[GeeTwo]

My biggest personal regret regarding robotics was the team's second year, and my first year (2013 Ultimate Ascent). I was mechanical mentor for our secondary function (that is, climb); my experience with antisubmarine warfare through the 90's and 00's (and before and since) set me up to work around another "more important" function. "Team Foundation" (drive and Frisbee thrower, herein referred to as TF) had built what they felt was a good device to accelerate and spin a Frisbee (and it was great after a few more tweaks). They had not yet figured out how to lift a Frisbee from "the stack" at the bottom of the ramp into the first spinning pneumatic wheel. I verbally pitched a pneumatically-powered "finger" that would push from inside the lowest Frisbee. No one got the idea, and everyone else's idea was much more complex, so I decided to take things literally into my own hands. I walked around the build space, "cutting" aluminum bar through metal fatigue, doing a bit of rough hand bending, and bungeed a working prototype lifter into place on the build frame before the discussion died down. Once people saw it work, it took little time before it was "plan A". TF put in metal bars to mount the cylinder, but they did not replace the metal-fatigue-cut finger. I tried to convince TF that the finger was a prototype that needed to be done right, but they resisted the idea and I GAVE UP TOO QUICKLY.
Fast forward to Bayou: By now, we were putting 95+% of our shots taken from our sweet spot (essentially under the pyramid) into the 3 point (highest) goal. Once we worked out how to load from the station (which was built well outside of specification), we were a real competitor - when nothing went wrong! We had matches where we forgot to close the manual pressure relief valve, and it seemed like everything that could be set wrong was set wrong in at least one match. Despite our extreme case of TRS (twitchy robot syndrome), we managed to rank #20 (to this day, the only time we've ranked in the top 24). We were not selected for playoffs, and might have declined if we had been, because we were not ready to compete. It seems that our jury-rigged pneumatic "lift finger" had jammed beyond any easy repair. Shannon (another mentor), I, and a couple of students worked in the pits for about an hour trying to fix the lifter mechanism. After playoffs had run through a full cycle and we were nowhere near finishing, we called off the repair and started loading the trailer. It was literally well over a year (summer 2014) before we revived the Frisbee launcher. These days, our "no mentor prototypes on the competition robot" rule would have ensured replacement by a student-built part, but I've also learned to be more assertive in fixing problems in general.

This is a particular case of the general rule given several times above: don't forget, learn.