Sweaty Palms and Busted Knuckles

Sittin' pretty.

We had a competition this weekend! Let's just jump straight into it all.

Competition Prep

In a word: rushed.

Thursday, during packing night - the robot was on the practice field, continuing some autonomous testing and other software tuning. Back in the lab, after a massive stretch of optional meetings with the lab open for more than 15 days straight, people were understandably tired. We grabbed the tools we knew we used most, grabbed the remaining parts for the climber and wheel of fortune mechanisms, grabbed totes, boxes, bags, it was a mess. Now, I may slightly just a little bit be an itsy, wee bit OCD and a neat freak, so obviously, I would take more time to complain about this if I could, but, I'm already late with this post and just want to get our weekend information out there! 

So take this as a lesson, for so many things - take the time to prepare correctly. Our messy packing led to a messy pit, led to messy repairs, missing parts and tools, and a generally frantic mood every time the robot was in the pit.

As a second lesson, update your post-match checklist: number 1 - Close your eyes, take a deep breath.

Onto the weekend. Glacier Peak is usually pretty nice to us.

(Head coach laughed at this. For her apparently Glacier Peak has never been kind. Let's see if we can remedy that...)

Volunteering to help build the field also grants some interesting photo opportunities.
A completely empty pit, Friday after pits closed.

How We Ride

On horses. Like Knights. Get it? Ok, I'm leaving...

Our team runs in a few, well-connected silos at competition. Our storefront team (and usually Chairmans presenters) have shifts in the pit, talking with judges, parents, other teams, and other guests. Our scouts fill the stands, marking quantitative and qualitative data about every robot in every match. Our drive team handles the match strategy and execution, and our pit crew owns the state of the robot. The scouts, and maybe storefront, are generally the only people on the team who actually know what our rank and overall performance is. Drive team isn't allowed to know, and pit crew ain't got time to know. I work mostly with pit crew, so this weekend report will have little to no information about rank, strategy, match play or etc, and much more about why 3-D printed parts in high-stress/strain locations are bad. Mmkay?

Truth be told, I looked up on Blue Alliance today to verify what I heard was our final ranking position at number 2. Legit, didn't know, didn't care. The other side of the fence is apparently the same way - my wife works with our scouting team, she doesn't know, doesn't care about the state of the robot, she just looks at what numbers we show on the field (cause, that's what all the other teams see...). Further truth be told, I think that's the best way to handle it. The final slide of our team-wide competition meeting was 3 simple words:

Do your task.

Scouts gonna scout, drivers gonna drive, Randy's gonna volunteer. Focus on your task, focus on what you can control, and trust your teammates to handle their tasks.

How Saturday Went

Frantically.

I feel like I need a thesaurus.

Saturday we finished up inspection (we... arrived late on Friday...) and ran 10 of our qualification matches. One match, all of our systems were fully operational (more on that later, but you can probably guess which match that was...). The other nine matches, something or another went haywire.

Our process for the pits and each match is generally the same - we have a priority sorted post/pre-match checklist, and we enact any immediate repairs. Often, at the same time. It's a small robot, yet we've been able to have... all 5 allowed people in the pit (wink?) on the robot all at once. Match 1 got things started off just right. After weeks of running the robot in the lab and on the practice field, our indexer gearbox roller interface decided that now was a good time to explode. We caught a power cell in the indexer on our battery connector, tore the ball, and sheared the teeth on one of our 3D printed gears. The robot came back, the crew descended on it, we swapped plastic for metal, ran through our remaining checks, and sent the bot on its way.

Spent the day in the pits with some of these awesome CyberKnights.
Cuddly CyberKnights.
Way too comfortable CyberKnights...

After our second match, we realized that we hadn't found all the issues caused during the first match. We replaced another plastic gear with a metal one, and replaced two additional cracked, warped, and... melted(?) 3D printed roller pulleys like-for-like. Again, we ran through as much of the checklist as we could in the time we had.

Our checklist is prioritized on a few parameters - what do we <need> to play next round, bumpers, battery, starting configuration, etc. Next up - what do we <need> for successful autonomous, then main cycles, then "specialties", in this case, end-game climbing. We perform an electrical pull test on all the exposed connectors, check any mechanical parts that extend beyond the frame perimeter, and do hand/poke check on all the rollers, shooter wheel, chains, belts, really anything that moves and should move, or is fastened and should stay fastened. If we have enough time, we'll tether the bot and run a software suite of functional checks - running a ball through a cycle, running the climber, and checking the swerve drive for any anomalies.

