4561 The TerrorBytes | 2024 Build Thread | Open Alliance

Welcome to The TerrorBytes 2024 Open Alliance Thread, HYPE!!!

The TerrorBytes are ecstatic to be a part of the Open Alliance for the first time ever. Our goal is to provide consistent, quality posts that can help teams.

Who are The TerrorBytes?

TerrorBytes Robotics, FRC Team 4561, otherwise known as “The TerrorBytes,” have been competing since 2013. We are located out of Research Triangle High School in Research Triangle Park, NC. Important focuses on our team are “Building Better People,” upholding FIRST values such as “Gracious Professionalism” and “Coopertition,” spreading STEM to our community, and having fun whilst learning.

Currently, the 2024 TerroBytes are a team of 48 students and 15 mentors. Going into the 2024 season, we are excited to get Hype! for Crescendo and have such a strong, dedicated team to work with.

What to expect from us this season?

The main topics of our thread will focus on:

  • Design Decisions
  • Season Strategy
  • Match Strategy and Analysis
  • Brief and succinct summaries of events/updates

Stay tuned in the near future as we have a lot to talk about for what we have worked on so far this off-season and what is to come. Hype!!


Socials:

30 Likes

i am excited to see what the TerrorBytes bring to this coming season! :]

6 Likes

Can’t wait to see your team and follow along with your OA.

4 Likes

This is going to be HYPE!

1 Like

I will be judging heavily from afar.

2 Likes

2022-23 Superb Season Summary


To help get acquainted and contextualization with posting on Chief, here's an overview of what we did last season and a few of the issues we faced and design choices!

Build Season

At the end of our 2022 season we purchased swerve modules (sds Mk4i) and used them at offseason competitions. When Charged Up was released, we saw an open field. Thus, we opted to use them for the added mobility and speed across the field. We held design discussions and began prototyping rudimentary intakes and other ideas before we picked a set direction to head forward with on the team.
Design Choices

Our strategic discussions resulted in us wanting to focus on high scoring. Our design team came up with a double jointed arm, but with two versions. V1 was created using MAX Tube (Thx Rev <3) and is what we competed with the whole season. V2 was intended to be a 971 (2018) “Exotica” Arm in which we would compete with at later competitions. This was intended to be implemented later as we wanted to give a robot to our Programming sub-team with enough ample time to get everything working. They worked with a mock-arm (image below) until we got the robot to them.|

Meanwhile we were investigating 2 intake designs: a horizontal intake and an Everybot-style intake. We tested both of the designs, opting for the one we could mount and get tested quicker (horizontal. The Everybot intake was sidelined as we were still figuring out how to package it nicely, but still continued working on it if we did ever decide to swap to it. With the mechanical and electrical work finishing we gave the robot to the programmers and drive team to get us prepared for our first competition (FNC Week 1 UNC Asheville Event).

UNC ASHEVILLE

We came into Asheville with a somewhat working robot. We were happy to seed #2 and make it to the Loser Finals with our partners 5727, Omegabytes, 6639, The Mechanical Minds, and 6729, RobCoBots. We were also thrilled to win the Impact Award and Julian Berla winning the Woodie Flowers Nomination!!! Hype!

More Info

Day 1 Asheville saw us having some good matches, but seeing that our shoulder joint was struggling due to the excessive wear placed on our shoulder sprocket and chain. This led to it eventually stripping the sprocket, leading us to have to replace it on Day 2. Day 2 saw us taking the more conservative approach of doing what we can (which meant not scoring high cones to minimize any stress placed on our arm joints which we eventually saw failure during our second to last match of quals).

Post Asheville Thoughts:

  • Need to reinforce shoulder joint (add a second sprocket to reinforce the joint)
  • Tune arm further to score high cones better
  • Want to ensure we can keep similar performance and final ranking so didn’t want to change too much
Pictures


Wake County

The Wake County event went well. We were consistent enough to go undefeated and rank 2nd after quals. Thanks to our alliance members 587, The Hedgehogs, and 8758, Tar Robo Turtles, we were able to make it to the Finals. We also won the Innovation in Control Award and the Safety All-Star Award for our Safety Captain Siri and all of her work. Hype!

