4795 Eastbots - 2023 Build Thread

Post 1: Offseason recap and intro to our team!

Team intro

We’re team 4795, the Eastbots, a FIRST North Carolina team based out of East Chapel Hill High School. We’re a student-led team with a mission to lower skill floors and make large scale impacts, both in our school and local community, all this to inspire the next generation of leaders in STEAM.

This year, our team is about 80 strong, with 12 leadership members and 8 mentors. We’ve worked for the past several years to innovate new training and engagement programs to support sustainable knowledge transfer, and provide a transformative team experience to each team-member, all while building awesome competition robots through student-led innovation. We’re psyched to be a part of openalliance for the 2023 season.

|676.3536x450.7893

What we’ve been up to - Offseason recap

Summer

This summer our team focused outreach and improvement of our in-house capabilities. Below is a brief summary of these activities, written by involved members.

Outreach

During the summer we did regular outreach events at local children’s museums, summer camps, and the local public library. Generally this consisted of taking our robot to the event, demonstrating its ball shooting, and talking to parents about our program/ programs appropriate for their children. We also attended outreach and robot demonstration events for a few teams close to us, namely 587: The Hedgehogs, and 4829: Titanium Tigers.

Tooling upgrades

We received the Gene Haas Foundation grant this year and decided to use it to buy new tooling. We decided to purchase a Shapeoko 4 XL. This machine isn’t ideal for cutting metal, but we were happy to make that compromise as we have a sponsor that can cut metal parts for us. The ~$2000 cost savings over an Omio was worth it for us.

Fundraising

Following the increasing formalization of the rest of our team, our business team has increased in number of roles and size. Because of this, along with our recent work to promote equitable technical training, we have received over $8,000 in new grants: a very large amount for us.

Competitions

!!!

We attended 2 offseason competitions in FIRST North Carolina this Fall, specifically the THOR 3 and Doyenne inspiration events. We were thrilled to win both, along with our great friends at team 4829: Titanium Tigers.

Fall

Annually, our fall season consists of efforts to recruit, train, and engage new team members. We have innovated a series of programs through the past 3 years to better accomplish this. Additionally, our fall offseason typically provides returning members with creative outlets that benefit the team and its capabilities. Our fall activities are described below by participating members.

New Member Programs

Recruitment
We started the year out by equitably recruiting students at our school. We do this by advertising our interest meeting in freshman english classes so every single freshman at our school hears a pitch to join the team. We definitely went a bit overkill on recruitment this season with over 70 new members registering for the team, and we’re still struggling with the amount of new members on the team now.

Skills Training
Our skills training program seeks to provide recruited students with the foundational knowledge needed to contribute in their respective sub teams, through 8 hours of classroom instruction over 4 weeks. We typically encourage members to train in multiple subteam disciplines, but this wasn’t possible this year due to our incoming class size.
Our training programs in general have taken considerable iteration, and we’re quite proud of the knowledge we’ve built in how to train for FRC over the past few years. Info and curricula can be found in this super document.

Design
Our design skills training is the 3rd iteration of a CAD basics program using onshape, and this year focused on slowly building CAD confidence in members. The slow and steady approach to CAD learning, along with persistent returning member teaching assistants to help our design lead, led to our highest success rate, with over 70% of our 25 design members CAD competent by the end of skills training.

Mechanical
The goal of our mechanical training is to have members who have never touched a drill before be able to meaningfully contribute to building a robot. They spend one month building a mechanical training kit in small groups and getting safety trained on our shop tools.

|501x375.9792

Electrical
The goal of our electrical training is to have first year members be able to wire most of a robot and help design basic custom circuits for sensors. We start out by teaching basic electrical concepts, then move on to learning about FRC specific components, and finally finish up by doing various wiring exercises.

