On March 8th and 9th the Robowranglers travelled to Amarillo to compete in our first ever District Event. After a few weeks of robot tweaks we came out swinging ending up 18-0 and emerging as Event Champions. How do we feel? Feeling good!
The Road to Amarillo is Long
Not the literal road. I’m talking about the figurative road. Well… the literal road is also pretty long.
<INSERT JOKE about that “getting there by morning” song or whatever it is that everyone keeps talking about. They say it like I should know what’s going on, but I don’t… so I just smile and nod and pretend I’m totally into it too. Haha, I like jokes!>
Since the end of the build season we’ve been hard at work making robot improvements. Notably, we redesigned our end-game mechanism. We also changed our ball scoring mechanism, removing the “over the top” option from our double-scoring claw.
Why’d we change the ball claw? We originally designed the “over the top” scoring option to enable the robot to score into the cargo-ship without ever lifting the elevator. We didn’t think it would add much weight to the claw to do it this way, and it thought of it as essentially a “free” feature.
Unfortunately, our design ended up struggling to handle different size balls without WAAAAAY over-compressing the fat ones (Turns out that +/- 1/2-inch tolerance is a real doozy). We had several fixes in mind, the obvious one being: “optimize the claw to score only out the front” and so that’s what we did. Now we have a backboard with a ton of compliance which easily accommodates all the different size balls, even those ones we’ve heard other teams have seen which are outside the legal spec.
We also played with our handoff-geometry a little bit to speed up our “lock and load” time. We didn’t get this quite right, and actually had to make some tweaks in between matches at Amarillo. More on this below…
We’re really happy with the result. Once we got the final issue with handoff tuned-in, our “ball game” felt solid. In particular I’m very happy with the ball-scoring motion at the cargo-ship.
As I said before, We originally prioritized an “after the buzzer” climb to the level 3 platform. This mechanism worked utilizing two gas-shocks and a cam-track system. It’s one of the cutest mechanisms ever to be on a 148 robot. The mechanism lifted the robot, tilted it, triggered the “break the ankles” linkage at the top of travel, shoved the robot forward, then lifted the foot off the ground… all in one motion, all powered by springs. RIP.
The new mechanism uses a PTO from our elevator gearbox to shove down the feet using a rack-and-pinion. The feet now have a mini drivetrain which scoots the robot forward until our front wheels are on the platform. Then we “break our ankles” using a mini pneumatic cylinder to get the feet out of the way, continue driving forward until the middle wheels are on, and retract the legs. Boom. Easier done than said. A picture is worth a thousand words, a GIF is worth a thousand memes:
Most of this is automated by software with our operator just stepping through the sequence. We haven’t had a lot of testing time on this yet, so Amarillo was basically one long series of improvements. As we got more comfortable, we kept speeding things up.
This new mechanism is way more versatile than the “pull the pin and cross your fingers” version we had before. We think versatility will be the name of the game when things start getting exciting…
This is our first year in district play, and so this was the first opportunity we utilized “unbag time”. Instead of doing a full practice day at a regional, districts only have a few hours of load-in time on the first day. To make up for this, they give teams up to 6-hours of time to unbag and work on their robots. How did it go?
We did three 2-hour sessions. Each session involved some manufacturing, some assembly of sub-components, and some retrofits of the competition robot. We pre-planned each session and The Boss made sure everyone stayed on task. We tipped the robot backwards, then divided work up by “top of the robot” and “bottom of the robot” tasks. This allowed us to have 6-8 students working on the robot at any given time (in addition to the work being done on tables nearby, the work in the machine shop, and the work at our fabricators.)
Man… talk about a super-power.
We got SO much done in this time. I feel like we could EASILY rebuild the entire machine from top-to-bottom in this time. It doesn’t even seem fair. (Sorry California).
Robowrangler Fun Day
Before the event began, the Robowrangler students did some West Texas exploring! They visited Cadillac Ranch, and put our own unique spin on one of the cars there…
The team also went hiking at Palo Duro Canyon and took the annual selfie. Sadly, I wasn’t able to make it this year (stuck at work).
The Battle for Amarillo
Being new to districts, I didn’t really know how the new schedule worked. I was speaking to a friend about it (from the Michigan District).
“Did you know, we get 5 hours of time to work on the robot Thursday night? The pits are open from 5PM until 10PM. Then we get like 2 hours on Friday morning. This is basically a full practice day, ON TOP of our unbag time.”
“Ah yes, the hidden time no one talks about… load-in time.”
It’s a good thing we had this! As I’ve said, we’ve been a little slow this year and the robot definitely needed the additional quality time…
Durability & Match Turn-times
At a 30 team district event, you get more qualification matches (we ended up getting 12 at Amarillo). This is good!
