It’s day 2 and we love the game. Following the plan we outlined above, we think we had a very successful kickoff weekend.
Day 1 (1/8)
We had some of the team virtual, some of the team in one classroom, and some of the team in another classroom. Each room (plus the virtuals!) were all connected via a Google Meet call, and we alternated between the rooms as to who was reading off the next part of the manual.
We estimated the “max score” to be around 143 points. This isn’t really the absolute maximum score as there’s no limit to how many cargo you can score in teleop, but it’s the max we realistically expect teams to get. Maybe by DCMP or CMP this score will be broken, but we’ll have to see!
We then split off into 3 groups (one group per room, and the virtual group) and each group made a list of
- Our robot will do X
- Our robot will not do Y
- It would be nice if our robot could do Z
We then shared those lists with each other, and that brought us to the end of our meeting on Saturday.
Day 2 (1/9)
Today we met again and decided to compare our 3 lists and decide, as an entire team, on what our robot will, won’t, and might do.
We decided on this:
Our robot will:
- Drive in auto
- Drive over bumps
- Drive below the low rung
- Score in the low goal
- Score in the low goal in auto
- Be able to carry 2 cargo at once
- Pick up cargo from the ground
- Score from fixed locations
- In this case, we are talking about a wall shot from against the hub
- Hang on the lower & middle rung
- Intake and outtake cargo in opposite directions
Our robot will not:
- Shoot from anywhere on the field
- Hang on the high or traversal rung
- Monkeybar from rung to rung
- Break the rules
It would be nice if our robot could…
- Score in the high goal
- Score in the high goal in auto
- Score from the launchpad
- Catch cargo from the HP
- Feed cargo to the HP
- Have the HP score in auto
We recognize that all of our “nice-to-haves” are unlikely, but we don’t want to immediately rule them out without further prototyping data from both ourselves and the community.
The low rung
We believe that fitting under the low rung is extremely important for us. If an alliance has 3 robots that are all trying to climb in the last 30 seconds, and none of them can fit under the low rung, that means they all must enter from the side, rotate 90 degrees, and then align to their respective rungs… Sounds like a major, major headache. To give our alliance room to breathe and a larger window of execution, we see it as highly important to be able to fit under the low rung.
Cycles & high goal
We are unsure about how fast high goal cycles can be. Cargo may take 4-10 seconds to pass through the hub’s upper goal, and our (extremely nonscientific and low sample size) testing is that a ball dropping out of the high goal takes about 4ish seconds to settle. Intaking a bouncing ball is not something we have done before, and not something we have seen teams do reliably, so it is an unknown factor for us. We think that the ability to catch a bouncing ball is probably rather important, but we aren’t sure how to do it, and will wait to see more results from the community first, as it’s not a high priority for us (in fact, not even on our lists). Low goal depositing the cargo essentially immediately onto the floor for pickup should be a relatively significant difference in cycle opportunity.
The cycles seem very similar in distance to 2019 thus far. In 2019, the good teams were hitting 12-14 cycles, and the elites were hitting 16-20 cycles in a match. I would hazard a guess that cycles could be maybe 50% slower than 2019 cycles, meaning a good team could hit 6-7 cycles in a match (where each cycle is 2 cargo), or 12-14 cargo, which is nearly a solo ranking point depending on how well your auto goes.
We began sketching out a few different basic ideas. Many of them resemble cargo passthroughs from 2019, such as the Everybot, 5406, or 1323. We view having opposite side scoring/intaking as massively important for our cycle times - turning is slow, and you should avoid it. For the most part, balls will be sorta-kinda scattered semi-randomly around the hub - you can score a lot by driving in mostly straight lines between the hub fender and the cargo on the field.
This also allows us the opportunity for a possible 4 ball auto, which would be a significant boost to our likelihood of a cargo RP in quals. We have not discussed auto too much in detail yet, but I think a safe goal would be a reliable 3 ball auto, and our reach goal could be a 4 ball auto. The team has not successfully used gyros or encoders in the past, so I’ve been introducing some of the veteran programmers to WPILib’s documentation on Ramsete and odometry in preparation for the season. We are getting a ¼ field carpet on Sunday, which we’ll use with the 2020 robot to develop paths starting early in the season (hopefully next week?).
As for the climb, we briefly looked at our COTS options, but realized we only need 1 stage to reach the middle rung. Because of that, we’ll likely be building our own single stage climber, as we don’t expect it to be a significant headache, and it will save us some money. That’s as far as we got on that front.
We specced out approximately how many motors we’ll need for each subsystem.
- Drivetrain: 4-6 NEOs
- Intake: 1 NEO or 1-2 NEO 550s
- NEO or NEO 550 depends on our roller/wheel size and drivetrain speed; we need the surface speed of the intake to be greater than the drivetrain’s velocity in order to safely pick up the game piece
- Ball tube / hopper: 1 NEO
- Shooter: 1-2 NEOs
- For the low goal, this is likely just a tiny bit of extra oomph to make sure it gets out of the hopper. It would be very slow so as not to whip it into the low goal that hard.
- Climber: 1-2 NEOs
- Total: 8-13 motors, mostly all NEOs, maybe a NEO 550 or 2
Quick intake test
We briefly powered on our 2020 robot to test how the intake worked on the ball. The ball seems much harder to grip than the 2020 balls, but it may behave differently when it’s on carpet. https://photos.app.goo.gl/ns1Qouwgs1WhVxnFA
While reading the manual and doing all of that, we had someone taking summary notes in a doc, which you can view here: Copy of Kickoff 2022 Notes - Google Docs
I gave the 2713 students some prior years to take knowledge from. Specifically, for climbers, we can learn a lot from 2013, 2016, 2018, and 2020 (special thank you to Everybot). As for ball mechanisms, our closest game pieces were 2016, 2012, 2020, and 2019, in that order, but they’re all vaguely similar.
If anyone has any questions please feel free to reach out via CD or Discord!