Programming
Our programming training is 8 weeks instead of 4 weeks because we’ve found that it’s difficult to train a member that doesn’t have programming experience in 4 weeks. We start out by spending 6 weeks learning Java and then move on to working on Romis. This course bleeds into the first few weeks of our next activity, ERC.

|501x372.3107

East Robotics Competition (ERC)
While skills training provides foundational knowledge to new team members, ERC provides opportunities for each new member to practically apply that knowledge. Founded in 2020, this activity splits our new team members into multiple (3 this year) ERC teams, each building a small scale, FRC analogous robot over a 6 week build season. This year, we were able to, for the first time ever, open ERC for participation by other teams in NC.

Our competition was held this past weekend (December 11th). More info about this will be posted soon by our ERC coordinator.

|328x214.9999999999997

Returning Member Projects

As stated before, our returning members have championed a variety of projects in order to improve theirs and the team’s capabilities and knowledge.

Elevator
Like many other teams, we’re building an elevator in preparation for the upcoming season, specifically its potential to be a pick and place game (this group of team leadership only has experience with shooting games). Our design is a 3 stage continuous elevator that mounts on a WCD drivebase given to us by 2363: Triple Helix. CAD of our elevator design can be found here.

Pneumatics Test Board
We haven’t historically had a way to test pneumatics on our robot without an entire electrical system, so we decided to build one that runs off an Arduino. This has two double solenoids but is pretty easily expandable if we want more in the future.

Build Season Prep
Our design team has been hard at work with a myriad of drivebase designs in preparation for the upcoming season. Specifically, we’ve been working to design a standard west coast drive (for our second robot), a swerve drivebase using 3” MaxSwerve (which we just recently purchased), and a specialty drivebase (as a starting point, should we encounter a terrain-focused game). More information about these designs will be posted soon by our design lead.

Conclusion

Thanks for taking the time to read this! Please ask any questions here in the thread, which we hope to update quite actively in the coming months. We’d be thrilled to share any materials or references, don’t hesitate to ask.

Thanks, and happy open alliance!

17 Likes

Always nice to have more NC teams in Open alliance :blush:

4 Likes

Post 2 - CAD/Code Release

We are proud to release the code and CAD for our 2022 robot, Giannis!

Code
CAD
Engineering Notebook

Key Features

  • 6wd WCD
  • 4 ball autonomous
  • VIW cutout intake
  • Automatic wrong color ball rejection
  • Shoots from anywhere between 1’ and 20’ away from fender
  • Topspin control biwheel shooter
  • 6 second ground to high bar climb
11 Likes

Quick Update

We’ll be trying to adopt a somewhat uniform post schedule come build season. In the past we’ve updated our build log daily, but with open alliance we figure a weekly consolidated post will work better. If anybody would like to see us do updates a different way, please let me know! This thread is meant to serve whoever reads it, so we’ll likely go with public opinion here.

A few other members will be out pretty soon with some pre-kickoff project-related posts, so look out for those. After that we’ll get into build logging!

Offseason Post - ERC Recap

What is ERC?

The East Robotics Competition (ERC) is a 3-year running training-focused robotics competition, designed to engage new and inexperienced team members in the technical and social elements needed for FRC.

At the start of the ERC season, our new members are broken up into ERC teams of 10-15 students each, and every ERC team is given a custom ERC game manual, which details the game and field for this year. Over the course of a mock build season of 6 weeks, these ERC teams independently make small-scale robots, with new members being the only ones to do any actual work on the robots and returning members overseeing the work and offering advice. And then, the teams compete against each other at the competition, which includes scouting, matches, awards, and food.

We’ve found that our new members come into the FRC season prepared and more able to collaborate and participate due to their experience from ERC. Our returning members are more able to take on leadership roles during the season because of their experience coaching the ERC teams.

This Year’s Season

Game

The ERC game this year was called Riptide Rescue. The kickoff slideshow (short and sweet) can be found here, and the manual here.
The winners of the 2022 East Robotics Competition were Eastbots Team 79 and Team 3822 (Neon Krakens)!

Our Team

With a massive influx of new members, we split our team into 3 ERC teams; 47, 79, and 95. Each team met twice a week, with each meeting having two teams in attendance, which allowed for intra-team communication and the encouragement of gracious and professional attitudes.