It also means, you have much less time between matches. This is not so good.
We’ve been concerned how tight the turn-times would be, and wondered if we would have enough time for robot maintenance. So, we spent some serious time improving robustness. We talk about achieving “every match” consistency. This year that meant we’d need to reduce the time it takes to go through “the checklist” or make sure the robot would keep running even if we didn’t always have time to go through the full list.
We spent a lot of time during the lead-up to Amarillo and the event itself working on “don’t let the robot eat itself” fixes. Things like “make sure the elevator cables don’t snag on the new end-game” and “don’t let the intake wires get grabbed by a defender robot.”
Fun fact, during our final unbag time we went to do functional checks on the robot: The end-game mechanism karate chopped a pneumatic valve and ripped the end off a plastic storage tank, then within 10 minutes the elevator snagged and ripped its upper-stage rope loose. Oops.
We wanted to make sure that stuff didn’t happen in Amarillo - even if it meant redoing a ton of wiring and tubing during the “load-in time”.
As many people know, the qualification schedule can make or break a team’s weekend. This is more true in some games than others. At an event with 30ish teams and 12 qualification matches, you expect to play with and against almost everyone at the event.
After watching practice matches, when the schedule came out we were wondering one thing: how many times would we play against 1296 and 3310?
We set the line at 2.5. I chose the “over” - expecting to play against them at least 3 times.
The qualification schedule 148 received at the Amarillo district is the most cake-walk of a schedule that ever cake-walked. While in any given match, on any given day you never know how it will play out… I’ve never seen such a gift from the scheduling gods. I don’t know what karma we spent in exchange, but I’m not complaining… Every time we had to play someone tough (and there were lots of tough robots at this event) we had plenty of help. We didn’t even have very many “tight turn” matches. Some teams had to play almost back-to-back. Other teams had to play against 148, then a few matches later go play against 3310.
We have a saying on our team, which we use when discussing strategy. “The good news is, we get to play with 148 every single match.” (At least, it’s usually good news).
The qualification matches went great. We solo filled a Rocket. We scored a lot of balls. We climbed onto level 3 a bunch. Wooo!
Take the Ws
Since the schedule looked favorable, and after seeing how a few rounds of matches played out: we decided to be a bit conservative.
In this year’s game, most high-performing robots start out the match scoring on the rocket, then if they get defended (which they should expect every match) they switch to “scoring wherever”. Very few teams are able to complete a rocket under even moderate defense.
We had lots of matches where we knew we’d probably win, but where we knew we’d probably not be able to get the rocket RP. What do we do in these situations?
We double-down on securing the win, and taking plenty of time for the climb RP.
“The team who makes the least mistakes is the one that’s gonna come out on top this weekend. Don’t let us talk ourselves into anything stupid out there.”
We tested things. We tweaked our game. We practiced different moves. We didn’t ask our partners to do anything they don’t normally do.
I’m most proud of the fact that we didn’t end up “chasing the rabbit” and just spent the whole weekend doing meat-and-potatoes improvement of our game. Continuous Improvement is the mantra. “Sometimes it seems like the only thing we’re good at, is getting better.”
It was kind of interesting… 12 or so of the teams at Amarillo had already played in El Paso the week before. I’m not used to coaching a week 2 event where I feel like the only guy who hasn’t played yet. As a result, most teams knew “the drill” while the Robowranglers were still figuring out how not to trip all over ourselves.
We struggled with two robot functions during this event. I know lots of teams enjoy hearing how 148 overcomes challenges, so I’ll tell you our two (self-inflicted) struggles:
The first one is due to a design error during iteration. When we implemented the new upper ball-claw, the front roller interfered with how we hold a disc. One of our DEMAND design goals was “we need to be able to hold a disc while legally in starting configuration”. This was a bummer. For Amarillo we added some smaller “fingers” inside the disc-grabber beak, and unfortunately didn’t get enough time to test and tune these before the event. (We did some basic testing, but didn’t do a full “practice like you play” test… oops.) We spent most of the event tuning these in, while being very inconsistent in auton. (Lots of dropped discs, lots of “not scored” discs, etc). By the end of the event, we had a solution which was “adequate.” Of course… adequate isn’t good enough. (We’re already tweaking some things so we can use the primary beak fingers to hold the disc at our next event.)
See that picture above? The robot looks beautiful! So fast, such a good intake! In fact, the intake is so good, sometimes after the ball gets shoved into the robot, then the intake grabs the ball and shoves it back out of the robot! Which… when I think about it… isn’t very good at all.
We call this a “looper”. A ball that comes into the robot, but then gets grabbed by the back of the intake roller, rolled over the top of the intake and shoved out between the bottom of the disc mechanism. Oops.
