottom of our robot and it really helped with of COG
Installing weights into the frame of our robot. As a second year team and (generally low scoring) this let us stay very offensive by using our robot for defence. Which is the reason we ended up coming second in one of our district events.
Seconding (fifthing?) the low CG, steel belly pan was a great move.
As crunched as our auton development was, the final code ended up being incredibly readable and easy to maintain. This proved useful at several, semi-panicked, critical moments.
Making all our parts relatively easy to replace. Totally saved us once.
Choosing to use Limelight for ground pickup. Made pickup of pieces significantly easier once we added it (right before champs).
LED usage. We developed a pretty good system for signaling all sorts of things to the human players through blinking colors.
Double linear extension. We considered using a pivot but eventually settled on two linear arms. This introduced some complexities but overall was a net positive in terms of control complexity.
I’m going to generally mirror all these for 6995 (except the bumpers one).
Focusing on driver automation using AprilTags and auto-align. More details here: Piloting Assistance/Automation on Good Teams - #5 by Amicus1
Going 25x25 for a drivebase. This massively helped maneuverability in the loading zone and helped us sneak through defense.
Adding “bumper alignment aids” made of solid steel to the back of our bumpers and solid bars to the edge of the frame opposite the arm extension. We had max-weight bumpers and max-weight bot and we rarely tipped.
Functionally finishing our major CAD elements by January 11th
Sticking to an aggressive build schedule which enabled us to start driver practice with a funtioning robot on February 4th
Attending a week 1 event
Stealing the Everybot intake concept and integrating a version of it into our existing superstructure
Designing around a linear actuator to move our high pivot arm (made programming a lot easier)
Prioritizing driver practice and Autonomous testing over creating a more complicated/capable robot
Using swerve drive for the first time in team history despite only getting MAX Swerve modules for the first time in Mid December
Picking 9143 as our third at SFR.
These bout summed it up for my team.
Focusing on simplicity and functionality over complexity and wow factor
not going for swerve (yeah yeah its great but having a drivetrain that we know was better)
accounting for the fact that we didn’t have swerve
getting actual auton coding and driver practice (unfortunanly after first comp but still helped a lot)
ditching ground pickup for our second comp (was too slow to do)
Bellypan and swerve
Becoming a charitable nonprofit!
Also, steel weights that doubled as swerve protectors.
Doubling down on preseason training.
i know i said swerve was prolly a mistake in that thread but i also think swerve was a good idea, i also think our virtual four bar, although tough to control, was very cool and i don’t regret it
One smart move I made was getting out of the way before the robot arm hit me in the face. Yes, the robot would have been disabled. But I might have been disabled too.
Picking 857 and 7220 at states .
In all seriousness:
Putting handle loops on the robot
Switching intake designs
Putting googly eyes on the robot.