Other Teams

For the first time, we were able to expand ERC to other FRC teams! Two NC teams participated alongside our ERC teams; the Hotbotz (2640), and Neon Krakens (3822). Both teams entered one ERC team and participated in the competition with epic robots.

The expansion of ERC from an internal competition to one that other NC teams could benefit from and enjoy allowed for a greater level of collaboration and teamwork, as well as the opportunity to learn from and compete with other talented teams. It was a fantastic opportunity, and we’re very appreciative of these teams.

The Bots

Robot Overviews!

Team 47

The Eastbots 47 bot included a rotary arm and wheeled intake. Its rotary arm was not wholly functional at the competition, but it could still fire its preloaded box into the mid goal, interact with the low goals, and play defense.

Team 79

The Eastbots 79 bot used a winch-driven rotary arm and wheeled intake, capable of intaking boxes from the ground and shooting them into the mid goal.

Team 95

The Eastbots 95 bot had a rotary arm and a clawed intake mechanism, and could intake boxes from the ground and place them into the mid goal.

|164x172.051912188434

Team 2640

Hotbotz’s bot was capable of pushing into the low goal, playing defense, and stealing boxes from the other team’s low goal. It also featured a work-in-progress telescoping rotary arm!

Team 3822

Neon Kraken’s bot featured an elevator and ground intake, capable of picking up boxes and placing them in the mid goal or on the high goal.

Challenges

Part of what we’re teaching at ERC is how to work well together, and at the start, some of our teams struggled with constructive criticism and good inclusivity skills.

Actively engaging our returning members proved more difficult than anticipated, and led to some of our returning members feeling like they didn’t know what to do and weren’t having their existing knowledge actively built on in the offseason. While some enjoyed coaching and mentoring ERC teams, others struggled to participate outside of ERC or the additional offseason projects we were working on.

The people within our ERC teams have at this point learned to work well with each other, but not yet with members of the other ERC teams. We’re going to work hard to make sure that there are opportunities and encouragement for our members to forge bonds outside of their ERC team groups during our FRC season.

Conclusions

ERC is an activity we’re very proud of. We feel it offers a very strong engagement opportunity for our new members to learn and exercise their skills, and for our returning members to learn leadership and mentorship skills. We’re excited to continue improving this activity and growing FIRST NC participation.

Any and all questions are welcome, thanks for reading!

This post was written by Reese, our ERC coordinator.

7 Likes

Hi, I’m Reese, the ERC Coordinator! Please ask me any questions about ERC, whether they be about the idea, planning, or event itself.

If your team is interested in participating in ERC next year, please reach out to us at [email protected].

5 Likes

Hello! I’m Wesley, the design lead of the team this year. During the past few weeks we have been working on drivebase designs for the upcoming season. The three types of drivebases we are working on are WCD drivebase, swerve drivebase and terrain drivebase

West coast drive

  • We had a consistent performance with our west coast drive drivebase last season since our second district event, and this season we hope to make some improvements in serviceability.
  • WCP ss flip gearbox, 6 wheel drive, four 6” traction wheel+two 6” rev omniwheel
  • Use short tubes between wheels for backing the bumper.
  • We switched from versaframe+gussets to maxtube+end caps for better serviceability. This also makes the side rails available for mounting belly pan and mechanisms.
  • 1*1 3/16” hole pattern on bellypan for electronics mounting.
  • Thumb nuts for bumper mounting. Last season we used wing nuts for this. It was hard to get it on and sometimes they got cross-threaded because the metal was too soft.
  • Link to the CAD

Swerve drivebase

  • The new rev swerve made swerve a lot more affordable, we decided to do a swerve bot if the terrain is appropriate.
  • Use the gussets that comes with the swerve module for mounting spark max.
  • Have both horizontal and vertical battery box design. The horizontal battery box sinks .75” below the belly pan for lower center of gravity, this also allows the game piece path to go above the battery if needed.
  • Same bumper mounting strategy as our WCD drivebase but with andy mark gussets instead.
  • Here is the link to the CAD