More Info

We were very consistent with minor connection issues caused in the venue as well as connection issues during hard collision. At this competition we implemented a high or low strategy where if our data suggested we would win a match we would focus more on scoring low to get links, however if a match would be close we would do high to try and ensure a win which worked well for us. Our alliance worked extremely well together, scoring over 100 points every match but with a tough match against alliance 3 dropping us into the losers bracket where we trekked our way back to finals. Where we were ultimately beaten by the powerhouse alliance 1.

Post Wake thoughts

  • Consistent high scoring would help with our playoff alliances
  • Pad the radio, ensure we don’t lose comms on collision
  • Change intake???
Pictures


FNC State Championship

States was when we finally implemented the Everybot, 2910, and 7890 inspired intake (there are more, but these were our main references). This intake was implemented two days before states, and was a scramble to get it working. This was our main upgrade all season as we wanted to increase cycles with a more touch-it, own-it intake. This was the only upgrade possible in which we could also deliver it to the program (Exotica arm was too out of the reach).

More Info

We had two matches during this competition where our robot disconnected and then proceeded to turn off. Initially, we thought it was a breaker issue, swapping it, though in the playoffs this problem occurred for the second time. We looked at our battery log and saw that it was the same battery allowing us to assume that it was indeed a battery problem. Though, we were able to sort this out and continue playing.

Also, adding a new intake leads to having tune new set-points in the pit. This is un-ideal and for this upcoming season we will not be changing code in the pit (programmers will talk about the caliber about this eventually).

Post States Changes

  • Repair taco’d sprocket
  • Investigate adding absolute encoders
  • Tune Arm for smoother motion
  • Drive practice
Pictures


WORLD CHAMPIONSHIP

States saw us take a break to give our children a break from the intensize build season. Though we still added absolute encoders onto the joints (instead of limit switches used to zero at the start of the match) to help make our arm positions work smoother.
Worlds was great, though technical problems (induced by attempted improvements on the robot) hindered our ability on multiple occasions. Though, we were happy to have performed the strongest we have at worlds this season. A great end to the season.

More Info

Post World Changes???

  • SUPER PIT (post soon™)
  • And More…
Pictures

We cant wait to share more! HYPE!!!

  • Alexander Trent (Engineering Co-Captain)

Some of these pictures were taken by Danny Levinson, we appreciate the work he does for FNC https://dannylevenson.com/FIRST/

Something Exciting

ezgif.com-video-to-gif

21 Likes

No thoughts, just need to highlight the intellectual milestone that was putting a (safety proofed) bench vise on an FRC robot

6 Likes

Very Belated Thor East and Doyenne East Recap


Thor East

THOR East was the first off-season competition we attended this year. We were happy with all the skills we learned and thankful for our alliance partners 7443 and 6729, helping us make it all the way to the losers finals. This event was “chill” and we were there for that.


Doyenne East

The TerrorBytes also attended Doyenne Inspiration East. This event was created to focus on the inclusivity and empowerment of young female and nonbinary students through FRC. We were happy with our performance in this event and the opportunities our students had on and off the field at this competition. Thanks to our alliance members 8429 and 6004.


Key Takeaways

  • Drive Practice is important to build consistency in drivers. At these events, we were rotating through new drive teams and had not allotted enough time to make all of them comfortable (most got some practice, but it was very inconsistent)
  • We ran TPU wheels here and plan to continue at future events
  • The robot needs work on it. Much of this event was spent jankly tightening the chain to make it work. We plan to make swapping a bent or stripped sprocket much easier for future robots. The arm is currently being rebuilt for our next event due to our main shoulder sprockets bending and stripping teeth. In hindsight, this has been a problem with us all season and we would love to focus on improving our DFM next year.

Future Plans

  • We are excited to attend Rumble in the Roads next week and have been prepping for that
  • We are working on a cube shooter currently (post about that will be soon)
  • We are adding limelight and allowing programming to continue testing new odometry stuff in preparation for next season (programming post soon)
  • Super-pit is continuing to be worked on (post soon)
  • Arm design is being modified to help reinforce our joints
13 Likes

Future post is gonna be SICK!!! :fire::fire::fire::fire:

3 Likes

Can you expand on your experiences with these, both in making them and how they held up over the weekend?

