Team 1540 Flaming Chickens | 2025 Build Thread

Welcome to Team 1540’s 2025 Open Alliance build thread!

We’re excited to be back for another year, hopefully better than the last! This year, we aim to post more often and regularly, with more photos and videos, and fewer blocks of text.

Team 1844 The Eggineers

This year, we started a JV team for new members, and those seeking a lower commitment. As of now, we don’t plan to share as much from 1844 as from 1540, to allow them to focus on developing their technical skills. However, this isn’t a strict rule, and we may share their progress based on interest.

Links

Website | GitHub | YouTube | Instagram

20 Likes

Kickoff Recap

Our kickoff started early in the morning, with a communal potluck breakfast. After this, and watching the kickoff livestream, we broke out into small groups to understand the game, and later into our school’s two teams to create priority lists and robot and mechanism concepts. While we were doing this a group of parent volunteers were building the field elements, which allowed us to have them done before the end of the day!

This season, we tried a couple of new systems at kickoff. Firstly, rather than working through the game as a whole team, we decided to break down into groups of 7-8 to go over our kickoff worksheet, heavily inspired by 6328. This seemed to work well, as it allowed for more of the team to directly contribute and gain a more clear understanding of the game. Secondly, we decided to break down game tasks into separate MSCW lists for 1540 and 1844. This splits all game tasks into categories of Must, Should, Could, and Won’t, based on how relevant they seemed to our season goals. In the past, we’ve typically ranked every action based on priority, but this has left exactly what tasks we want to perform ambiguous and led to us making some robots beyond our capabilities. We hope that having clear tiers of priority will make it easier to cut or postpone certain tasks if we find ourselves out of our depth or falling behind.

Points Analysis

When all coral is scored to maximize points, the majority of points come from Level 4, however they are relatively evenly distributed. One particularly interesting thing is that given scoring in the Trough will likely be significantly easier than any other level, and that it is worth 28% of points scored optimally, this could be a very strong year for low dumpers.

When scoring all algae on the field, the most possible points come from scoring them all in the processor, though this will likely only result in a total benefit of two points, as we found it is very easy for the human player to score in the net. This will only be a concern based on how close to the match is, or is predicted to be, as if you are outscoring your opponents, the slight decrease in points gained may be worth it for an overall higher score. This logic also translates to only scoring the algae on your side of the field, though with lower values.

Overall, this shows that in an alliance going for the maximum score, coral will make up the majority of points. However, algae certainly isn’t negligible, as it makes up around a third of those points when scoring all algae.

1540 Concepts

As 1540 is aiming high this season, our priorities list reflects this. We’re aiming to be able to be the primary scoring robot on most alliances throughout the season, with the ability to achieve a six ranking point match as close to solo as possible.

It’s quite ambitious, just like our season goals. However, as this is one of our strongest student groups in recent history, we’re confident that we can pull it off. We’d like to focus on coral, as it has the highest potential for points, as well as the Coral RP. This list is currently incomplete, such as having no form of intaking in Must, because we want to verify many with prototyping and more in-depth game analysis before making a final decision on some tasks.

Initial Robot Concepts

We’ve initially gravitated towards two primary robot archetypes. The first is heavily inspired by 1678 in 2019, with an arm that pivots 180 degrees on an elevator. The primary change from this design would be a wrist on the end of the arm so that we can pick up fallen coral, while still placing them vertically, and making the arm long enough to reach the ground. This would allow us to cycle back and forth from the loading station to the reef without turning, saving some cycle time. The primary disadvantage is the mechanical complexity, alongside having three degrees of freedom, which may be more than necessary.

The second involves an intake 90 degrees from our scoring mechanism. This would allow us to pick up game pieces from the floor, and then pass them off to an arm on an elevator, likely similar to 3647’s 2019 robot. This concept means that we won’t need a wrist to properly align the game pieces, as well as one fewer degree of freedom on the scoring zone. Its main drawback is that we need a separate intake mechanism, rather than one that can also be used to score.

One of the biggest issues we’ve found is ensuring that both of these can properly hang on a deep cage. We believe one of the easiest ways to do this will be to grab the cage and then rotate up, similarly to 148 in 2010. However, it’s currently unclear whether or not we can properly integrate this into our initial concepts.

1844 Concepts

1844’s Season Goals:

As Catlin Gabel’s new, introductory robotics team, 1844 had more limitations with fewer (and greener) members. As a result, we set lower goals for ourselves. However, due to our performance in Oregon Bunnybots and potential as a team, we believe our goals are still achievable, especially with help from 1540 and the shared workspace.

This discussion revealed we primarily wanted to focus on algae, with the ability to score coral in the first level of the reef. This was because we believe that most teams will focus on scoring coral, leaving algae to be completed in the background, making a robot that can score the net and processor well a valuable partner. In addition, we noticed that an algae in the net is worth the same as a level 3 coral. However, there is also significantly less algae on the field, meaning it is very possible that all algae will be scored before the end of the match. The ability to score coral would allow us to continue playing offense through the remaining portion. Scoring coral in L1 can also be useful for autos as they contribute to the RP and are worth more during auto.