Terrain drivebase

  • We decided to do a 8wd 6” pneumatic wheel drivebase, with wheel shafts drop under the rail for maximum ground clearance. The belly pan has 6” of ground clearance.
  • #35 chain, with WCP ss flip gearbox.
  • Sheet metal plates outside the tubes for mounting wheel bearings, chains are exposed so that when they touch the ground they can help move the robot.
  • Pockets on the aluminum plates on the side rail for weight reduction and aesthetics.
  • Custom bearing block for chain tension, use wcp cam tensioner to adjust.
  • 1x1 rails outside the wheels for bumper mounting.
  • CAD link
6 Likes

Curious on how you are doing the threads for the thumb nut? Just worried if you destroy the stud that you could be in trouble and your backup would be replace the Maxtube (or flip it over).
Is the T Gusset to help you get to the bumper backing rules?

So we’ve used studs with wing nuts for about 3 years now and have never damaged threads on a stud. If we did, the studs are just 8-32 bolts held in with a locking nut inside of the frame rail, so they can be dropped and replaced fairly easily.

The main reason we’re switching to thumb nuts here is for ease of access. The wing nuts last year on the front and rear would touch the bumper mounting brackets on each rotation and were a real annoyance, so we switched to standard non-locking nuts. We’d often have to break them loose with a socket if they were over tightened, so thumb nuts seemed like a good move.

We thought about toggle clamps or sliding latches, but this method seemed way simpler and cheaper.

Yep, although it’s only really needed with a 30” long drivebase. This is actually a little flawed though: inspectors cleared it at our first event iirc, but we were told to add box tubing to the sides of the gusset at our next event.

We (6500) used thumb nuts last year and we’ll do it again this year. So much easier than wingnuts

5 Likes

Offseason project - Elevator

Like many other teams, we’re building an elevator in preparation for the upcoming season, specifically its potential to be a pick and place game (this group of team leadership only has experience with shooting games). Our design is a 3 stage continuous elevator that mounts on a WCD drivebase given to us by 2363: Triple Helix. CAD of our elevator design can be found here.

|281.35714285714283x375.7171712472425

  • plain stocks+gussets for structure and used our own bearing block design.
  • Used our own gearbox and it can be clamped onto the bottom stage tube.
  • 3DP Helical spools for rope.
  • A frame to reinforce the structure.
  • Eccentric spacers for tensioning bearings, plastic sliders between tubes instead of bearings.

We were not able to finish the elevator’s carriage stage and put it on the drivebase. But we gained a lot of information that will be helpful in later elevator designs.

  • Tube tolerancing was bad and the tubes are not parallel. It binds when it’s retracted and it becomes wobbly when it’s extended. A fix to that could be using tubes with predrilled hole pattern like maxtube.
  • The pulleys are not in the middle of the tube, causing one side of the stage to rise higher up than the other. This can be fixed by minimizing the pulley size and putting the pulleys as close to the center as possible. Or we can have two sets of pulleys to balance out the force.
  • The gearbox took up some space at the bottom, so we can not utilize all the travel of the tubes.
  • Eccentric spacer might not be necessary.
  • Plastic sliders can produce a lot of friction when it is pressed against the tube, bearings might be significantly better.
  • Might need to switch to endcaps for structure.
  • Might need to use cots bearing blocks for actual robots.
6 Likes

Configurable bumper

We found that no matter what drivebase we are designing, the bumper design is always similar, especially after we have come to a solid bumper plan. and we want to spend as little time as possible on the bumper design during season. So we worked on a configurable bumper. You can import it into any robot or drivebase design and configure its size and bumper cutout size.

Here is the cad link.

3 Likes

Post - Pneumatics Test Board & assorted CAD releases

Pneumatic Test Board

Goals:

  • Cost effective (costs ~$50 to the team)
  • Controls two double solenoids
  • Easy and intuitive to use
  • Retains all safety features of roborio+PCM

BOM

CAD

Code

Video Demo

Physical Component Layout

The physical component layout and plumbing is identical to a normal robot, retaining all of the required safety features.

