I’d say 968’s 2007 gearbox really kicked off the use of aluminum gears and hex bores in FRC. It didn’t really exist before this. pic: 968/254 Transmission
Team 968 and Team 60 also set a standard for exceptionally clean wiring, back before most teams did.
I’ll throw in 254 from 2014 with (what I believe to be a first in this program) their implementation of path planning/following. I think many would agree that this paved the way for “modern era” FRC path planning that is now almost common place amongst many teams (although it’s still mind blowing to see every year IMO).
357 brought mecanum wheels to FRC in 2005 with Jester Drive. They had to hand pour each roller and assemble the wheels custom. 190 fielded a mecanum robot that same year.
Agreed completely that 254 in 2014 brought path planning/following into the forefront of FRC, but teams had their own more primitive implementations going all the way back 2008 (and possibly further). I know getting those algorithms figured out were game changers for 1114’s 2008 robot (a lot of which was based off of 111 in 2003’s work).
Slightly before my time there. I didn’t know that it had history prior to 2014 (I’ve never read/heard much talk of it before). Very cool to hear that there were various implementations of this in earlier years too!
The Pink Team’s nesting extension arm was a direct copy of the 177 arm used in 1997-2001, 2003, 2005 and the same basic design is still used in our elevator mechanisms.
Edit: video link showing the 2001 Iteration that is probably the best known version:
2 Edit: Watching this video I realized it also contains footage from the match that caused the rule “the scoring object is considered part of the robot to be created” when our robot fell out of bounds but was balanced on the ball so we were not touching an area outside the field, and therefore not DQ’d.
https://www.youtube.com/watch?v=wUnjfGgQbQ0
The first Roller claw was simultaneously developed by 45 and 177 for the 1998 game. I believe we debuted them at the same event.
Edit: Adding pictures.
VersaPlanetary gearbox is the biggest game changer in my 18 years in FIRST.
If we are going COTS then this is a different discussion…
I am happy to have the nostalgia trip though.
Also from 254 that year: Cheezy Vision. Sure, they weren’t the only ones “breaking” Autonomus that year (1114 did something similar with a Kinect), but to have something so modular and portable (in the software sense) that it could be released to other teams for championship (only requiring a webcam on the driver’s station) yet so effective that it was explicitly banned the next year.
Really, if you make something so effective that it gets banned for being too effective (not just because it damages the field), you deserve to be on this list. 71’s 1997 detachable goal cap comes to mind (alongside a few other designs already mentioned in this thread).
Fairly certain 900 should be somewhere on this list from what I’ve heard then 
Another team based on that sentiment would be 95 with their vacuum powered drive base to help provide extra traction/normal force.
Remember when Harpoons passed inspection, and made as far as the side of the Einstein field?
This may or may not be a better photo of that intake
If you’re referring to the 2015 900/1114 harpoon bot, I’m not sure how I feel about it being on this list. Notable, yes. But other than all non-cots cheesecake being eventually banned in 2020 due to the elimination of bag day, it wasn’t ever ruled illegal (in a design/concept specific way); likewise, it hasn’t been copied or really built on in events/seasons since. But I stand to be corrected.
I was more referring to the numerous rule changes they have been responsible for over the years.
This post here makes reference to the shear number of times this has happened (at the bottom of that post).
71 “created” rules twice… 1997 (the intentional removal of parts rule) and 2002 (the carpet rule, although a lot of teams were running studs/cleats that also weren’t so nice to the carpet). The latter I’m not counting since the rule was more of a “don’t break our field” rule more than it was a “that’s so effective it ruins the game”… although I could see one arguing that it was a bit of both.
469 didn’t change the rules in 2010 either (but it was indeed innovative)… they just capitalized on what came out looking like an intentional loophole per a Q&A response that explicitly legalized such a design. Although one could argue that the use of agitators in this years goals (to randomize which chute returned cargo exits) was a direct response to said design 12 years earlier… otherwise a crafty team could design a robot that could shoot in a goal in a way that would always return the cargo out the same chute (and be parked at said chute to immediately recapture and score the cargo over and over).
I would say that this is more enforced by the rules requiring cargo to hit the floor before being scored again (as was also required in FTC Velocity Vortex)
This video has some footage of our first telescoping arm in 1997, and our first roller claw in 1998 (also on a telescoping arm). https://youtu.be/RbfgBgXh9zo?t=160
StangPS is absolutely the godfather of all of the modern state estimation and drivetrain motion planning you see in FRC today. When I was a student in high school I was blown away when they posted what they had implemented, and it inspired me to go understand exactly how it worked (after I got over my initial reaction of “it’s so good that it’s not fair!”). This wound up giving me quite a heavy push towards robotics / autonomous systems in college and my career.
I take immense pride that my teams (along with so many others) have helped move the ball forward in these areas over the years as well, and at this point what comes in WPIlib makes all the early attempts at autonomous mode look amateurish by comparison - but we stand on the shoulders of giants.
One could get around that by using the low goal and using the ground as part of your floor intake… the agitator randomization ensures that one can’t do such consistently. But we’re splitting hairs.