Initial Robot Concepts

We also came up with several early robot concepts. The first of these involves an intake for both algae and coral, by using one set of rollers to guide the algae into the robot, and another to hold the coral at the end. The algae would be fed into a shooter to launch it into the net, or outtaken into the processor. Coral would then be outtaken into the lowest level of the reef, while an elevator with a hook can latch onto a shallow cage. The primary flaw in this design is a lack of a way to dislodge algae from the upper levels of the reef, though the climbing hook may be abe to do this with proper planning.

|598x796.8594506110908

The second concept involves a combined intake/shooter for both algae and coral. This would be placed on a pivot, which could then reach to the ground to intake algae, or go to the loading station for coral. Once again, the main issue with this design is the lack of a way to remove algae from higher on the reef, and that it lacks a climb.

|624x831.3627526070819

|624x831.9999999999999

The third concept involves a slapdown intake for algae into a shooter, heavily inspired by 254 in 2014. This would also have a climber in a box, which if angled properly or allowed to pivot, could also be used to remove algae from the reef. It would also have an intake on a pivot on the front of the robot, in order to grab and place coral on the lowest level. The primary issue with this concept is the sheer number of separate mechanisms, and the potential lack of space that comes with that.

The final concept involves an intake for algae on an elevator (for both L1 and L2 algae, perhaps L3) feeding into a flywheel shooter. This would also have a place on top of the shooter to store one coral for the auto RP. This design works very well in concept for meeting 1844’s Must section, but leaves no room for any of the mechanisms in the Should section, with the existing mechanisms already taking up the entire space.

6 Likes

Initial Prototyping

The last few days, we’ve mostly been focusing on prototyping and testing how game pieces work, alongside some initial design for our scouting system.

BunnyBots 2022

Each season, we host an off-season event called BunnyBots to allow new members to get experience building a robot before the build season, and we noticed that the coral looked very similar to the tubes used in that game. As such, we wanted to share some of the resources from the game, and a couple of our robot CADs. If any other teams who participated are also interested in sharing their robot designs, that would be very near.

Game Manual

Event Livestream

Eliminations in particular have some really interesting robot concepts, especially 2471 and 2521.

Roller Intake Robot CAD

Clawed Robot CAD

Apologies for the mess the CADs are, we never expected to need to share them publicly. The roller intake generally worked better for us, though it did have some issues with jamming. The claw was a little too loose and the shape was slightly off to grip the tube properly, so we wouldn’t recommend taking the design directly, but the concept could work.

Coral Intaking

Most of our prototyping so far has focused on intaking coral. Compliant wheels seemed to work pretty well, especially intaking the coral parallel to the rollers. Perpendicular did work as well, though our prototype had some large dead zones.

We also did some prototyping of using mecanum wheels to vector the coral into a central position. We had some success with moving coral placed perpendicular to the robot, though it was difficult to get it to turn properly into the intake.

Deep Climb

We’ve had a decent level of success with the deep climb. The prototype involved “hugging” the cage with 2x4 slabs of wood, one on the opposite side of the cage from the robot, and the other on the side closest to the robot, we then used string attached to the far side to tension the system and lever it off the ground.

We’ve found that to achieve deep climb, the robot does not need to go fully sideways as we originally thought, but rather just get lifted slightly above the ground. We have been looking at the 2010 for inspiration for climbs.

Algae Shooting

As 1844 is interested in shooting algae into the net, we’ve also done some testing with a simple launcher. We initially tested with a design that combined the intake and shooter, however it failed to generate enough power to shoot very far even with a motor. We’re hoping to try with a more standard hooded flywheel shooter, of which there is a lot of information already in FRC and will be better for our goals. We powered the shooter with a drill at first and it went surprisingly far.

Scouting System Design

We’ve also started designing our scouting system for the season. Quantitative scouts will enter data across four pages, for Auto, Teleop, Endgame, and Post-Match. They will primarily be collecting where each team scores, and how often, throughout the match. They will collect some qualitative data at the end of the match, in rating each team from 1 to 5 in terms of driver skill, where 3 is average, 1 is very bad driving, and 5 is exceptionally good, all in the context of the event, and notes on anything interesting they did during the match.

Qualitative scouting will be collecting, well, qualitative data. This will primarily be in the form of notes, but we also have a couple of other systems. Firstly, we’ll be rating each team on an alliance with the best team getting a 3, the second best getting a 2, and the worst getting a 1. We’ve used this for the last two years, and found it gives us very accurate bases to start our picklists, and acts as a decent substitute for qualitative data. In addition, we’ll be having our scouts track which robots played defense against which other robots, and how effective this was. We hope this will allow us to make more targeted defense picks, so that we’ll be able to be a specific alliance we’ll likely play with than a generally fine pick who may struggle against the biggest threats. Last season, we tried applying roles to robots, so what could track which teams played in what way. However, this was difficult for many quantitative scouts to deal with, as they often didn’t have enough game knowledge to properly identify strategies, so we’ll be moving this to our qualitative scouts this year. Finally, auto paths. We’ll track which teams have the programming for a specific path, beyond how many pieces they score in auto.

