FRC 1757 Wolverines 2023-2024 Build Thread

Moderato Cantabile

(moderately and singingly)

Build Season Week 5

Intro

Another build week means more things to share! This week we’ve been very productive with finishing up some of the CAD, starting to cut and build our intake and shooter, and managing some other business.

CAD! CAD! CAD! CAD! CAD!

Climber

Here is the climber. We made additional changes.

The main limiting factor behind getting the drive rails completed is finishing the climber to a point where we are happy with. A major constraint is the space to drive it, given that it must have enough force to pick up ½ the bot in a very cramped area.

These maybe-mounts for the passive hooks we initially were considering for simplicity, but will be opting instead to use the shooter A-Frame in order to maximize our internal space usage.

Elevator

We manufactured some parts for the elevator.

A key take away for this is to always check your zeros when CNCing. We had to redo one of the beams from a very slight offset. We hope next week we will be able to test our rigging ideas for the elevator.

Current main

Here is our current main.

Stuff we built

Intake

Over the weekend we cut out the side and carriage plates we needed for our intake. We currently have the first prototype assembled, but we have learned a few things during the process. For one, we did not CAD the fasteners, which turned out to be a mistake on our part during the assembly process. The rollers had to be cut down in order to give way for the bearing attachments.

We also opted to wrap the roller tubes in silicone rather than neoprene, which ultimately proved to be much a more arduous task to put on, but is certainly more grippy.

Here’s a video of it in action:

Sticker + Batteries

Over the weekend we had some fun with making stickers. I love stickers, you probably love stickers, we all love stickers! One of the first things we did was make stickers for our batteries, as we just got new batteries and they needed labeling. It was also a perfect opportunity to refresh our Cricut skills, so these are how they turned out:

We also took the opportunity to mark the batteries we got at worlds last year, and didn’t already have a label

A Note on Labeling Systems

Last year we made the decision to adopt a new Numbering/Labeling scheme for our competition batteries. The goal was for the names to be distinct from each other and simple. Labeling them 2023-1, 2023-2 ect wasn’t a good idea in our opinion. Many ideas were kicked around including naming the batteries after Pokemon, seems like a great idea as a list to draw from, there’s lots of options (at least 151 in fact) but after an hour-long debate between the mentors and students over which generation Pokedex we should use as a source of names we realized that the amount of choice was getting in the way of simplicity. Eventually, we settled on the Greek alphabet but that character set is going to run out next year given our current rate of battery acquisition.

Bumper

After that we decided to run a test with making a sticker and printing it onto some fabric, which was very successful. We created a happy flower for our first test. Our goal for this year is going to try and make our custom team number stickers for some extra aesthetics. Last year our bumpers ripped quite a bit and we had to repair them for DCMP, so this year we are trying another bumper fabric, but the consequence is we now have to do our own bumper numbers

Camera mounts

One of our team members printed out some camera mounts made of TPU plastic. We really like how flexible and durable it is, it is the first TPU part we have on this bot. We tactically borrowed these from 6328’s cad. Code has not had a chance to implement these positions and test vision.

Outro

It’s been another productive week here at 1757. Our midweek and next weekend will be among the most critical as it is the week before February break, so expect a lot (and we mean A LOT) of tasks to be completed/started during that time. So now it’s a matter of how much we can take advantage of our surplus of time. In the meantime, we will update you on everything that happens, and as always, please reach out to us if you have any questions!

Best wishes,

Team 1757


Written by: Baili, Liam, An, Luke

P.S.

Programming post coming soon…

7 Likes

Hey all, it’s Landon (vice technical captain of 1757) here to give a couple more updates on that intake I designed and we now have tested more extensively earlier today:

  1. As you can probably see, we are using two distance sensors (linked in a previous post above), which are wired into the motor to be used in code (in that video they are not active). Through some testing with the Rev Hardware Client (using them as limit switches), we found that activating it on the lowest sensor yielded the most consistent and important result (that the note stays in the intake). We have made (but not tested) a code change that will use that behavior when intaking, but instead use some simple logic to center the note in the intake while holding by moving such that both sensors become active.
  2. You may be wondering about the issue of torque as we are direct driving with the Vortex onto this: while we were too based off some issues, we have apparently solved them. Effectively, it came down to the motor not being able to continue outtaking when a note is held due to the resistance: we realized it was more due to the angle with the floor (if we lifted it can go). We also added a bit of extra space on one roller that was rubbing which almost fully removed the issue.
  3. To improve our torque situation, we opted for voltage control in the Hardware Client over percentage, which is what we used in this video. Using voltage made it consistently spit out the note from a stop, which percentage only did if the motor was stopped and then started again.

