2974 Walton Robotics- Build Blog 2024

[madelinep]

Welcome to the Build Blog for 2974 Walton Robotics!

This is our first build blog, which we’re starting in an effort to give back to the FIRST community and to share our building and programming processes, resources, and updates. We’ve learned from and been inspired by other teams’ build blogs and open source CAD/code, and we want to share in the same way. We also plan to join the Open Alliance in the 2025 season.

CAD: (coming soon)
Software: GitHub

Who we are:
2974 operates out of Walton High School in Marietta, GA, part of the Peachtree District. Our team has about 40 students and the support of many mentors in business and technical fields, teacher coaches, and community sponsors. We have 8 departments: Build, Programming, Outreach, Presentation, Strategy, Communications, Creative, and FTC, where our students mentor two middle school teams. Team members can work across multiple subteams depending on what suits their interests, and the student leads of each of these departments make up our Executive Team. We will be attending the Anderson and Carrollton PCH events, with the hope to qualify for District Champs and then Worlds in April.

[madelinep]

Our plan:
Since Kickoff, we’ve been strategizing, brainstorming, prototyping, and beginning to manufacture. After some testing and multiple iterations of CAD, this is our current plan for mechanisms.

Ground intake

• Our intake will be set behind our bumpers, and it uses roller bars with .2 inches of compression. This is a mixture of two of our earlier intake designs: an intake that would fold over the bumpers of the robot and use stacked rollers/polybelt, and an intake that would have a fold out extension to “touch and own” notes from the ground and bring them underneath the robot into a feed or directly into the shooter. (both CADs pictured below). We prototyped the over the bumper intake using stacked wheel rollers instead of polybelt and discovered that the bottom rollers were good for grabbing and owning game pieces, but were not as good at conveying them between the bars to go into the chassis.

(Current intake)

(Early over the bumper intake design)

(Early bottom intake design)

Top/bottom shooter + climber

• Our shooter is set up with axles above and below the note and will use flywheels and timing belt to settle the note in from the intake and fling it out to score in the speaker. We tested this with .5 inches of compression on the 2 inch height diameter of the note, which seemed to work really well.
• This will function as our amp/trap scoring mechanism as well. In prototyping we were able to use a top/bottom shooter to repeatedly make it into our constructed trap from a few feet away, shooting at about a 70 degree angle.
• We also realized that if we raised the base of our shooter by about 2 inches, we could lower it to hit the chain on the stage and use it as a climber. We’d also get more space/spare weight on the robot without an additional climbing mechanism, so we plan to test out climbing with the shooter. This is a tentative idea and we may go back to one of our other climber possibilities depending on how well this works, especially since we want to make sure we aren’t putting unneeded stress on the design, but we wanted to prioritize the shooter to give our drive team enough practice time before our first competition.

Transition from intake to shooter

• We’ll be using a set of rollers with polybelt and a curved barrier as an extension of the intake to turn the curve from floor pickup to our shooter.
• We briefly planned a full conveyor to smoothen the transition from the intake to the shooter, and we prototyped PVC rollers wrapped with surgical tubing to better grip the note along with a guiding ramp. But when we tested this conveyor it was a bit slower than just continuing the intake and tore up the note slightly, so we decided to have the more direct connection between the intake and shooter instead.
• The conveyor is still the most up in the air. All current designs include some configuration of active wheels with a variety of passive rollers and surfaces. Despite this we are moving forward with the shooter and intake into the manufacturing phase while still conceptualizing the conveyor.

(Current cover/roller setup, polybelt around blue rollers)

(Early 3D printed ramp setup)

(Conveyor prototype)

Electrical board

• We initially tried to keep everything on the top face of the robot, but constraints imposed by the shape of our robot meant that having electronics both above and below the central part of our belly pan was the best way to manage space. We’ve calculated our shooter’s max angle of incline to be 30 degrees above ground if we want to shoot from the center line into the speaker. That angle is low enough that it restricts the maximum height of our electronics underneath it, which limits possible placement of both the battery and the mini PC that goes with our yet-to-be-placed camera (a See3CAM_24CUG).