Electronics Box

The goal of being as cost effective as possible meant I had to find an alternative to the FRC control system. I decided on using an arduino along with relays to control solenoids, and a brushed motor controller to control the compressor. We wanted to use solenoids which we already had which run on 24V so I added a boost converter to step up the 12v from the battery. Additionally, to make it more user friendly and lightweight we used FTC batteries we had laying around.

|318x412.0625690472007

Simulated TinkerCAD circuit

Recommended Improvements

  • Add a way to easily remove the battery from the test board
  • Add right angle fittings to solenoid outputs to make it easier to plug things in
  • Make the electronics box bigger, I ran out of space to put electronics in it
  • Use heatset inserts instead of pocket nuts to secure the two printed parts together
  • Add handles to carry the board

Compact Pneumatics Setup

In addition to the test board, I wanted to find a way to use lightweight components from Andymark more efficiently.

NOTE: This has not been tested on a robot yet

CAD

Goals:

  • Remove as many failure points as possible
  • Take up as little space as is practical
  • Be as lightweight as is practical
    The way I accomplished this is by putting the entire high pressure side before the air tank and low pressure side mounted to the other side of the tank.

High Pressure Side

The emergency relief valve, high pressure gauge, and pressure switch/analog sensor are all mounted on the output of the compressor. From there, we have a short tubing run into the tank using right angle fittings to prevent tight bend radiuses.


Low Pressure Side

We have the regulator, working pressure gauge, and dump valve mounted on top of the tank. There is a strain relief bracket with ziptie slots on the tank because we have seen a lot of leaks come from the fitting coming out of the tank.

SB50 Clip

We saw a major failure in a playoffs match at DCMP where our battery cable was unplugged when a robot went over our bumpers and contacted it. This was partially due to the PDH end of the connector being hard mounted and that we weren’t putting a ziptie between the two connectors before every match. In order to prevent this from happening, we designed a clip that holds the battery connectors in place. We might end up mounting battery connectors to our robot using a TPU printed part to allow for compliance.

NOTE: this has not been tested on a robot yet

CAD

Goals:

  • Easy and intuitive to use
  • Can be installed in in <5 seconds
  • Prevents battery from being unplugged

The clip we ended up with uses two M3 screws that go into the SB50 mounting holes to hold the clips together, and two separate printed parts that attach to each other using these magnets.

3 Likes

Happy kickoff everybody! Kickoff recap post coming soon.

Build Season Update 1 - Kickoff!! (And early strategy)

Saturday - Kickoff overview:

What a game! We had a blast watching the stream, eating pizza, and doing strategic analysis!

Immediately after the game reveal (and lunch, of course) we broke into 4 groups for manual reading and task analysis. Here tasks were identified and analyzed based on difficulty and value (using this sheet), and archetype brainstorming began. For many (50ish) of our members this was their first FRC kickoff, but their experience with ERC (our “East Robotics Competition”) seemed to help a lot for understanding of task analysis and strategic design. We purposefully avoided mechanism brainstorming at this stage, to avoid associations of archetypes with designs.

After this, we returned to the main group and each group presented some ideas and noteworthy things.

Here is an abbreviated list of the items we discovered preliminarily.

  • High goal is quite a reach forward and up.
  • The platform is narrow! Only 32” for each robot on average (without wiggle room).
  • Climbing seems very high value, especially in autonomous.
  • Links are super high value and may make low/mid quite viable.
  • Full field sprints seem common.
  • Shuttling! Momentary control rules seem to dictate that a robot with a sub 3 second field sprint time can ferry many gamepeices at once from feeder to community.
  • Additionally, multiple non-momentary gamepeice control is… allowed in community??

The weekend - Strategy, analysis, and conclusions:

We met online several times over the weekend to analyze the game tasks further and evaluate first several tasks, and then the merits of the archetypes we saw.

Before we get into that, though, here is a bit of info on our competitive and technical goals, set prior to the season:

  • Finalist at our first event (week 1: Asheville)
  • Peak relative to the North Carolina meta at our 2nd event (week 3: Wake county) and win.
  • DCMP finalists.
  • All as a simple but effective (the Easbots way!) well practiced cycler with advanced software controls and autonomous.

