Ipswich TIGERS 5459 | 2024 Build Thread | Open Alliance

Team 5459 - Ipswich Tigers

Who Are We?

Team Ipswich TIGERS is based in Ipswich, Massachusetts. The team was founded in 2015, and has grown massively since! This year you can find us at the UNH and Pine Tree events, and hopefully even further as well. We also hold a yearly coding bootcamp for students and teams in the area every summer.

Team Structure

Our team is strongly student-led. We have five primary sub-teams: business, strategy, mechanical, electrical, and software. Each sub-team has a student executive who makes sure everything runs smoothly. We also have several technical mentors, primarily helping with software and mechanical. In the off-season, each sub-team runs their own training workshops.

Goals

Last season was our most successful so far, and we plan to keep that momentum. To us, that means preparing in the off-season for success during the in-season. Furthermore, our team is working on organizational and our project management skills to ensure a successful season.
We plan to have a robot done by Week 4, in order to give software and the drive team time they haven’t had previously. Furthermore, it is our goal to build a robot capable of being an alliance captain or first round pick at any event we attend.

Links

6 Likes

Updates from 5459!

Pre-season we’ll be doing one update at the beginning of every month, so one early December and then one right before kickoff detailing our kickoff plans!

Mechanical:

Since the last post, and during October the team made a lot of progress.
Mechanical made some sweet swerve module covers after an unfortunate accident where we sheared a Falcon 500 right off its face plate when attached to our MK4 module.

Here they are on the robot!

Also, the mechanical team has been running some shop training and build practice for newer members. Soon they will be doing some CAD review as well!

Software

Software made a lot of progress! We’ve been going over programming with new members, and have got our swerve autonomous working after previously being broken at N.E.R.D. Furthermore, we also have functioning vision tracking for AprilTags using our LL2+. I can’t attach the videos here, but please join the Discord and check them out!

Goals for November

Software:

  • Work on angle and extension control for a robot arm
  • Continue programming classes

Mechanical:

  • Continue shop training
  • Review CAD / CAD Practice robot
  • Design + build smaller swerve chassis

Business

  • Reach out to sponsors and design some sick team merch!
1 Like

Summary

We’ve had a very productive first week! Currently, the team is planning on using a single mechanism on a pivot to accomplish both intaking and scoring in the amp and speaker. We won’t be doing the trap, as the team found this outside of our resources and not within our strategic design.

Kickoff Day

On kickoff day, the team met for our yearly pancake breakfast tradition. Then, before the game reveal, we reviewed the previous years games (2008-2023). We found this useful to immediately recognize potential designs. Everyone then read the manual and took 1678’s Rules Test.

Goals

Currently, the team is working on our CAD. At the same time, the software sub-team is working on our subsystems and getting a scouting system up. Our goal is to have a complete robot by 509’s Week 0 event.

Design

We’re using a 25x25 chassis, with an extension and gap similar to that of Team 95. We’ll have our intake arm (similar to Quokka’s Quagsire) drop down into that gap. We’re slightly modifying the intake-shooter mechanism design from Ri3D Quokkas to have a full-width intake.

For our climber, we’ll react off the truss, with a single climber on one side of the robot with a axe-head designed to catch the chain on either side.

Controls

We’ll be using a LimeLight for AprilTag detection and a pivoting arm to accomplish a continuous range of shots from varying angles on the speaker. We’re also using PathPlanner for autonomous pathing.
We found tag detection to be accurate up to 27’.

Analysis

Links

CAD

CODE

4 Likes

I’ll augment Toby’s post with a few of the why’s of our decision making process:

Why is our need, want, and no list the way it is?
General: A small, fast robot that can go anywhere on the field, including the stage. This game has some really long cycles with very few protected zones. Being able to drive across the field quickly in any of the three lanes is going to be key to success no matter what scoring action is being taken.

Drivetrain: We finished a swerve drivebase in the offseason, this game favors swerve by having a completely flat field. Not a very hard decision there. Being unaffected by notes while driving was a requirement, whether that be by driving over them or pushing them out of the way. Don’t want to get stuck on one, don’t want them getting in our swerve modules.

Intake: Ground intake reduces lineup to the Source, particularly important given that the Source is across the field from the drivers. We prioritized width over being able to get notes in the corners of the field.

Scoring: We thought that, given that the kitbot scores notes in the speaker, we should be able to score in the amp for coopertition and amplification. Scoring only in the amp doesn’t allow for a great autonomous, so speaker scoring is also on the list. The trap is the first no here, as it only provides value in consistently getting the Ensemble RP. Little value in elimination matches, high difficulty, restricts climbing to only the center. Long distance shots into the speaker are also a low priority, but they fit with everything else we’re doing. Likely a project for later in the build season.

Climber: Given that we aren’t doing the trap, harmonizing on the chain is the best way to get the Ensemble RP somewhat consistently. The simplest climb in this game is two hooks in the center of the chain, so complementing that is the key to ensuring success in harmonizing. A buddy climb would offer some advantages similar to the trap in gaining the Ensemble RP, but is just as challenging. De-climbing is unnecessary given that we aren’t scoring in the trap. The side climb that reacts off the truss (shown above) seems to be the best way to climb consistently with either a partner scoring in the trap or a partner with a center climb.