Bumpers

• We originally made C-shape bumpers that had a continuous piece that covered the back and both sides of the robot and one piece for the front to leave a gap for the notes to go underneath the bumper into the intake. We then revised this design to have all sides of the bumper be the same height above the ground as the piece over the intake instead of having the C-shape piece be lower, because the unevenness would create an anchor for the robot to tip over.
• Since the bumpers are all set to be the same height now, we are making the bumpers into two c-shape pieces (symmetrical) with the seam of the two pieces coming together being on either side of the robot instead of on the front and back of the robot. This is to help reinforce the front of the robot frame and help prevent bending/deformities forming over the course of the season as the robot experiences wear and tear.

[madelinep]

Other ideas:
We explored a few more avenues of CAD/prototyping that we aren’t planning to use on our bot at the moment, and here they are.

Side roller shooter

• We used this prototype to test different sideways note compression levels– .5, 2, 4, and 6 inches. As expected, the highest levels of compression made the shooter significantly less effective. After testing we determined that this style of shooter wouldn’t be able to shoot the note at as much power as we would like, even with motors powering the rollers at full speed because of the amount of contact the note made with both the wheels and the top and bottom plates.

(CAD and prototype of side roller shooter)

Two-wristed arm shooter/intake

• This concept included a pivot point at both the joint between the chassis and the arm towers and between the arm towers and the shooter/intake, which would function as both. The double joint would give the mechanism the flexibility to intake from the ground or the source from both front and back and to shoot into the speaker. But to raise both wrists that much or flip the robot around from intaking on one side to shoot from the same side would take too much time in auto, so the double wrist didn’t really get fleshed out in CAD.
• This design also included pistons between the plates of the shooter/intake, which could be actuated to widen the mouth of the intake when receiving notes from the source instead of the ground, then closed again to shoot. But given our separate ground intake, this isn’t useful because it doesn’t add anything to just a shooter, and the lack of pneumatics simplifies the robot.

(Unfinished double wrist assembly with shooter/intake–has top wrist but not base wrist)

(Pneumatically opening shooter/intake)

Frisbee launcher

• With the frisbee shooter concept, we decided to design a one wheel shooter that uses a hopper system to gather the notes. A horizontal axle with gears at the bottom of the system allows it to pivot around the base of the side plates to aim toward the speaker.

Scissor climber

• This possible climber (one on both sides of the robot) would take up less space than a telescoping climber because of its ability to compress down, and would also have had less weight than the telescope. However, we decided not to go with an additional climber for now because using a single mechanism to both shoot and climb gives us back even more space. It also has the potential to take away extra complexity while allowing us to focus primarily on shooter manufacturing.

Amp/trap mechanism

• Our potential trap mechanism would grab the piece from the shooter and store it in the roller assembly, then rotate and extend to either the trap or amp, and score within the different components by activating the roller.

• We also were considering using two of the same telescoping arms that is the base of our amp/trap mechanism, along with hooks, as a climber.

Alright, that’s the first few weeks of the season summarized. From now on we’ll have regular new posts, whenever we make any interesting discoveries/solve any problems to share about. Feel free to leave any questions or comments!

Special thanks to these teams’ build threads for inspiration throughout the design process:

Team 95 Grasshoppers
Team 111 Wildstang/112 Plus One
Team 3847 Spectrum

[zache]

In previous stages of the build season, our team has traditionally relied on familiar tools like Google Docs and Google Sheets for task conveyance and storage. However this approach encountered challenges this season, leading to delays and a shortfall in prototyping goals. Recognizing the need for a change, we transitioned to using ClickUp in place of the familiar Google suite. This shift proved transformative for our team’s efficiency and organization, particularly in in-person meetings and task delegation. ClickUp’s features, including its timeline capabilities for Gantt charts, have played a pivotal role in enhancing productivity and ensuring that our team stays on track with evolving project requirements. This strategic adjustment has significantly improved our ability to manage tasks, coordinate efforts, and maintain a streamlined workflow throughout the current build season.