Okay, back to the analysis.

The first task we analyzed was the platform task, which we found through some match sims to be critical to match scores. Outcomes here:

  • Our robot should be able to climb the platform with impunity.
  • Our robot should be able to ease the alignment of alliance partners for a triple climb.

Next, we examined the merits of a feeder only robot vs. a ground and feeder robot. This was done with rough match simulations with cycles drawn onto a screenshot of the field.

Takeaways:

  • Ability to intake gamepeices on the ground in auto placement will be very useful in shortening cycles in the beginning of the game.
    • Likely especially so in week 1-3 North Carolina, where many teams will be unlikely to take their ground loaded gamepeices early in matches, if at all.
    • Also enables autonomous performance.
  • Ground intaking becomes more of a circumstantial bonus after first 40 seconds or so, by which time pre-placed ground gamepeices are gone.
  • Conceptualized a few feeder only robots to see if there’s a significant complexity reduction from ground intake.
  • Decision: Our robot should be able to intake gamepeices from ground and feeder.

After this we considered the following archetypes:

  • Cone only all levels (conly)
  • Cube only all levels
  • Cube and cone mid + low (cubone)

We first split into groups to do cycle time analyses, where we could vary pieces scored per game and predict match score using a few google sheets. These also factored in alliance contributions, informing our decision based on what event at which we’d like to peak (working better with better robots aligns with peaking at DCMP, etc.). Here’s our first sheet.

Next, we analyzed match simulations using an awesome spreadsheet, made by a @tooniegoyal. This allowed us to quickly simulate matches and analyze scores including links. These predictions did not factor in climbing or RPs; scorer archetypes were evaluated independently of these.

This analysis seems to show the power of low/mid cubone for consistent scoring specifically with links.

Note that cubone-all-levels robot was considered, but the additional complexity of both iterating a dual capable end effector and all levels transition mech (what we’re calling the mechanism moving an end effector between positions) seemed to conflict with a rapid development process (a key goal for our 2023 season).

Also, it was deemed possible to score high cubes without reaching high by shooting across the 10” or so gap. Upcoming testing will confirm or deny that.

We did some additional analysis to confirm our 3” MaxSwerve drivebase will be able to approach the platform at its starting angle, which, with 4.5” bumpers, it could. We plan some further testing to ensure the 3” wheels don’t bind on features of the ramps or the cable protector, but we’re tentatively swerve go for 2023. @Wesley_Li also made these grid and drivebase mockups, for anyone interested.

What our competition robot will be able to do:

  • Ground and feeder intake of cones and cubes (not of tipped over cones for v1)
  • Low and mid grid scoring of cones and cubes
    • Potentially high shooting of cubes

This informed a few competition robot constraints and objectives:

  • Be small (around 24” square) to improve mobility through congested areas and platform sharing
  • Be fast and agile (likely with swerve)
  • Low center of mass (for low tippiness; tipping moment analysis coming soon for varying drivebase sizes)
  • Be simple (with modularity and potential for through-season iteration)
    • With the ability to be electromechanically complete by the end of week 4 (including with swerve shipping delays).

The robot will possibly look something like this:

We will be building two robots this year, and a priority list for our second “Beastbot” is coming pending further discussion of what will best teach our new members and contribute to iterative competition robot success.

Final thoughts and schedule stuff:

We’ll be assuming a regular schedule of Wednesday and Sunday posts. We already worked on stuff today (yesterday? Geez is it really 3am?) that will go in the upcoming Wednesday update. I and @andreab will be your primary posters.

With that, happy charging!

-Lucien

7 Likes

Week 1 - Wednesday 1/11 Update

We started off running this week, and have found some interesting things.