I wrote a post detailing stuff earlier. To summarize, prior to these events, we were mainly struggling with getting the “correct” infill that would be durable enough to last a competition and having tread leftover at the end of the day (at minimum). We ran 30% TPU 95A Overture Black Filament gyroid infill of the Tj^2 V40s with PLA+ hubs. The 30% were better in resisting any wear and could probably be stretched for a whole event (two days), though the current plan would be to swap at the end of the day or before playoffs to a fresh set). We are going to run different variations at Rumble on the two robots to get some more real match testing in.

Also here is an image of a wheel after Thor and Doyenne (same set on both days).

We did like the grip (it was at minimum comparable to the stock Neoprene SDS wheels). After Rumble, I want us to do more scientific testing of the wheels and how much more grip they do have (trying to achieve the 80%? more CoF that 88 found in their testing when compared to stock Neoprene wheels).

Edit: I hope that we can find if the ~28 hours of printing time per set is worth the 20 minutes it takes to swap new tread onto billets.

8 Likes

I would also love to share this wonderful workshop slideshow made by @StevoTheMadman. It details our full-season strategy and design retrospective. This follows our series of workshops that we have been conducting over the summer and off-season (more to come) to teach our team new skills.

7 Likes

It's been a minute, here is some recent things we worked on...

TerrorBytes Holiday Party

We hosted our Holiday Party last week. It was a fun event with many festive activities. Here are some pictures of us transforming our lab into a very spirited space. This was a great way to end the year and send off our team members into the Holiday season.

We turned super-pit (writeup soon) into a festive environment. Fulcrum wasn’t safe from it either.


MiniFRC: 9.5

We also hosted another edition of MiniFRC Charged Up last week. We were excited to see so many teams come out and compete (21). As always this was in the spirit of getting newer students trained with fabrication, electrical, design, and programming skills able to be learned from competing in MiniFRC. Marshall wrote a great post about how they, at 900, are continuing to push the limits with MiniFRC. As always MiniFRC information can be found in our previous posts and on our website (newly renovated, and very much still a work in progress). Big shoutouts to Alfredo Systems for again powering another MiniFRC. Thanks to all their work behind the scenes, we are fortunate enough to host these events. Be on the lookout soon for things Alfredo Systems and The TerrorBytes are cooking up for future MiniFRCs (summer maybe?).


Prepping for 2024

We have been busy behind the scenes “sweating” to be ready in time for kickoff (OMG 17 days!!! HYPE!!!). We have been working on training rookies through projects curated by our leads. Design is working on a 2018 Design Challenge together, Fabrication has been working on training on tools, Controls has been working on teaching people the controls system and getting them familiar with everything, Programming has been programming away, and Business has been cooking up many things including obtaining sponsors, outreach, and social media.

We have also been frantically trying to get the lab organized for kickoff (kind of there, kind of not). That is getting rid of the random stuff we’ve hoarded through the years, pre-stocking (ah the wonderous process of figuring out what the heck we need), and finishing super-pit. Shoutout to Fabworks for helping us get sheet-metal parts efficiently to have the super-pit “done” (it will never be done) before the break. We were able to finish up the top-shelving bits and get them inserted last meeting.

10 Likes

Nah, I just get to take blame for it. That’s 100% my amazing students and they really do deserve it for this one… them and you folks for putting on an awesome set of events.

2 Likes

Our 2023 Off-Season CAD Release:

MiniFRC Cad Releases (because why not)
Fun CAD From Rookies
25 Likes

These look very well done, what did you use as reference material to CAD these?

1 Like

Having designed the 254 copycat, it was a mixture of lots of time watching matches, BTB, and looking at their manual and robot’s pictures obsessively. The work 254 does is pretty cool and looking through their thread, they explain things with great detail. I do want to note that this 254 CAD is not perfect at all and there is a lot of things broken with it, but most of it works enough, and to me it was a finished product for something that wasn’t going to be made IRL. Mainly, it was just about learning how they did things and doing the things that made sense to me. It also was focused on the things that they did, I would love to take the opportunity to add the many improvements to this CAD in the future, just ran out of time before Crescendo.

9 Likes

Strategy Sheet Release


Here is a google sheet I have made that is a basis for modeling scores for different levels of robots. This was made by me in preparation for kickoff tomorrow with the focus on making this a quick process once the game and manual are released. Inputting in the different variables for Crescendo will help determine for us, and hopefully others, what strategies are strong for optimizing ranking points.

How this works?