[madelinep]

This week’s updates:
New shooter design

• Our former flywheel setup wasn’t going to give the note as much velocity as we would’ve liked, so instead we switched to a Wildstang-inspired shooter and added carbon rollers. This week we finished cutting 14” churros and completed CNC for the side plates.

Intake progress

• We’ve finished cutting and tapping intake axles. We’re done 3D printing our custom bearings and cutting down churros, so our final manufacturing step is cutting carbon fiber bars and CNCing side plates for the subsystem.

Chassis

• This year we’re using SDS MK4i swerve modules with Krakens for the drive motors and Falcons for the steering motors. This week we swapped out the driving Falcons on our practice programming drive base with Krakens as well and are using those replaced motors as steering for our actual bot.
• We have cut and measured the wooden frame for the practice bumper set and have started making the frame attachment brackets. We will begin making the outer brackets and assembling the practice bumpers soon and we have decided on a font for our team numbers on the bumpers.

[philosophermurr]

Programming progress

The first thing we did during the season was generate swerve code with Phoenix Tuner X. We used our test chassis, which we then had to further tune. We ran SysId tests to figure out the kS value for our steer and drive motors, and then we used Tuner X’s tuning functionality to determine the kP needed for our robot to drive smoothly.

After we generated swerve code, we started working on pathing. We planned out our autonomous paths, which we then tested on our test chassis. As of right now, we aren’t using vision, but it’s in our code, and we’re planning on adding cameras to our robot as soon as we have AprilTags we can use. For our pathing, we’re using PathPlannerLib for functionality like event markers, but we’re generating our trajectories using Choreo.

We’ve also been working a lot with simulation, specifically for our shooter’s aim. Since it works similarly to an arm, we’re using WPILib’s arm simulator, which has helped us figure out how quickly the shooter will be able to change angles. We started out without an external encoder, but we’ve decided to add a CANCoder, which we are using as a FusedCANCoder, to allow for more precision in our aim. We’re hoping to be able to use this one shooter to score in every field element, so we want to be able to control its angle extremely precisely so we can hit small targets like the trap.

Overall, programming has been making sure all our code is almost ready by the time we get a robot. All of our code still needs to be tested, but we’ve done the majority of the writing. We want to spend as much time as possible tuning, rather than having to write code later on.

We’ll also start having weekly programming updates to the build blog, so stay tuned!

[madelinep]

Here’s what we worked on this week:
Intake-

• This week we adjusted our gear spacing on the side panels and finished assembly of the intake. For now we’re using polycarb rollers with rubber covers, but those may switch to carbon fiber since they would be easier to swap out and stronger rollers. We also attached the intake to the front frame perimeter and will be testing it and making any adjustments in the next week.

Conveyor-

• We added a powered roller above the note path between the shooter and the intake to spin in conjunction with the shooter carbon fiber rollers, which will be connected with polybelt to the lowest shooter roller. There are also 1/16” HDPE top and bottom guides to help the note path, as well as a bottom idle roller with central rubber wheels and smaller TPU wheels on the outer edges to better grip the note within the conveyor.
• We also began manufacturing parts for the conveyor, including our guiding plates and the axle for the lower idle roller.

Shooter-

• Today we CNCed new shooter plates. However, we found a host of issues, including missing holes and incorrect spacing which we’ve fixed at our most recent meeting while reassembling the completed parts of the shooter. We also found the optimal spacing for the floors in between the plates to support the note. We need to cut carbon fiber tubes and attach the pulleys for those tubes for the bottom of the shooter.
• The shooter pivot Falcon plus MAXPlanetary is planned to be underneath the bottom HDPE plate of the conveyor (in CAD above). We’re planning to print the top 60T shooter pulleys for the system this weekend, and both the top and bottom tilt axles have been cut and tapped.
• Also a correction to last Saturday’s post, our shooter is inspired by 6328 Mechanical Advantage’s design.

Chassis-

• Our belly plate has been CNCed using ⅛” aluminum instead of our previous plan of ¼” polycarb, since we wanted to give the bottom of our chassis back a little extra clearance. Most of the inner chassis parts have been manufactured including the 2x1s that our electrical panel will lay on and brackets for more structural support, and we plan to finish assembly of the chassis tomorrow.