This was the second thing we struggled with during Amarillo. We tried a bunch of different things: making sure the elevator wasn’t lifting up, making sure the drivers weren’t doing something dumb (they weren’t), tweaking the sequence the drivers use to pickup a ball. In the end, we added a free-spinning roller to the bottom of the disc mechanism. This prevented the ball from “grabbing” and caused it to “spin in place” a little instead of getting squeezed out over the top of the intake. (This all happens so fast, it just seems like the ball pops straight up into the ball-claw, like it should). By the end of Amarillo we didn’t have any hand-off issues. (Yay!) But it was annoying seeing us botch ball pickups in some of our matches.
Pick a Color, Any Color
For the first time, the Robowranglers brought coloring books as give-aways. We even let you take a crayon from our selection. “Pick whatever color you want.”
Thankfully our “slow and steady” strategy paid off, and we ended up seeded #1. We selected our #TeamIFI bros (and close neighbors) FRC#3310. This was an easy choice: their stats were fantastic, and we already had the opportunity to play two qualification matches with them.
Plus, as they say…
“Once you go Blackhawk, you never go Backhawk.”
Our elimination run was fairly textbook. We ran a very conservative “split the field, take the points” strategy. 3310 went left, 148 went right. Our wonderful 3rd partner (FRC#2657) scored a few points, then played positional defense. In half the matches 148 climbed onto hab 3, in the other half 3310 did.
We faced moderate defense in the semi-finals and finals and mostly employed the “they can’t stop both of us… just let it happen” strategy (again, keeping things very conservative, avoid mistakes). In a few matches we did some cross-over moves, drawing the defender back and forth between us.
“Did you… did you bring the defender over to our side, then leave them with us?”
“I mean… we just made an introduction: Hello, have you met my friend 3310? Why don’t you guys get to know each other! I think you have a lot in common… you guys talk for a while, we’re going to go back over to our side of the field and keep scoring.”
“Maybe they got the two black robots confused?”
Fun fact, the last time I coached with Paul Copioli was on Einstein in 2008! This time, things felt a lot less stressful…
We ended up having one fun moment in Final 2. We lost a significant amount of robot functionality very early in that match. At the time, we thought it might have something to do with the pneumatic systems, but struggled to diagnose the problem mid-match. What did we do? We switched to fullback and just tried to protect 3310.
It turned out that the XBOX controller used by our operator had an issue mid-match. (Ugh. I hate losing matches to XBOX controllers.) Thankfully, 3310 and 2657 had more than enough scoring juice to pull out the win. I mean… our elite level defense on 1817 definitely turned the tide. Definitely.
We also were honored to be recognized with the Quality Award by the judges! We love that award! We hope the coloring books played a role in their judgement of our quality.
The team did our typical debrief at the hotel afterwards. The message is the same as always.
Did you Have Fun?
Did you Make Friends?
Did you Chase Excellence?
I’m proud that everyone seems to be focused on the bigger picture. As The Boss said: “this is just one step on a much larger journey.” Continuous Improvement is a way of life in the House of Black, and we’re already making tweaks to every aspect of our team (not just the robot).
My favorite part of this weekend?
As I’ve discussed, this is a VERY inexperienced Robowrangler team. We have LOTS of rookies.
We HAD lots of rookies.
“All the rookies, raise your hands… You’re NOT rookies anymore! You’ve fired shots in anger, and been through your first competition. If you were holding back or hesitating at all, that time has past. Step up, take ownership of this season, and let’s keep pushing.”
Next week, Greenville ISD is hosting a District Event. The Robowranglers are not competing there. We (of course) have people volunteering, and we’re helping with setup/takedown. It’s gonna be a little weird having our normal meetings in our lab, while there’s an FRC event happening down the hall…
Why aren’t we competing in Greenville? Honestly - we get enough grief as it is. We don’t need anyone saying we’re getting special treatment as the “host team”. Our season schedule worked out great as is.
We have another week and a half until Dallas. Then we go straight into State Championship (if we qualify). After that, World Championship is right around the corner.
We’re not good enough.
We haven’t had as much practice as we’d like. We haven’t had a ton of time for auton tuning. Amarillo represented more “tune the robot” time than we’ve had at home.
We have some tweaks to make to our game. We have some half-finished subsystems to add to the robot.
We’re not good enough. I got into an argument with a friend over this. I didn’t say we’re not good, I said we’re not good enough. We want to keep pushing, and we’ve got plenty of time left to push…
The big question (as always)…
How do we balance our limits and our wants to get the best possible outcome?
I’ll let you know, once I find out. On to Dallas. This is Fine.