If anyone has questions, feel free to ask!

6 Likes

Prog Post Numero Uno

Alrignt Chief Delphi! Its that time of the year again! When enough students sit down and say “lets do code for a whole day” and thats the only thing that gets done.

Luckily timing worked out perfectly last Tuesday with our school system calling a snow day, and given that compared to last year our technical captain kept saying we were 3 weeks behind we decided to sit down and just code code code!

Note simulation

A strong part of our codebase every year is being as close as we can to being able to run complete driver simulation, able to be run before the bot is built.

For this year, that means taking the output state of the shooter, and showing how the note will fly through the air


A pretty simple class, just ballistic physics simulation

One piece that our simulation currently lacks is accounting for how spin changes the note trajectory. Currently we just take the average

Note sim in action

Field constants!

So since 6328 didn’t have very kindly borrowed constants in their code, we decided instead to just make our own. How? Uploading the field into blender!

One of our software guys has been doing blender for like 4 years, so he loaded up the glb file from AdvantageScope, aligned blue origin, and then took positions of elements.

While we don’t have a seperate fieldconstants.py file, you can find our constants here

The center position of the speaker is positioned in the middle of the opening, which is used for align and trajectory generation. If teams want any other positions on the field we’d be happy to generate them!

Trajectory Calculation

We already did the math for this, so implementing it wasn’t that bad

and here is the result

Another promising sign is that two individuals made each part, the trajectory generation and visualization, and since they match up with eachother it gives confidence in how this will work on an actual bot.

This does require us to have vision though, so we will also need to tune our vision and have our full field positioning system working

Moving shot

Alright! Theory explanation!
So consider a bot moving to hit a target, relative to the bot, the target is moving in the direction opposite to the bot velocity. The distance that the goal relative to the bot will travel depends on how long it takes to get to the target. Uh oh! Problem! Now that you compensated, it will take a different amount of time to reach the goal location, so you update based on the new timing estimate for a few iterations until its within reasonable error.

In practice, it looks like the following:
since our trajectory generator function returns airtime to target, this is pretty simple


The function also accounts for the proper angle the robot should be at in order make the shot.

In simulation it yields the following result:

Nom nom nom (note intaking simulation)

In order to really complete the whole “driver simulation” goal, we need to be able to have positions to simulate the state of the intake, which means putting notes on the field. Given our field constants we already made, putting the notes into simulation wasn’t difficult.

Thats the entire note simulation! We also simulate our beambreak sensors hooked up to limit switches on motors in this process, allowing further code to be simulated properly. The result if the following

Its not the most complex simulation. One easy improvement is changing the pickup position and radius to match the real intake, instead of just saying “note is within 0.5m of the robot”

Ensuring team success

This is more of a sidenote, but one of our code policies is to have at least two other students of the team review the code before being merged into main.
This has two main pros/cons

  • Pro: multiple students on the team know whats going on with code, not a single source of failure
  • Con: if one student has a bunch of back to back, individual complete changes, they have to wait for other students to completely understand to use on the bot

The con part has hurt us especially nearing competitions last year, where we just had one student who made a lot of last-minute integration changes to fix the bot, but we had a “magic branch” which defeated the purpose of main.
More students this year have stepped up, and we hope that by just finishing faster these on-the-fly changes are mitigated

Ok cool. So what?

We historically get our bot done criminally late in the game. (the drive rails havn’t even been cut yet) simulation and code helps us stay ahead and ensure that the time between robot being mechanically and electrically done, and when the bot can be competition-viable is reduced as much as possible.

Oh and everything shown in this post was done in 24 hour timeframe
Its a wonder what no school and dedicated students do!