For pitscouting, we believe that any relevant data can be collected during matches, so all we will be asking scouts to do is get a picture of each robot.

In order to increase scout engagement, on Friday evenings, we’ll be offering anyone interested to compare teams in a series of one versus one matches, where they pick which is better in each pairing. This will then be translated into a list that will be aggregated and used in picklist meetings, to help give the wider team a voice without needing to stay as late for the meetings themselves. We aren’t entirely sure how to get this to work, but we want to make this somehow related to alliance selection, and the person with the most accurate list will win some prize.

10 Likes

Even More Prototyping

Some more results of our testing with game pieces!

Cycle Analysis

One of the biggest questions we’ve had this season is how many cycles the best teams can get per match, and how we can then match this. To do this, we took the Elite category cycles from 2019, doubled this to account for a slightly shorter cycle, better motors and swerve, and added five for auto, as this seems to be close to the physical limit. This got us 19.2 cycles per match.

We then used this as a baseline to test how much time we would lose from turning in each cycle, to determine the value of pickup and scoring on opposite sides of the robot, which is something we’ve been debating. To do this, we plugged the cycles into PathPlanner, one driving straight and one rotating 180 degrees. The path closest to the coral zone loses 0.23 seconds, the side closest to the driver station loses 0.05 seconds, the far diagonal time loses 0.01 seconds, and the far side gains 0.02 seconds when not spinning. This means that the average time lost per half-cycle is 0.0675 seconds, or 2.7 seconds across a 20 cycle match. We’re still not exactly sure how this should influence our design, since although the time we gain by intaking and scoring on opposite sides seems insignificant, there are other factors like reduced driving difficulty (for both manual and automatic alignment) to consider. However, it’s nice for us to have some numbers to point to.

Coral Intaking

We tried having a roller made up of a solid block of compliant wheels, which would then lift over the bumper. This didn’t really work, as the coral just slipped out to the side.

We also tested using a piece of coral as a stand-in for something to rotate the coral as it is intaken. This seemed to work, though was a little inconsistent.

Coral Scoring

The main design we tested involved two rows of rollers with compliant wheels on the top of the coral, while it was just allowed to slide along a slanted bottom. This managed to successfully score on both the vertical and diagonal sections of the reef. However, when this design did miss, it tended to shoot quite far into the reef, possibly getting stuck and losing the game piece for the rest of the match. While we didn’t test scoring in the through, it should work, it would just need to be released less forcefully.

Algae Intaking

We tested a design using two sets of compliant wheels, and a curved structure to fit the algae. This worked quite well, and held onto the algae even when it was repeatedly shaken.

We also tested a design where algae would be bounced off a roller, and then flipped up and over it into the robot. This managed to kick the algae into the air, but wasn’t consistent without something to bounce it off of.

Algae Shooting

Last, but by no means least, we’ve put together a more traditional flywheel shooter for algae. We tried first running it with a drill, which while not as powerful as we would like, looked very promising.

As such, we decided to add some motors, which worked very well. Even with only a motor on the central flywheel, and a drill on the top flywheel, we managed to shoot into the net. We expect that adding a motor to the top flywheel will only increase the power of our shots. Consistency was somewhat of an issue, however we believe that was mostly due to human error.

15 Likes

There was one more coral intaking prototype tested yesterday that didn’t make it into the post above. This one had a single horizontal roller and ramp meant to test vertical rollers for re-orientation. While that didn’t work as well as we hoped, we noticed an unintended side effect:

8 Likes

now the question becomes what happens if there is a second row of wheels

Thanks for posting all of these prototyping videos.

8 Likes

Indeed! I love these non-traditional game pieces. I can’t wait to see the variety of intake designs this year.

We had a rough winter last year and spent the majority of our usual prototyping/design time out of our lab. OA posts and tests (such as yours!) gave us a ton of ideas and confidence to execute them. It’s also a great motivation for students to practice documentation, technical writing, journaling.

3 Likes

Woah nice post with a ton of info!

Jumping down to scouting:

Awesome to hear, may be useful to smaller teams that can’t field that many scouts too!

Roles (or combinations of) will be very important this year IMO. Synergy between alliance partners will be higher up my list of considerations compared to (maybe all?) prior seasons.

Dare I say Pairwise comparisons? :eyes:

7 Likes

Yes! We’re huge fans of Pairwise comparisons and have plans to integrate this form of analysis into our scouting system. Currently our plan is to have separate SVD matrices for different categories (offense coral, offense algae, defense, etc) and have quantitative (match data) scouts do a pairwise comparison of the robot they just scouted and the robot they previously scouted for each / several categories at the end of each match. Our qualitative (alliance data) scouts will also be filling out tier lists for the robots they watched every match.

We greatly appreciate your research and documentation on pairwise comparisons — it’s fascinating (but complex) math which you explain in a clear and (relatively) concise way. We’re excited to checkout your specific implementation of all the math once the backend is open sourced!

2 Likes

Super excited to see what you all can do with the method! Please keep sharing.

Still working on that lol

3 Likes