Mechanical

Mechanical so far has tested 2363’s old 2019 intake (thanks triple helix/@Nate_Laverdure, we love your old stuff!) using our new pneumatics test board (try saying that 3 times fast). We found this actually worked really well, at least for gripping both gamepeices.

https://drive.google.com/file/d/1tiPQYg8aaD94ZHV1uvKNCbFtbEsXG-pP/view?usp=sharing

We’re on the way to a full test with active wheeled intaking.

Mechanical also worked on building some of design team’s prototypes, namely the OTT intake (see later in post) and the compression tester. We got the latter working but it literally exploded on first use (@siyeng can vouch lol).

Design

Our design team conceptualized end effector designs and split off separately to begin design considerations for a universal arm platform we can build immediately. We planned out some end effector designs for fabrication into testable prototype mechanisms, and we also designed a more dedicated compression/wheel testing prototype that we could iterate very quickly.

We conceptualized a few end effectors, namely 3 variants of a pneumatic gripper (wheeled and passive) and an intake that grabs cones from above (Over the Top: OTT) and vectors using mecanum wheels. We also designed an intake using HYPE Blocks and 1x1 dowel wood for more rapid testing of wheel compressions.

We also began designing a 1DOF rotary arm. The aim here is for very fast iteration and development, where we can design and build this by the end of next week and integrate end effectors later. Here are a few concepting and design planning sketches made during meetings (used an arc constraint to inform arm and upright lengths):

And here is the v1 design, using a MaxSpline dead axle configuration. We came up with a more clever way to use MaxSpline bearing blocks to constrain the shaft without needing tube passthrough, and several other review points (design review notes in this doc). Those are listed in a doc here. We hope to have a v2 design to review or verify for production on Saturday.

This mechanism should serve as a modular platform moving forward (enabling our aggressive week 4 completion schedule), potentially for both of our robots. Because this mechanism scores and intakes outside of frame perimeter, rapid stowing and deployment will be critical, and the robot will return to a “safe home” position each cycle. Thus, the arm needs to be super light and very fast moving, ideally a motion cycle (280deg) in <0.5sec. We did some ReCalc math and it seems possible to get it that fast without browning out or turning our battery into a black hole. We shall see.

We played around in CAD with configs of this using the Triple Helix intake, and found a few things: notably that:

  1. Cube shooting with a 1DOF arm seems hella doable, and potentially faster than scoring top cubes by extending.
  2. Intaking from ground is a pretty favorable angle (45deg, for optimal cone drive forward scoring angle), and… potentially allows for picking up of tipped cones from the rear (using some sort of end stop that rotates the cone into the desired angle?

CAD (this will likely become our robot CAD) can be found here. The arm and uprights are configurable in length for maximum playing around with different configs potential.

Design team also came up with this interesting handoff intake concept for a full width dual gamepeice intake. Here’s a video showing how it works. We might prototype this if it jives with our timeline and concept.


https://drive.google.com/file/d/1SRK741lfUujQFS2ik4BCiE9Q80x1q3DA/view?usp=share_link

Software

Additionally, our software team began writing auto-balancing platform code, as well as researching motion profiling schemes for rapid rotary arm control. We also began researching coprocessors for AprilTags.

We were able to get a running test with AprilTags tonight, here’s a video:

https://drive.google.com/file/d/1ZYPO01W-Sxtp9P2tvUqW-pnm_BieskpS/view?usp=sharing

Let’s talk about our drivebases. We’re currently working on the following chassis:

  1. 2363 Drivebase as vision based path following testbed
    • 2363 cone intake is being integrated onto this for testing
  2. Swerve drivebase using Rev MaxSwerve
    • For programming to develop teleop drive code while mechanical works on the…
  3. Dummy drivebase using casters as stand-ins for MaxSwerve
    • Allows for the robot to be built and semi-tested and modules to be integrated later, maybe week 3 (since we’re expecting modules to arrive by Saturday).
  4. Our old robot: Giannis
    • For data logging testing atm
  5. The Beastbot
    • West coast drive chassis based on our pre-season posted design, for our new members, with some returning member leadership roles, to develop a robot on.

Electrical

…is working on wiring all these freaking drivebases.