Questions? Comments? Concerns? Compliments? Want to know what songs are in each video? Feel free to ask away!

Best wishes,

Team 1757


Post written by me! Luke!

15 Likes

hi landon

3 Likes

send carbon fiber pls and thanks

2 Likes

u can have it if u come pick it up >:D

2 Likes

Allegro Con Brio

Build Season Weeks 6 AND 7 (February Break)

|449x203.6959701824396

Us during weeks 6 and 7

Intro

OHHH BOY! WE’RE SO BACK!

It’s been two weeks since the last build update. Since then, we have been effectively utilizing the time generously given to us from February break to make substantial progress on our bot. That includes all of our subsystems and our bumper, so we have a lot to go through.

Bumper

Like last year, we are resorting to a two-piece bumper for easier access, and like last year, the bumper mounts are based on 971’s design. We also decided to customize our font for this year instead of using the FIRST KOP predetermined ones. We cut out our own numbers using Cricut, to which we then printed onto the bumpers fabric. We bought VX-21 X-Pac ripstop fabric so our bumpers will resist more damage. Last year our bumpers got really worn.

|359x239.6536175046749

Our finished product

One slight issue that we did come across was slight sagging. Some imperfections allowed the bumper to remain less than ½ inch above the ground, but ultimately, it was an easy fix.

|770.3703703703703x516.2857142857143

Notice how the corner of the bumper was slightly close to the table

We really love how these turned out. We hope this year our bumpers won’t rip as much, as we are using a different bumper fabric this time. Sadly, if the bumpers do end up ripping we will need to re-print the numbers again.

Elevator

After much messing around in the CAD, we have constructed the Elevator.

|564.9178001555038x375.99999999999994

In progress

|574.3008849557522x381.77347473622734

Standalone

|383.7443609022556x256

This is how it looks in the context of the drive base, when put into place

Here is it operating by hand :

While assembling the elevator we did run into some issues of bolts getting stuck and being difficult to install which caused us to waste a lot of time on a few bolts.

We later figured out that our CNC router’s tube fixture origin was off, causing cross holes to be misaligned. Since across the 2x1s we are using holes as clearance, we pick one hole and can match drill.

Intake

Our intake is done and oh me oh my is does it look awesome. Nothing has really changed too much in terms of how it works, with the only big thing was while constructing the intake we realized that we could fit a tensioner in for the chain. Before that we didn’t think it was possible when we were checking the CAD, lesson learned from all of this: double check everything.

Currently the intake takes in a note and then passes it off to the shooter from the back of it.

|632.7206477732794x420.62331166460103

Shooter:

After some testing with our first prototype, we decided to rebuild the shooter with new smaller side plates because the initial large size caused interference with the handoff mechanism.

And then when actually building the shooter, we also conveniently discovered that we cut the side plates with the wrong bearing hole selected, delaying construction even further (we all love when that happens).

|897.0520148952822x595.8766947841424

Shooter in disarray, workbench messy

|452.50223290965295x301.94370300207527|430.52279635258355x287

Shooter in its completed state, workbench clean

|427.86026200873357x320.3688562225367

The shooter in context of the main (still a work in progress)

It turns!!!

Outro

It has been an extremely productive two weeks for us here at 1757. Our bot is finally coming together after weeks and weeks of hard work. As we enter the final stretch of the build race, we will update you on everything that happens, and as always, please reach out to us if you have any questions!

We hope you enjoyed this megapost!

|642x1141.4229078736003

Best wishes,

Team 1757


Written by: Baili, Liam, Duanmu, Henry

6 Likes

Moderato Apassionato

Build Season Week 8

Intro

Hey all, welcome back to another build post. We hope that if you participated in a Week 1 Competition, it went well for your team. As for us, we are still chugging away with construction. This week, we have nearly finished our climber subsystem as well as our wiring.