Once the manual starts, input all the variables for their points here:
image

Hopefully, GDC makes it so that there is only one GP (game piece) and stuff like that, but having extra built-in for any situation isn’t always a bad idea. In the calculations, removing GP2 and Endgame scores is done nicely by being tallied up towards the final points separately.

Estimating cycles is the next tricky part. Using previous games though, one can estimate how many cycles different levels of teams can complete, allowing for this to be a scalar of the variables.

"Simulating" Matches

Once this is done, you can go to the sheet titled “Simulate Matches” to basically input the 5 base types of robots I have provided and what the final scores are going to be like.

Changing the robot levels in your match using the dropdowns can help determine what scores are needed to win against varying levels of robots while also giving insights into what strategies work nicely in optimizing the most ranking points. One thing to note is that the conditions for RP will need to be edited manually tomorrow, but the example formulas can be a great start for that.

Weighing Different Aspects of the Game

This is our 2023 weights for what we wanted to do in a season. We will create another, very similar one of these tomorrow when we figure out our strategy. This is a nice template for how we have done it historically, rating the things we want to focus on a scale of 1-5, or if we want to just ignore it.

Important Notes

This system isn’t a perfect way of simulating the game, but using safer/smaller numbers is always a good way to determine what # of cycles is correct, what match scores will look like (assuming there isn’t a limit in score), etc. As always feel free to ask questions and point out any flaws (I rushed making this today).

13 Likes

Crescendo Strategy/First Looks


TLDR: Intake ground, climb + trap, auto paths, shooting is fun, do some AMP.

First off here is an updated strategy sheet with our weighted table also inside of it designed for Crescendo. The logic isn’t entirely perfect, though we made it work for today.

Theoretical Strategy

Our team wanted a more structured approach to our season schedule and decided to implement a much more strict and structured strategy freeze, 1 week into the season. With our first look at the game and tried to analyze how our team wanted to play the game. We generated a weighted decision table to try and see an initial idea of what we want to do with the intention of relooking over it once we have had a bit of time to think over and look at the game.

First off, the lower the number the more preference we are putting towards it. Here are the main takeaways from our strategy discussion (all theoretical still).

  • Intaking from the ground is valuable to decrease cycle times and cycle more notes due to the ones started on the field already

  • Buddy climbs seem not to be worth as much effort as scoring a trap. A trap allows for one other person to climb and guarantee the RP, if we can find a buddy climb (i.e. some sort of in-line buddy climb + brace we would try to see if it’s implementable. Rather we want to focus on hoping our alliance partners to physically climb on their own early on and guaranteeing we can get the trap note. However, adding a buddy climb later is not out of the question. This will result in at least 11 points if another can climb in the match, securing the RP.

  • Analyzing the points awarded, if there is a match with a strong robot, scoring into the speaker is the most optimal strategy as after auto (which scoring into speaker rewards points), we estimated that if one robot is scoring if there were to be an amplification the robot reliably can only score one game piece (unless if some shuttling happens or it is early in the match). This would mean that for every 3 cycles, scoring solely in the speaker is only a singular point loss, as opposed to scoring two in the amp and one in the speaker. We felt that the point loss was worth it as shooting into the amp potentially has the ability for it to fall out, meaning shooting into the speaker from a further distance is more feasible. However, we are looking into how exactly we can try to use something to score into the amp as at higher levels the handful of points you gain per three cycles could be beneficial on your own and scoring becomes much stronger when you can score with multiple robots.

Autos and Match Play

We feel ambitious to prototype various shooter mechanisms to shoot from anywhere in your wing. Using the field CAD, we feel shooting from anywhere is feasible with enough margin to make it work. Shooting while touching the stage or from the bottom of the speaker is probably more reliable, though we will work towards tuning our design and prototypes to help reduce cycle times.

Here on the field diagram, we outlined guess optimal shooting locations for the speaker (indicated by the pink x’s). The light blue line is a 2 amp auto that would be huge for team play as right after auto all three robots can potentially score into the speaker.

The red is the speaker auto where scoring two into it rewards a nice 10 points. I wouldn’t be surprised if teams can make 3 or even 4-note autos, though we’ll see.

image

The red auto shows a 4-note auto, this would use a similar idea to what teams did in 2022 where they stop short of a ball and roll forward to get an extra game piece per stop, with only 4 notes per side I think it would be feasible. This auto would be potentially good during an earlier competition where we are less likely to encounter teammates who can score multiple notes in auto but cutting the amount of notes shot would most likely be easy enough if partners use those notes.