Electrical-

• We cut out the central electrical panel today, and we’re going to start adding components to it tomorrow so we can rivnut the panel to the central 2x1s in the chassis this week. We also changed the under-bellyplate electrical cover to 1/16” poly to keep it thinner than the main bellyplate.

[VishruthT24]

~2974 Strategy Process~

Kickoff

• At kickoff, we followed an intricate process that would allow for our whole team to get involved with making impactful decisions when it came to our robots strategy for the upcoming 2024 Crescendo season.
  • First, we watched the game video as a team three to four times. We feel that watching the video multiple times allows us to gain a better understanding of what the game is really about, and clear any nuances and/or questions that may arise based on the video.
  • Second, we broke out into groups and delved into the game manual, we poured over assigned sections of most importance (scoring, robot limitations, etc.) and had veteran members lead younger, less experienced members through the game manual. From here each group wrote a summary and presented it to the team as a whole for everyone to gain a good understanding of the game manual. After this, we typically assign members to further read the manual on their own, as it is still beneficial to do.
  • Third, we create our strategy based on the information our team collects. This year, for example, we wrote a program that would calculate the maximum score of your robot depending on previous years’ cycle times and other prerequisites.
  • Then we go onto the engineering portion of the robot, in which we brainstorm ideas for the physical construction of our robot.

To sum it all up, we used the RAC Method:

• Read the rules

  • Use what you read to gain a detailed understanding of the game and what strategies could be beneficial.

• Answer the “What & Why”

  • What is our strategy for qualification matches and playoffs? Why do we want this to be our tactic of choice (JUSTIFY)? Our strategy is what drives our robot requirements.

• Cite the “How”

  • How will the robot perform the functions we need to meet these requirements? Keep concepts simple for now. Ex. ground intake (yes) vs 2-bar roller intake (no).

What our Strategy department has done since kickoff:

• In the weeks after kickoff we have met as a department almost every week. First, we would just bounce ideas off each other until we had a set and developed strategy based on input from the rest of the team. Then we began splitting off into assignments. A group has been using AppSheet by Google to develop a scouting app we can use as early as our week two competition. Another group has been creating a Google Form primarily asking objective questions that we think are important to know from a match. The Strategy Director works closely with our Engineering Department to make sure our robot can adequately complete the tasks we need to accomplish to be able to compete at an elite level on the world stage and with the presentation and outreach departments to see if there is anything we can do strategically to improve our Impact presentation and preexisting Outreach programs during competition season. Also, we plan on presenting an Intro to Crescendo training to our team that will cover all FIRST game updates, and rule changes that team members may have missed or need clarification for before the competition.
• We will post another update soon to explain how we have progressed with the competition nearing as well, thank you!

[madelinep]

This week’s updates:
Our robot goes fully to electrical today, and we’ll hopefully hand over to programming on Monday!

Chassis:

• We finished the chassis with the final addition of riveting all supporting gussets and all inner chassis components to the belly plate.

Shooter:

• We finished manufacturing our shooter this week and will be attaching it (more than just briefly for the photo above when we were testing conveyor) today. Although this shooter version definitely won’t be our final one, we hope it will last through Anderson!
• With the exception of one roller belt, all of the pulleys we need have been added. We added rubber grips to our two bottom carbon fiber rollers for more grip on the note, and did other work to make our shooter more efficient including swapping our previous pulleys for ones with freer clearance and added a top glide plate for smoother flow through the shooter.
• For the pivot, we added the Falcon plus MAXPlanetary and ended up running into some trouble with our belt setup early in the week. We had to change our 40T belt between the 18T gears from the bottom pivot axle to the Falcon axle to a 36T belt. We had to recut our shooter towers to better fit the 100T belts we had, though even with the new towers cut to exact center to center belt distance from ReCalc, the belt was still a bit loose. So we printed another set of top pulleys with an extra tooth added, which we’re testing today. We also cut 24T pulleys for the bottom axle out of aluminum stock and got everything about the pivot set up for adding the shooter today, since it was easier to add the bottom axle of the shooter pivot and the shooter towers to the chassis all in one sweep along with the internal chassis frame because of the compact spacing of our intake/conveyor/shooter path.