Software: Making things as easy as possible on our drivers is the general goal. Eliminating manual aim while shooting with AprilTag detection will mean faster cycles.

Why base our design off of Unqualified Quokkas?
We had a similar concept on Monday of last week when we began discussing what designs would meet our goals while reducing the number of degrees of freedom necessary to accomplish all game tasks. We then saw the RI3D design and it checked off all of our boxes. We considered handoffs similar to Cranberry Alarm, with the addition of blockers to ensure that the Note went in the Amp, but

Why the bumper extension?
We wanted to protect the intake from damage while shooting and traversing the field. Any hits to the end effector in this design could damage the arm, especially side loads. Preliminary dimensional analysis showed that it was very hard to keep the intake within the bumpers while shooting from the subwoofer. When we saw the 95 method of mounting the intake outside of the swerve chassis yet within the bumpers, we saw a potential solution. This meant bringing the swerve down to a 25x25 square to allow for the 10 inches of bumper extension we needed to fit the end effector in its lower position. This gives us the advantage of a full-width intake that’s fully protected, especially relevant for center line autos.

This does come with the disadvantage of putting some weight outside of our swerve chassis, we’re going to try to keep most of the weight between our swerve modules and maybe add some casters under the bumper extension to prevent tipping during deceleration.

3 Likes

CAD

We’ve just finished our CAD! Aside from detail work, our robot is entirely planned out. We now have a climber arm on our robot CAD, which fits under the swing of the pivot. We’ve also added sponsor plates. We reworked the dimensions on the intake and shooter to make it so that the shooter has a width of 14".

Mechanical

Mechanical has been working away on the robot! The goal currently is to have the robot mechanically finished and wired fully two weeks from this post.

Programming

Programming has got a significant amount of work done! We now have our code sketched out for all our subsystems and have begun working on commands. We have tested pivoting our arm, pivoting our arm based on the distance to a tag, and centering on a tag on our protobot, which we’ve found to be a great resource.

Scouting

We are now using ScoutingPASS 2024 and are working on learning Tableau for data analysis.
We’ve also sketched out our pit and match scouting questions:

1 Like

ITS ALIVE!

This week the team got a lot done! The robot is now driving and we just need to wire in our end effector sensors and motor and we’ll be cooking! From there, the climbing arm needs to be on the robot as well.

Mechanical

The mechanical team has been working on our climbing mechanism, as our end effector is now mechanically done. We’re also just beginning on a pair of new bumpers.

Software

Software just got our pivoting arm working. Next up is setting it up on the bot with the LimeLight. We’ve also been working a lot on auto paths, with paths to do anywhere from 3 pieces off the center (preload + middle), to the 4 piece auto using the 3 close notes, and a 5 piece similar to 1678’s. These are yet to be tested.

We’ve also finished up customizing our scouting software, using only quantitative metrics for match scouting and a custom pit scouting set up.

Electrical

The electrical team has got a lot of work done. Our drive base is functional, and now we’re working on extending the data cables, power, and beambreaks from the end effector to the electrical board.

3 Likes

Week 0 Competition & Build Progress

Our team has made some great progress! We competed at the official Week 0 event and ended up being the first pick to the 3 seed alliance.

We managed to diagnose a lot of issues with our robot, and as a result are making a few changes, namely switching from Neos on our end effector to Kraken X60s. Our climber also broke after being bent during our first attempt to climb, so we’ll be reworking it to be more robust.

We are also going to begin working on a second swerve drivebase so newer members are kept busy.

Goals


Currently, we have a week long vacation, and the teams goal by next week is primarily to have the robot climbing again and shooting with our new Krakens. At Week 0 we only shot from flush with the subwoofer, or into the amp, so now we’re moving onto shooting from a variable range (up to the edge of the wing) using our LimeLight.

Here’s what our working climb looks like:
5459climb

We also currently have a functional one note and drive out auto. Now, we’re working on autos that begin on either side of the subwoofer and do both 3 and 4 pieces, where the 3-piece scores two middle notes.

This match highlights our auto well:

Other Reflections

During our Week 0 event, we had some issues with scouted data, primarily a lack of recorded data for some teams. To prepare better, we’ll be “pre-scouting” a Week 1/2 event before our Week 3 event. We’ll also begin collecting the statistic of the average time of full-field cycle teams for teams, as to have a better idea of their possible output.

We also noticed that when scoring in the amp or lining up flush with the subwoofer, the driver station we where at had a large impact on the difficulty of doing so and thus the time taken/wasted. In Elim’s and match strategy in general this will be important to keep track of.

Also, when running cycles we noticed that it was ideal to have two teams moving on the outside of the stage and one underneath. Fortunately, we can go underneath the stage if needed. This will probably play into our alliance selections in future events.

3 Likes