Climber

We stole parts from Andymark Climber in a Box from parts that our 2022 bot could have been. Instead of the 2-stage design that Andymark sells, we are modifying it to a 3 stage design to give us the extension we desire.

|1665.4860767740533x1112.5529601393218

Climber in progress

Shooter

For our shooter, we attached the gearbox and some reinforcement plates to the shooter in order to support the gears on each side, attaching top/bottom for each side, so it meshes better. Before, the shaft holding a gear and pulley was only being held on by 1 bolt-washer side and the tension in the belt (not very ideal).

|228x329.52581476800015

Our other modification of the shooter was the addition of another small side plate. This was added to create extra stability, as currently, one of the gears is only being supported by a hex shaft and a bearing. Adding another side plate would mean that the gear is meshed better with other side plates, giving it more support.

|719.5076113954538x478.2859883327814

Shooter - not significantly changed

We glued gears to shafts to reduce tolerance because our research says we should have half an inch of tolerance in order to have the accuracy we are desiring. We employed this technique last year with good results, however this does mean the gears used cannot be reused.

Elevator

Our elevator had a few changes made to it. We realized a fundamental flaw in our design with tensioning: since the gears lock together the two belts, it wasn’t possible to tension the entire belt path without first disengaging the gears. An issue we noticed while putting on the moving hook was that it would hit the elevator. This wasn’t an issue when we were looking at discrete configurations, but we originally didn’t account for the in-betweens. While the moving hook would not intersect during the topmost and bottommost spots it would intersect during in-between spots. As a result, we cut one of the elevator bearings on the elevator and replaced it with a plate and teflon tape…

|459.71381352541x305.41136788689647

We changed this

|682.0413669064749x455.3296370853156

Elevator in profile

Updated Main

After installing the shooter and parts of the elevator last week, this time, we focused primarily on wiring the bot as a whole. This is in preparation for testing the code during the week and weekend, so we’re looking forward to that.

|477.3034351145038x318.1484741960773

Drive base - Bottom view - look at all those wires

In addition, we also just attached the intake. For those counting, that makes 3 out of the 5 subsystems that are available for testing. Expect a lessons post based on what we discovered from testing.

Closing

This week has overall been pretty busy, and our bot is getting close to being competition-ready. We have two full weeks left before our first competition at Revere High School, so we hope to wrap up programming and building and begin testing our subsystems. We will update you on everything that happens, and as always, please reach out to us if you have any questions!

Best wishes,

Team 1757


Written by: Baili, Liam, Duanmu, Henry, An

3 Likes

It moves!

the last couple of posts have come out late, but since the bot is mechanically together, software has had a chance to start progressively adding bot functionality, here are a couple videos from the past week of software finally getting hands on bot

Questions? Comments? Concerns? Compliments? Let us hear them!
Team 1757


- Luke

6 Likes

Fortissimo

Build Season Week 9

Intro

Hello everyone, we hope you had a good build week 9 and a good competition week 2. This week was focused on finishing our climber and achieving critical completion on our robot.

Climber

We learned some mistakes while assembling the climber. One of those mistakes was us misinterpreting the CAD because of how complex it was to understand, causing details to be lost. In the future we want to make sure our CAD is more readable so anyone on the team can assemble parts easier with less issues.

The climber works by having springs with a constant force that always want to push the climber to max extension, but there is also a rope that connects to the top and to a motor at the bottom which constantly holds it down. During climbing it simply releases it upwards and then pulls back down where it then connects to the passive hooks at the bottom of the robot. These passive hooks simply just move to make space for the chain then go back into place holding it in using springs.

Here’s the climber in action:

Main

Everything has been built, things like climber just needs some fine tuning. Next week, we will be primarily focused on the software of our bot.

|411.16732684131125x273.4964700789897|460.9307178944404x306.05968142403026

Wiring

We are almost finished with wiring our drive base. For the orange pi’s we attached it to our main, and have begun the process of taping down some parts on our main so they are protected. down. Something we learned while wiring is that when designing the robot you should factor in how tight the wires are.

|593.3055555555555x390.893523579836

Passive Hooks

We tested the bot’s passive hooks to examine its performance on the chain. The bot, as expected, tipped forward allowing the intake to successfully score into the trap. The bot with and without the battery and bumpers still gives us the same results as the intake in the front is where most of our weight is.

While testing the passive hook we noticed that the trap we built was 2 inches higher than the schematics causing our intake to not fully go into the trap’s entrance. After lowering the trap by two inches, we got the result we wanted with our intake fully going into the trap’s entrance.