Intake:

• We attached our intake this week, and switched the polycarb rollers to carbon fiber with RoboBitz end adapters to keep the intake lighter and stronger.
• We did some testing with both the intake and conveyor on (below, we’ll post videos soon), and we’re able to intake off center as seen, since once the note touches the conveyor roller it sucks the note fully in.

Conveyor:

• This week we finished most of our conveyor, including top and bottom guide and powered top roller, though we’ll be adding the polybelt to transfer energy from the bottom carbon fiber roller of the shooter to the conveyor when we install the shooter today.
• We switched away from the idle roller plan in favor of just extending our bottom HDPE guide (by adding wood inserts and sanding them down along with the central chassis 2x1s to follow the same angle as the 3D printed supports) because the compressible rubber wheels were slowing down the note as we sent it through the intake. We also added a slit for a piece of flat stock in our supports for the bottom guide since the HDPE was bowing across the gap between the supports.

Electrical:

• We ended up having to reCNC our main electrical panel yesterday because some Loctite from screws within our mini PC holder got onto the plate and weakened the area underneath. We also had made some changes to the design of our main panel that we wanted updated anyway, including just tapping the PDH into the 1/4“ poly instead of having to layer an extra set of nuts under the components above our panel and adding wider bottom supports for our Mini PC, which will definitely simplify maintenance during competitions. But we got everything we could without the panel set up, like power pole connectors going into the PDH for the steer falcons, so we can wire as efficiently as possible today.
• We also got our battery holder printed from TPU (complete with cool racing stripe, totally not just because we ran out of black filament) and mounted it with our Anderson connector attached onto the belly plate. We printed swerve covers to hold our main breaker and probably radio as well.

Bumpers:

• We also got quite far on our bumpers, which are done by our JV Build members! Our team numbers were ironed onto our set of practice bumpers early this week. In order to make our bumpers fit better around protruding nuts on the robot’s frame, we added divots to them. We also finished cutting the 2x2 brackets to their correct length, and we got through around half of our wood pieces for countersinking.
• As of now we have finished our first set of bumpers and just need to place and attach them to the robot.

Design-
Now that we are almost complete with the robot from the intake to the shooter, we are now looking at potential designs to climb and make the shooter work more effectively with the amp as the rest of our bot gets handed over to programming.

Amp:

• We’re considering a servo attachment that can slide out above the shooter or rotate above the shooter and act as a hooded shooter. The design will be flexible HDPE and curve down as we run into the amp.

Climber:

• Our most prominent design idea right now is a winch system in which a hook will attach to the top of the shooter or newly implemented amp mechanism and swing up to attach to the chain when deployed. This hook would be detachable from the shooter itself. The idea behind this was to make sure we can bring the bottom of the robot as close to the chain as possible. We are working on CAD for this design over the next week.

[madelinep]

Here’s video from testing our conveyor/intake setup on Saturday. We also realized that our shooter needed an extra support churro in the bottom corner to keep the edges of the shooter from bowing in, but that churro interferes with the polybelt in between the top conveyor roller and the stationary bottom roller on the shooter so we’re adding a new NEO above our intake/conveyor top plates to run the conveyor roller.

Intake/conveyor testing 2/17/24

Also the robot is mechanically done (ish)! We’re wired and everything is mounted, and at this point we’re onto adding sensors and adjusting and testing our shooter further. More updates to come this weekend!

[philosophermurr]

Programming progress

We took a break last week, which is why we didn’t have an update for you all. Sorry!

However, we’ve made quite a lot of progress this week. We started out by simulating the flywheels for our shooter using WPILib’s flywheel simulator, which we’ve been using to determine how long it will take for our wheels to spin up, allowing for more accurate auton timing.