Wrap-up

We’re also collaborating with 4829 Titanium Tigers on a field build, progress is good! A few of our members go over there each meeting day. Pics will be posted in the Sunday update.

That all said, happy week 1!!

-Lucien

8 Likes

Wow that’s some amazing progress, keep it up! Would you mind sharing the CAD of the ‘handoff intake concept’? It’s a very interesting concept and we would love to take a more detailed look.

2 Likes

Yeah totally, here you go!

And the gripper used in the concept is a lightweight pneumatic-only gripper with a shooter piston.

2 Likes

Week 2 - Sunday 1/15 Update

Howdy y’all, it’s the Eastbots again with another video (err, chief post). Here’s some end of week 1 updates:

Design

We held another design review for our rotary arm. We made a couple key changes after our first review, including:

  • Switching to 3D printed pulleys instead of rev’s (doing printed polycarb for the first time!)
  • Making our encoder gears smaller (and also 3D printed)
  • Adding an A-frame like support from carbon fiber rods
  • Decided to go with 2 in-line neos with normal max planetary gearboxes to turn the arm instead of the 90 degree ones
  • Adding a couple hardstops
  • At today’s design review, we confirmed the viability of these changes and also decided on a couple other minor changes, like adding standoffs between the motor mounting plates. Here’s the design review doc (link).

We designed a prototype of a clamping end effector that operates like scissors. It is similar to the Triple Helix intake we have been testing on, except the geometry is a little different and there’s potential for more wheels.

We also designed a similar end effector which does not clamp, it uses only wheels to retain the game pieces. It uses entraption stars and REV compliant wheels, driven by one NEO motor. We wanted to get this machined as if it was a real mechanism and test its effectiveness as soon as possible, though we plan to keep prototyping and iterating this concept as well as our other end effector concepts. (Can be seen in the first picture)

Our second robot team (also dubbed ‘beastbots’) has also done some CAD for a telescoping arm concept, a two-joint arm, and also tested the feasibility of a cone shooter (no pics of that sadly).

Mechanical
Our mechanical team has mostly been working on machining prototypes this week, as well as machining some parts for our drivebase(s), and improving organization. We continued testing Triple Helix’s intake and added 2 motors to it, as well as started preparing to mount it on a stick to the triple helix drivebase. The OTT intake has been assembled, and the pneumatic gripper is partially done. The handoff intake and the claw intake are ready to be machined.

Our REV Swerve modules finally arrived (with about a week’s worth of delay) and we assembled them all on saturday. So far we’re pretty pleased with them, though we’re monitoring chief for any updates from other teams since we already saw some things that weren’t mentioned in the instruction manual (here’s a notes doc). We haven’t been able to actually make our drivebase drive yet, though. We will have a separate Maxswerve post next week detailing our findings!


On another note, we are continuing our quest to increase our organization: we just got a nice new rolling tool chest.

Software
Our software team has been experimenting with apriltags through photonvision, running temporarily on a Raspberry Pi 3 and a Lifecam. We bought an Orange pi 4 LTS for the future (here’s a comparison sheet of coprocessors). We’re also looking into snakeyes and the new limelight 3. We also plan on having the ability to recognize reflective tape, but we have more experience with that.

We wrote swerve code for the first time- we diagnosed some issues with it after running it in simulations. We used the REV library as reference. In parallel, we’ve also been working on writing drivebase code for the tank drive triple helix drivebase.
Our new members have continued working on their Romi courses and also learning about PID control. We’ve also started exploring autonomous routines for swerve- we decided that we would most likely use path planner or Helix Navigator.

Electrical
Our electrical team has been very busy flashing electronics, configuring motor controllers and crimping our new motors. We have has trouble crimping our NEO 550 motors, we’re gonna use wago connectors instead. We also wired the triple helix drivebase and the Swerve modules.

That’s all for today Wildcats (that’s an easy chapel hill high school joke, don’t mind me. And to quote Lucien: geez is it already 3 am? - it’s not but it feels like it)

  • Andrea
6 Likes

These designs are amazing! I can’t for the competition!

1 Like