Programming updates

Note Camera

Part of our plan is to use robot vision to have a note camera that will find the closest note on the field for us and automatically go over to it. It works by aiming a camera at the floor and then using photonvision to get the angle of the note relative to the robot. Some math is then done to get where the note is relatively on the field, then the bot goes over and picks up the note.

April Tags

Right now with April Tags we’ve seen a lot of talk about inconsistencies with them and have been brainstorming ways to try and mitigate the issues. One of our ideas is to make a field calibration similar to 971’s field calibration tool. The idea is to have pathplanner automatically generate a path based off of the april tags. At the moment we are still considering it, and we’re still evaluating the cost-effectiveness of implementing it since we are cutting it close to our first competition.

Closing

Overall, this week has been a busy one, we have made good progress on the climber getting it assembled, as well as making good progress on the drive bases wiring. We are entering crunch time, as we only have one full week left before our first competition. We will update you on everything that happens, and as always, please reach out to us if you have any questions!

Best wishes,

Team 1757


Written by: Baili, Liam, Henry, An, Duanmu. Edited by Sean, Luke, Liam

6 Likes

Such a neat robot architecture! Really looking Forward to seeing you guys in action.

1 Like

Looking good!

In most cases of using adhesives on metal assemblies you can disassemble them with heat. I bet you could take everything apart with a heat gun or small torch if you needed to.

1 Like

Vivace Maestoso

Build Season Week 10

Intro

Hey all, we hope everyone had an exciting week 3 of competitions. This week, we focused on finishing our bot, programming autos, and doing extensive testing to make sure we were ready for our first competition at the NE Greater Boston event this weekend.

Programming

Autos

Over the weekend we’ve been tuning our autos to make sure they’re all ready for our competition. While testing autos we did run into a few issues as in the beginning as our autos were not as smooth as wanted them. Part of our issue was we had our gains set too high causing our autos to not be as smooth. Another issue was our acceleration and velocity were kind of weird near the beginning also, along with some of our vision being a little inaccurate. Near the end of the day we were able to get some more progress towards autos.

Here are a few videos of some auto runs (not all of them were good):

somehow driving through the stage

bumping into the speaker

nobody was hurt

Final Preparations

Sponsor Panel

To accomplish this, we utilized a Cricut Maker and the following glossy vinyl paper recommended to us by Team 1153. We took advantage of Cricut’s “Print-Then-Cut” feature to first print the logos and then cut them according to their outlines. At first, we struggled with printing the stickers as the ink rubbed off the page, but we solved this by first toggling “Use System Dialog,” and then adjusting the paper to be “Heavy Glossy” in preferences, as well as lowering the quality to “Draft Quality” in advanced. After printing all of the stickers and putting them onto the panel, here’s the final result.

This panel will go on the elevator part of the robot, as that section was the easiest to put the sponsor logos on.

Picture of the panel on the bot in CAD

Parts

We’ve been manufacturing extra parts for our bot in the event of part failures. For example, during testing, we noticed that one of our intake side plates broke fragile during testing. This is one of the disadvantages of over the bumper intakes as if your intake isn’t strong enough, it can just break on you. To prevent this we decided to put a 1/8in aluminum reinforcement plate onto the intake plates in order to prevent the plates from breaking so easily.

Recap of Robot Skills

Here are the final skills of our robot ahead of Week 4 of competitions:

The intake will be able to elevate, so it can collect both notes from human players at the source from the ground and the floor. This approach also allows us to score into the speaker and trap. The elevator uses 2 falcons driving the intake up and down, and also helps us for climbing. The elevator subsystem then works in parallel with the climber on the trap in order to help us score into the trap by having us lift ourselves onto the stage’s chain and then having our elevator lift the intake up to the trap to score. The climber we have has a custom hook and uses parts from the andymark climber in a box. Along with the climber we also have 2 passive hooks that will allow us to keep hanging once attached. Our shooter is separate from the intake, giving us the precision we need to make far shots.

Here it is, our final robot. Meet Proteus!