We’ve also added support for aiming during auton in simulation. We accomplished this with our aimAtSpeaker() and toTarget() methods. We’re running aimAtSpeaker() every time we want to shoot, and then toTarget() throughout our auton so we’re always aimed at our target. For our aim, we’ve also slightly changed the way that we’re calculating our aim target—the target changes based on where we are on the field so that we always have the biggest target possible. Essentially, our robot is always going to aim at the far end of the speaker from where our robot is, unless it’s close to the middle of the speaker, in which case it will aim at the center of the speaker instead.

We’ve started trying to implement shoot-on-the-move, referencing Mechanical Advantage and Team Rembrandt’s methodology.

After getting the robot, we started testing our teleop functionality. The robot can intake and shoot notes, but the aim has not been properly tested as the build team is planning to add a CANCoder to it. We could still manually adjust the shooter angle, so we tried scoring in the speaker and amp.

We could score in the speaker with little difficulty, but the amp required some tuning. We had to check whether the motors had spun up (which we use our spinUpFinished() method for) before shooting which allowed us to be more accurate and consistent with our amp scoring.

[madelinep]

We’re participating in PCH’s robot reveal night tonight, so be on the lookout for info on our bot Shosty (as in Shostakovich) from them!

Since we’ve handed the robot off to programming we’ve had just a couple changes to mechanisms.

Shooter-

• We are in the process of adding hard stops for the shooter above our battery to make sure the angle of the shooter can’t go below 22.5°. We’re also planning a programmed stop point.
• We also have manufactured a belt tensioner to add to our pivot belts since even with our new printed gears to keep the belts tighter, we were still skipping some teeth.

Conveyor-

• We’re now powering the top conveyor roller through a polybelt connection to a Neo mounted above our intake. We also changed the supports for the bottom HDPE plate of the conveyor to a higher curved angle once we got our shooter mounted and saw how it sat at the end of the intake/conveyor note path.

Climber-

• Our current design includes a winch system to pull the robot up along with hooks extending from a pivoting attachment on the shooter. Arms will reach out to open the trap for the stabilized shooter to shoot into the trap. We hope to have this winch setup ready by Anderson.

Electrical-

• We added a chromatic sensor into the top guide plate of the conveyor to register notes being intaken, and added a beam break to the shooter to see how far through the path the note has gone and ensure it’s in the right position before shooting. We plan to add LEDs and/or an on-screen indicator on the driver station for feedback from these sensors.

[VishruthT24]

Strategy Update:
-We went through Scouting Training with our whole team this week. In past years we used a regular Google form, however this year we decided to use a different approach. We have developed a competition management app from which managing our whole team seamlessly becomes a reality.
- With the app, team members can see assignments for when they need to be in pit, when they need to scout, when lunch/dinner and other meetings occur, and when to go over to our CanBot. In addition to this, the major feature of the WRT Scouting App is the integrated scouting form, which allows members to scout from within the app without having to use multiple applications at once. Since some competition venues do not have the best internet connection, we are able to also sync data entered into the app without any internet. All data gets saved onto the device you are using and updates afterward when a user has a proper internet connection.

• Along with this, as a Strategy Department we have been working closely with drive team to develop the best ways to score and move across the field while being faced with defense.

• At the same time we plan to assign preliminary scouting assignments so that our scouters get more practice with scouting through other week 0 and week 1 competitions. This will also give us more data on what to expect as competition going into our week one competition.

• The open-source template version of our management/scouting app is linked below. If you have any questions feel free to contact us on here at competition, on Instagram, or by email. Finally, by feeding in your team’s specific information and enabling all bots when data is entered, helpful SMS notifications can be enabled.

• Please make a copy of the application from the portfolio link to make edits:
  https://www.appsheet.com/portfolio/446067747

[philosophermurr]

Programming progress

We can now control the aim, and we added a preset to shoot from the subwoofer as one of the buttons for the drivers. In addition, we had to make changes to the way we intake so that the note wouldn’t shoot before the wheels had spun up. Specifically, we had the note come up to a beam break sensor, and then slide down until it was no longer breaking that beam.

We then implemented a state machine for our robot to ensure that the robot is ready for each state of scoring. It uses a VisiSight and a beam break sensor to determine where the note is in the robot so that it can turn off certain motors once they have served their purpose. It uses a collection of triggers, which turn motors off and on based on what sensors have been tripped.