The green auto is an auto that could theoretically be used later in the season when other people have more consistent main staged autos we could do an off-side auto to get more notes closer for after auto, depending on speed we could in theory shoot as well but the potential to have 2 notes from further away close is very beneficial.

-Alex and Panav

18 Likes

Week 1 Update


With school (and meetings) getting canceled due to weather we got set back a little, losing our Tuesday meeting where we were planning to get a good chunk of work done and writing a lot of this post but we manage and still have a lot to talk about!

Game Elements and Prototypes

Our practice elements are coming together. Thanks to our fabrication team for getting these together we have more on the way with hopes of our game elements being done by the end of Saturday

Prototyping Questions:

We are working on getting more of these prototypes underway but we have gotten a start on the list with more on the way

  • Intake:

    • Ideal Compression → OTB Prototype
    • Roller Positions → OTB Prototype
    • Wheels Types → OTB Prototype
  • Elevator:

    • Calculate Torque
  • Climber:

    • Hook Designs → Design this
    • Angled Top Bits → Makeshift Elevator Prototype
  • End Effector/Shooter:

    • Compression → Shooter Prototype (Fully Built Version)
    • Spin → Hook motors to shooter prototype
    • Wheel → Put in stealth wheels, billet wheels, and weighted ones

Initial Prototypes

Shooter V1 prototype testing: we have had middling success trying to shoot with just our two flywheels powered off of drills and have plans to test with Neos to try for a more consistent speed on both sides as well as adding a top to the flywheel so we can more accurately insert the notes.

Test 1

Test 2

Test 3

Test 4

Battery Education

We just got some new batteries in and used this as an opportunity to teach our electrical rookies about battery best practices and how to set up and crimp some of the connectors.

Architecture

  • Be small → Current frame is 30x30, but we are going for around 24-28” x 24-28”
  • Btw, Alpha Drivetrain frame constructed
  • Be short → Optimizing Robot Height to allow for stage path
  • Be fast → Kraken X60 Drive, Neo V1 Turn
  • Score in Amp/Speaker - Ideally same mechanism
  • Climb onto Stage at minimum

KrayonCADs

Our design subteam brainstormed a lot of different architectures that the robot could potentially use and the two we figured would be the most competitive were these two designs:

Vertical Elevator/Offset Shooter



Pros:

  • Shooter Mounted Low
  • Can go under the stage

Cons:

  • Separate Intake Mechanisms for Trap and Amp
  • Packaging the second intake to work for Trap will be tricky

Titled/Vertical Elevator Integrated Shooter

Pros:

  • Same Intake for all Mechanisms
  • Handoff allows for the elevator position to be low while intaking
  • Elevator used for only Amp, Trap, and Climbing
  • Pivoting shooter allows for variable distance shots
  • Potentially Add a Buddly Climb in a later version

Cons:

  • Everything integrated into the elevator
  • Passthrough necessary to shoot
  • Passoff from OTB intake into the shooter

More Design Stuff

We are almost done with our Strategy/Design freeze. We recapped our goals for the season with our design subteam. Saturday, we will be working on finalizing what our robot architecture will be.

We took a closer look at our decision matrix to see what we feel are the most effective things to focus on when playing the game, noting at the bare minimum we need a ground intake to amp/speaker scorer with a climb. We also know how important the trap can be for getting the Ensemble RP. With that in mind, we want to do enough to try and trap to not rely on human players + double climb.

Low rankings/things we are not focusing on:

  • Source Intake > ground is superior
  • Harmony and Buddly Climb due to us wanting to maximize us on our own. By week 3, we expect to see almost everyone climbing onstage. Hamrnoy is nice, but not a direct strategy, though we will optimize being smaller.

In essence, we believe that some sort of elevator (angled or vertical) with some sort of pivoting shooter + ground intake would be able to achieve our goals the best, we plan to do a bit more diving into the geometry and the best ways for things to go together and share our findings with our team during our Saturday meeting.

CNC Issues

In addition to all the design talk, we managed to crack our baseboard for our shopbot when trying to cut prototypes!!! So we are working on getting that fixed ASAP.

Future Goals:

  • Finish Field Construction (1/13)
  • Finish Architecture and End Strategy Freeze (1/13)

- Alex and Panav

Fun Times

13 Likes