Conclusion

10 weeks ago, we started planning our bot. Now, after many busy weekends and nightly calls, it is done! Everyone here at 1757 is very excited for our week 4 competition in Greater Boston this weekend and we can’t wait to see our bot in action on the field. We’ll cover everything that happens, and if you have any questions, as always, please feel free to reach out to us! Prior to Revere, we also plan to finalize our technical binder, stay tuned for more info!

Best wishes,
Team 1757

Written by: Liam, Baili, An
Edited by Sean, Liam, Baili

9 Likes

As we finish Packing for the WPI District Event this weekend, I wanted to share the First Edition of the 2024 Technical Binder, which we use to serve as a living documentation of the robot, and hopefully explain why we did some of the things we did.

1757 2024 Technical Binder_WPI Final.pdf (4.6 MB)

Like last year, From this point on, we will be publishing updated versions for any subsequent events we qualify for.

-Steve

4 Likes

Such a cool design!

Did you guys run into any issues with so many of your mechanisms in a cantilevered position? This game is pretty brutal so I was curious how things were holding up.

I also noticed at your first event some stray Notes were lodged in your bot, any plans on how to prevent that or alleviate that if it happens again?

Best of luck! We’ll be rooting for you.

1 Like

Not on 1757, but this did happen to them in a match.

The intake got hit when it was outside of FP, just happens cause OTB + cantilever, but they got it replaced before their next match.

I also noticed at your first event some stray Notes were lodged in your bot, any plans on how to prevent that or alleviate that if it happens again?

The geometry of the inside of the bot makes it really difficult we attempted a fix for playoffs last weekend but it was pretty ineffective. i suspect Driver Practice with the bot is probably going to make the most amount of difference and luckily we were able to spend a few hours on the WPI practice Field this week

1 Like

Upon reviewing video (shoutout using an Instacam X3) and code logs, we’ve determined that its self-inflicted from code issues and driver override. One of the major issues of the bot is the intake roller is a 1:1 with the motor, and if there’s any unneeded friction we can stall the note at 80A instead of doing our centering sequencing, getting stuck midway.

The driver has the ability to force it down through using the “feed” button, consequence of the feed button is that if it magically fixes itself and stops stalling the note can get caught inside. We are working on software and driver changes that will prevent this self-inflicted indicated from continuing.

1 Like

Greater Boston Post-Comp Review

(sorry for the late post, it’s been a busy week)

Intro

We finally got to see our bot in action on the field in the long-awaited New England Greater Boston event at Revere High School! After an entire weekend of practicing, competing, and improving, we feel pretty good about our initial impressions, although there is still massive room for improvement. We have a lot to cover, so let’s just jump into it!

Our Competition in Statistics

  • 17th in OPR
  • 7-5-0 record in Qualification
  • 20th in EPA
  • 16th in Auto EPA
  • 20th in Teleop EPA
  • 17th in Endgame EPA

All data shown here is from Statbotics

Summary: we were firmly one of the mid-table bots at Revere. Not particularly excellent at anything, but we are there when we need to be. This notion translated very much into our on-field performance. We could also say that we underperformed per expectations, not just because we performed worse at this event than last year, but also because we underperformed on EPA given our rank (based on the graph). We benefitted somewhat from having stronger teammates in earlier matches, but we shouldn’t let that discourage us from improving.

Technical Issues

Intake

During one of our qualification matches, our intake’s spline-xl pivot shaft bent, causing our intake to not work properly. We fixed it in the end, but we learned how to prevent it in the future. While we were driving we had our intake out, and when it collided with another team’s bot it caused the shaft to bend. In the future, we’re going to try not to drive with it always out since that could cause history to repeat itself.

Our intake backstop also bent inwards from the shooter wheels hitting against it. To fix this we plan on changing our endstop spacer of the shooter by using an acetal spacer instead of an alu spacer.

Climber

During our qualification matches, we found out that our hooks could only just barely reach the chain, making it virtually impossible to hang and score trap notes. To fix this issue we redesigned our hooks to make sure they slid onto the chain easier.