Finally, we started testing basic auton paths, like shooting and leaving.

[madelinep]

Build updates:

We’ve had a busy week driving, finding out what doesn’t work about our note path (ie. breaking things), and making changes to the bot. But first, here’s the link to our CAD!

We ran into issues early on this week within parts of the note path as we got more practice with the full bot in, all of which we were able to fix successfully and fairly quickly. When we were doing cycles, notes were getting jammed occasionally in both the conveyor and shooter.

Conveyor-

• The polybelt that transferred power to the roller was getting too hot because there was too much torque and the crowned pulley on the Neo started to freespin, so we had to switch to a timing belt and shorten the conveyor roller to add a pulley to it. We also made our gearbox thicker to support more of the axle.

Shooter

• The shooter pivot was stuttering some when we started driver practice at the start of the week. We added more tension on the 100T belts to keep the system from succumbing to gravity and skipping, which has worked well.

• Our motor to bottom axle pivot 36T belt broke last night- not from a particular tension overload, the belt was just lower quality and the teeth had apparently been stripping off the belt throughout its use. We considered swapping the pulley setup for gears or chain and used ReCalc and WCP calculators to settle on doubled-up 31 link # 25 chains with two 18T sprockets to find the closest fit given the parts we had, without having to remill one of our inner chassis 2x1s (which is what we would have needed with gears because of our current center distance).
• We had to take the intake and front swerve modules off to fix this, so we also added access holes to the frame to allow the lower pivot axle to be taken out by just taking off a swerve module, instead of the entire front half of the bot, which we were planning to do in the future anyway.

• We swapped our guide plates to 1/16” polycarb instead of HDPE to be able to see note alignment within the shooter. This will let our drivers know if they have to adjust their distance from the Speaker before shooting, since non-centered notes in drive practice have had to be scored from nearly underneath the Speaker.

Intake-

• Our intake has been working well, we just made a couple minor adjustments this week, adding a printed gear protector and putting slick tape on the outer edges of the rollers

Winch climber

• We got our parts CNCed for assembly today. We’re switching from servos to fiberglass rods because we did not want to add more weight to the shooter.

• When adding the winch to the left side in CAD, we noticed it would not fit as it was because the Mini PC extends out over the central chassis 2x1s we were basing dimensions off of. To accommodate for this, we removed a piece from the side plate that it would interfere with and moved the motor, which meant we had to change the gear ratio to 30:30

• Also we’re making the side plates of the winch connected by a 2x1 piece of stock to make it easier to input the winch into the robot as a module and attach the winch more to the belly plate of the chassis.

Bumpers

• We fixed the wrinkles that were found in the fabric and made new custom brackets, since our original ones were hitting the plate to hold on the intake. We also added spacers to increase the bumpers’ height to leave more clearance for the note to pass underneath.

[zache]

Here’s our robot reveal video!

Walton Robotics 2974 Robot Release Video 2024

[philosophermurr]

Programming progress

This week, we’ve been working on autons. We started by adding another note to our shoot and leave path, and then we added a third note. We have a relatively consistent 3 piece auton now.

In addition, we’ve been polishing all of our teleop functionality. First, we added a second beam break sensor at the tip of our shooter to more accurately determine when we’ve shot a note to make sure the robot doesn’t go back to its idle state too soon. We also added a podium shot for our driver as well as amp scoring functionality.

For scoring in the amp, we took inspiration from ORBIT’s amp scoring method, where they flip their shooter down to force the note into the amp. We’ve found that this is extremely consistent.

Our auton paths and teleop functionality are pretty much finalized for Anderson, but we’re going to go through everything and make sure it works as intended. If possible, we want to slightly optimize our 3 piece to make sure we don’t run out of time.

[madelinep]

First off, we won Anderson yesterday with Team 343 Metal in Motion and Team 3091 100 Scholars! Our Dean’s List nominee Grace also won and will be competing at District Champs! We’re super proud of our drive team, pit, scouters, pit and Impact presenters, and more for making the competition a success.