Our third, fourth, and fifth matches happened in rapid succession, twice not even returning to the pits. After our third match, the front roller stopped rotating - we found a suspected fault in the motor itself and during the 4th match prepped its replacement. We also found that our climber did not engage, caused by a re-assembly error after fixing one of the roller shafts after match 2.

Just before our 5th match, we were able to swap the intake motor and re-pin our climber PTO shaft. Our 5th match was glorious. Match 26 at Glacier Peak, 9 power cells scored in autonomous, and power cell after power cell finding its home in the Outer and Inner ports during the tele-operated period. 55 in total, as told to me by head coach. Finish it off with a double climb (unbalanced, but, oh well...). That was fun. Lets do that again!

Sometime during the weekend, the fruit wars were continued...
I don't know who struck first, but I do know, that I didn't eat either of these.

The following match we started having Limelight issues, and failed our climb. The return to the pits showed that we had sheared more 3D printed teeth, this time on our climber brake. We had one spare back at our shop, and another mentor grabbed their keys to fetch it.

The rest of the afternoon was spent chasing climber modifications and pneumatic issues. We installed 2 additional air tanks, believing there was an air pressure issue with the brakes. We installed new parts after every match to bring our second climber tower closer to completion. By the end of the night, tired and weary, we apparently had won all but one match, but had fixed or replaced at least one part on every subsystem, except chassis.

How Sunday Went

Better!

But there's always room for improvement.

I've been hanginaround, this town on the corner... bar... shield generator...

The robot-side team goal for Sunday was pretty straight forward - two matches, no issues, make us look good before alliance selection, and get a rest before elimination rounds. We worked all morning, a skeleton crew stayed during opening ceremonies, and we finished getting the second climber arm attached, plumbed, and software tested. We were ready to go for our two matches, and both matches went super smooth! The robot returned from the pits, we ran through our checklist, and gave back a shiny, tip-top shape bot.

Before lunch, Alliance Selection. Everyone watching the top 8 teams, staring at clipboards and phones, pondering who would pick whom. I, naturally, was in the pits with the robot. Apparently we got picked up by the number 1 alliance. Cool beans. The robot was patched up and ready to go.

Ish.

We had a small problem with the climber. We now had both sides fully up and running - two arms, two brakes, fully operational. Except the brake material, again, was printed plastic. We had sheared 2 teeth on both sides, so we had contact of maybe 4 teeth in total to hold the robot. Going into quarter finals, of course we made sure to tell our drive team about the perils of climbing... oops. (Oh man... there's actually a grammatical character to intone sarcasm... what the heck is it again..?  Oohhh backwards question mark - ⸮ - irony punctuation. Uh, anyway) We actually told them before the second quarter final. We also continued our Limelight issues, not getting the same accuracy as we had previously, which also impacted our autonomous. We started running the simple auto, guaranteeing a few points, opposed to high risk high reward.

Semi-finals, we showed off. Victory spins. You know it. Definitely not because we blew a pneumatic supply line and couldn't do anything at all. No climb, no ball pickup, all we could do was drive. Might as well drive! A single piece re-plumbed, and another note taken and we were back in business.

Yep, I make the kids do paperwork. What broke, what was changed, what was fixed.
We're going to be adding all this to the wiki so over the summer I can remember just how long a day this was...

I feel like eliminations went quickly, again I apologize, I know we played against some awesome teams, I was just busy staring at our bot, watching every subsystem move, trying to capture any oddities to look into after the match. Thank you Glacier Peak teams for the fun matches!

Thank you volunteers for pointing out our drive coach's clipboard! (⸮Which was most certainly not left inside the robot...⸮)

Finals I did pay attention. We played against the host school, 2930 Sonic Squirrels, a fellow 4900 number with 4918 the Roboctopi, and an old nemesis... 2976 Spartabots. <Fade to black> The year was 2018, Houston Texas, Minute Maid Park. CyberKnights, joined by another swervy-groovy-lifty machine stood opposite the regal green Spartabots. We lost that day, but vowed revenge. <Return to present> 2020, Glacier Peak, Finals Match 1. The time for our revenge has come!