Performance in Qualifications

Data for our team TELEOP ONLY (amped speaker, unamped speaker, amp, miss):

M3: (0, 1, 2, 3)

M7: (0, 1, 1, 3)

M23: (0, 0, 0, 1)

M29: (0, 1, 4, 2)

M33: (0, 1, 1, 2)

M41: (0, 0, 4, 0)

M45: (0, 0, 7, 1)

M59: (2, 0, 8, 0)

There was a clear story to our progression in qualification. Especially in our earlier matches, we struggled to shoot accurately into the speaker, meaning our conversion ratio was a lot lower than we wanted it to be. This was mainly down to the aforementioned struggles regarding intake feeding, and since we couldn’t climb, our endgame wasn’t great either. However, as is often the case in Revere, our performance in the latter stages often made up for initial struggles. Especially in matches 45 and 59, we proved capable of being an effective amp bot, and we did eventually improve our speaker accuracy. Even though we won neither of those matches, it did prove that, if paired with the right bots, we could prove to be effective. Also near the end, we started overperforming our EPA and getting consistent one-note and leave autos, putting us in a good position for alliance selection.

Scouting

For our scouting, we learned a lot. We reverted back to our old system of Google Forms and spreadsheets to collect and analyze data. The feedback we got from our team has shown us that it needs some work, at the moment it feels hard to work with and could be simpler. The system we used was difficult in how students had to type out the amount of Notes that were scored, so they had to look away from the game to write that down. Moreover, some questions could have been more specific or clear.Alliance Selection + Performance in Playoffs

During alliance selection, we were selected as the second pick of the 5th alliance with teammates 2877 “Ligerbots” and 246 “Lobstah Bots.” An interesting note is that 246 were a lower rank than us, but ultimately a more attractive first pick because they overperformed on EPA.

Our first match was against the 4th-seeded alliance, and our first course of action in tele-op was getting a note stuck in the drive base that we couldn’t get out until we bumped into another bot halfway through teleop. We were able to score a few amp notes and one speaker note, but 2877 and 246’s efforts were already more than enough to move on in the upper bracket.

Elim 2 - note gets stuck inside our bot

We did not get a note stuck in ourselves in our next match against the formidable 1st-seeded alliance, and we served as mainly an amp and occasional-speaker bot. Despite increased contributions, we couldn’t come close to their firepower, and we dropped into the lower bracket, where we ultimately fell to the also-formidable 3rd alliance.

We appreciate teammates 2877 and 246 for being gracious enough to pick us to join their alliance, and even though we would consider ourselves unlucky in the playoffs, our season is far from over. If we pick up where we left off at WPI, there is plenty of time to improve.

Outro

Overall, we are content with our performance, but we still have lots to improve on. Our season has only just begun, and now we must feast our eyes on WPI. In particular, we would be incredibly happy if we managed to improve our speaker consistency and figured out how to climb the chain, so there’s lots to work on and look forward to this weekend! If you’re going to that event, feel free to stop by our pit and say hi, and as always, if you have any questions, feel free to reach out!

We keep marching! Let’s CRESCENDO!

Best wishes,

Team 1757

Written by: Baili, Liam, Henry

Edited by: Baili, Liam, Duanmu

5 Likes

WPI Post-Comp Review

(trust us, we tried getting this post out as soon as possible)

Intro

Well, that sure was something! There is no looking back now, as we are officially off to the New England District Championships after a much-improved performance in our 2nd competition of the season at WPI. After less than a one-week turnaround from Revere, we were able to demonstrate more of what our bot is capable of, and on top of that, we earned ourselves a very-much-welcomed District Impact Award! There is a lot to discuss, so we better accelerate into it!

Our Competition in Statistics

(All data is pulled from Blue Alliance, Statbotics, and 2370.Viperscout)

  • 6-6-0 record in Qualifications
  • 21st out of 36 in OPR
  • 14th in Teleop Note Count
  • 9th in Notes placed in Amp
  • 12th in EPA
  • 13th in average points contributed
  • 10th in Teleop EPA
  • 23rd in Endgame EPA
  • 10th in Melody RP EPA

LOOK AT IT WE ARE RIGHT ON THE LINE

Technical Difficulties

Intake

In our first match, we had a problem with notes slipping in the shooter, being prematurely ejected into the air rather than being able to be fed into the shooter. Fortunately, the problem was addressed mid-match, and our intake was mostly effective for the remainder of the competition. The issue stemmed from how one of our rollers shattered, causing our intake to act wonky. We tried fixing it, but didn’t have enough time so we had to go a match with it broken.