Next, some of our build takeaways. We had no major mechanical issues at comp, which is partly due to the note cover we added at the start of the week. The polycarb sheets we added both make it easier to shake off notes that fall/bounce back onto our chassis, and given the wheel marks left on the front cover, protect the top of our intake/conveyor well. (Photo from TBA)

Also- our intake. Our ability to touch a note and run with it is one of the main reasons for our quick cycle time (which our Strat lead will elaborate more on in a post soon). A lot of the questions we got from people coming up to our pit were about the intake, which can be seen in the CAD link from my post last week. Our drivers won quite a few on-field bouts of competition for notes, since the wide span of the intake means we don’t need to slow down at all for alignment and can just grab and go. We saw the effects of a few front collisions especially during playoffs, with our front 1x2 above the intake and the front right 1x1 bowing in up to about a quarter inch at one point from a hit from another bot’s corner. The front of our robot is fairly robust- the stock we have there is 1/8“ thick, so we plan to just replace it with identical pieces and print more copies of the note guides we needed spares for.

We had been on the lookout for the potential warping or cracking of the 1/4” polycarb side plates, but they seemed to hold up well. Those pieces and the bumpers did their job of protecting the intake rollers, but bumper brackets took some damage as well.

Electrically the robot was working smoothly all comp, the only bigger issue we had was with our ethernet switch. It just stopped working between our ops check and one of our playoff matches, which was fine since we don’t have cameras on right now which is what the additional switch ports are needed for, and we were able to route our connection from Rio to radio with just the RPM. We’ll probably keep the switch off until we have cameras on, and replace our current TP-Link switch with a new one.

Between now and Carrollton Week 4 we have a few improvements to make. We’ve already built what we could of the climber off the robot and will be installing it early this week, but we had left it off for Anderson since we wanted programming to be able to focus on autos instead of climbing for the time being. We’re also going to be cutting down on the length of our shooter, both to keep our robot light and to shorten intake to shoot-ready time. Cameras and LEDs are also planned additions for either before Carrollton or before District Champs. Updates on those and more next week.

[VishruthT24]

This past weekend our team participated in the Peachtree District Anderson Qualifier Event in Anderson, SC

• Some takeaways from a strategy perspective are listed below:
  • Pit - Scouting; is a huge deal in this game, knowing the autonomous paths of your partners is a make or break when it comes to developing successful match winning strategies for Crescendos
    • A lot of times matches were won in Autonomous, as teams would have to chase to catch up to the score when three really good teams got to pair together.
  • Regular Scouting; It depends per team, and what your team is actually looking for when it comes to scouting. However, making plan prior to competition is important, you know what you are looking for and your scouts’ data will actually come to fruition and lead to strategical input.
    • We were able to use our scouting data collected for essentially every single robot at Anderson for every match they played to see what teams would be able to statistically best complement our robot.
  • Overall we plan on going deeper with scouting before our second competition. More prelim scouting with younger members of the team pitching in, and a better data analysis approach using tools like Autocrat, and AppSheet.
  • We have updated our app, and plan on pushing out our second version prior to week four as well. With any questions about the app just respond here or message us on one of our social medias and we should be able to help.

[grac]

17.3.24 Programming Update

Since our last post, we’ve been busy optimizing our robot and creating new autons! After our Anderson event, we finally got our new wheel radius characterization method working. It’s inspired by 6328 Mechanical Advantage’s wheel radius characterization technique, and it is so much easier than the pushing-and-measuring method we used before!

Our autons have also improved significantly :D. After Anderson, we noticed that our robot did not follow our Choreo paths perfectly because Choreo doesn’t account for battery voltage, so the robot was sometimes slipping when it tried to keep up with the path. We fixed this by constraining the torque more. We also added several new autons! During Anderson, we upgraded our 3-piece auton to a 3.5-piece auton. This week, we also added a 4-piece auton and a 5-piece auton! In general, we also worked on optimizing our auton paths in Choreo a little more to allow for faster autons.

Currently, we are still working on tuning our arm for faster and more accurate aiming. Additionally, to make robot set-up during competitions more accurate, we also are adding a limit-switch based system on the robot to make the arm start in the same place every match.