We failed. For a few minutes. Then we succeeded? That was... weird...

I can't say much about the decision on the field, and nor can I do anything about the decision on the field. I watched the robot come off the field, joined my pit crew, and got the robot ready for the next match. I did my task. This time it benefited us, other times, it hasn't or won't. It is what it is, and if it's outside of your control, just close your eyes, take a breath, and get ready for the next one.

Finals were great. Fun matches to watch, difficult matches to watch. Both times we came out of autonomous behind - we would have to outscore them in the tele-operated period.(Note: Win the autonomous game.) We were fully prepared to climb - we had 4 teeth holding the climber and one spare. Either 25 points, or 5. (Note: win the end game.) We had some of the hardest defense played against us in the first match, so we planned for the second - drive smart. (Note: always drive smart.) Our drivers, with the limited visibility this field offers, tried to stay protected but as far forward as possible - the best for our current shooter tuning, and battery capacity. Oh that was fun - right before finals we found out that our pit didn't have power and our batteries weren't charging. (Note: charged batteries work better than not charged batteries.) In the end, we stayed safe, drew a few fouls from our defender, and then went to close out with a climb, and finally, after all competition - we broke something on our drive train (suspect the encoder...). One weekend, and something now broke on every single sub-system. (Ok, nothing mechanically broke on the shooter, but the Limelight is part of the shooter subsystem, and I blame software entirely. :-D <3 )

That being said, we did score enough to take the win, and took the event. The team cheered, cried, hugged, screamed. I stood there, thinking about our next match, just 5 days away. I may be broken in this regard - I was happy, but the weekend was far too busy for me. Goals for this week - make pit crews lives more boring.

Thanks to our awesome partners!

Oh man, I didn't even thank or congratulate our allies. 2910 Jack in the Bot - once more, a fast, swerv-y, groov-y, bally picky uppy and shooty machine! You've been so fun to see these past few years and your robot once again looks amazing! 4173 IMVERT - while we won't be attending an event hosted in your house this year, we are glad to have you on our alliance! Thanks for working well with us today, and good luck both teams at your next events! (Spoiler... 2910 is also going to West Valley...)

Also, as head referee and as someone involved at a high level in Unified Robotics for its entire existence - I would like to send a warm congratulations to team 4131 - Glacier Peak Chairmans Champions - Iron Patriots. I enjoyed seeing you step up during all of our Unified events, and if you have this same passion and energy for all your outreach and robot programs, this was definitely well deserved. Congratulations!

On Deck

5 days until West Valley. Not a lot of turnaround time. Plan is reliability, reliability, some spare parts, and reliability.

Let's not do this again.

Naturally, the robot was unloaded Monday afternoon. Software ran a systems check, taking a look at the drive train, trying to identify our issues in the final match. 5 minutes after they finished, (and 4 bolts!!??) the robot was in two pieces.

Four bolts. Yep. Just four.

I've decided I actually like this aspect of no-bag. Normally, we would rush a 6-hour fix or upgrade session, and pray our previous efforts and reliability were retained or improved. Today, we actually disassembled most of the robot, and were able to understand the ramifications of our designs and decisions. Bearings getting coated in dust, wear patterns in aluminum, bolts which seemed, a little too loose on disassembly. Good notes on how it all looks when taken apart should help us improve it when it goes back together.

Amperage, voltage, pressure, angle, you name it, we might log it.
After competition we pour over logs, checking for why we had driver reported issues, or identify faulty components.

In a 6-hour unbag, we wouldn't take apart our shooter and identify where and how the bearings and shafts were wearing.
I really enjoy that we can add another level to our post-competition meetings.

While we had a fantastic showing, and displayed just how competitive our robot can be, we can't rest on the laurels of one meet. Our plan is to continue developing this robot for each competition, lightening what we can, adding our final subsystem - wheel of fortune, and improving our cycle times and accuracy. Software, drive team, and pit crew should and will be putting hours into each part of this robot between now and Houston.

Hope you all had a good, hard, fun, inspiring, week one competition.

Mentor, reviewing the blog post - "You didn't put a quote at the end, did you..."

I'm tired.

-B