Shooter

One problem that continues to bug us until now is shooter inconsistency. While we feel that our accuracy has sufficiently improved from our time at Revere, there is still equally sufficient room for more improvement. According to our own scouting data, we had roughly [%] accuracy on the speaker, and more often than not, we would resort to placing notes in the amp so that partner bots could shoot amplified notes rather than the other way.

video

To fix the issues of inconsistency, we plan on switching our shooter’s motor over to a Kraken motor as those have been more reliable for us.

Performance in Qualification

We had a very similar pattern in our performance in WPI as our performance in Revere, except this time, the first half and the second half had even starker differences. In fact, they were more or less complete opposites in terms of results. For the first 6 matches, we had a 1-5-0 record, and we were languishing in second from bottom in the table. However, for the final 6 matches, we recovered with a 5-1-0 record, with all 5 of our wins coming consecutively at the end, which was kickstarted by a “revenge” match against 1768 “Nashoba Robotics”, whose alliance soundly beat us the match prior. From that point, we were constantly leading our alliances in terms of amp output, and having better luck overall with alliance partners and bot efficiency skyrocketed the team from close from bottom place up to a seemingly improbable 15th. After down and out in the first half of qualification matches, we are absolutely ecstatic with our recovery performance, and as we can see from the graph below, we probably deserved it.

It should also be noted that during qualification since our climber was actually working this time around, we were able to contribute to endgame in more ways than just parking.

EPA progression from QM1

Alliance Selection + Performance in Playoffs

During alliance selection, we were selected by 9644 “NEIA Robotics” and 5000 “Hammerheads” to join the 6th ranked alliance as their 2nd pick. In the alliance, we had the highest EPA out of all 3 bots while Hammerheads lead the alliance in average points scored during qualification (7th overall).

There was a recurring pattern during playoffs that applied to most teams, not just our alliance. They would often prioritize teleop points more than they prioritized climb and park points, probably because their score:time ratio was too low to be worth compromising. As such we only 3 endgame points in our first match, while the 3rd alliance we were up against only had 1. Ultimately, their superior teleop performance got them the win, so we dropped into the lower bracket.

Our next match was against the 7th-seeded alliance, but ultimately, by sloppily giving away penalties and Hammerheads dying momentarily, we lost again, meaning we were officially eliminated after 2 matches. Our team performance wasn’t exactly stellar, but it was already an achievement for us to reach the playoffs after such a rocky start in our qualification matches.

Impact Award

Prior to the awards ceremony, our runs of bad form in both Revere and WPI meant that qualifying for New England Districts this year as one of the top 90 bots was going to be harder than normal. Thankfully, to our delight, we received the Impact Award due to our team’s extensive outreach activity, meaning we automatically qualify for Districts!

Here’s what the judges had to say about us:

“This team’s howl is making their community’s interest in FIRST shoot upward. Their cool library of courses interests the fledglings of their colony in science and engineering. The work they are doing to educate and empower is just the tip of the iceberg. The FIRST Impact Award goes to Team 1757, Wolverines!”

We’re extremely grateful for this award and would like to thank our business subteam, more specifically our outreach subteam for making it possible for us to be in another moment of history for our team. Great job out there!

Outro

First of all, we would love to thank teams 9644 “NEIA Robotics” and 5000 “Hammerheads” for inviting us to join their alliance. We would also love to thank the judges of the event for considering us for Impact. They’ve all made this event unforgettable for all of us at Team 1757, and we are excited once again to be at District Championships. We will continue to keep working on our bot to make sure the issues we encountered during our WPI comp don’t happen again. If you are going to NEDCMP, feel free to stop by our pit and say hi, and as always, if you have any questions feel free to reach out.

|2048x1365.2289431276677

Great job team! Let’s CRESCENDO!

Best wishes,

Team 1757

Written by: Baili, Duanmu, Liam

Edited by: Baili, Steve

2 Likes