Cartesian Robotics #8561 | Build Blog 2025

[ege_dgny]

CARTESIAN Build Blog | REEFSCAPE | 8561

WELCOME BACK!!!

#8561 Cartesian Robotics is excited to continue its Open Alliance tradition!

For the 2 upcoming days, we have 5 planned posts:

• Workshop Upgrade
• Our Off-Season Work and FTC Success
• The Kick-Off Event
• Season Callendar
• Season Goals and Strategy

After the planned posts, we plan on posting at least every 2 days!

Our History

Cartesian Robotics is more than a team. It is a community. 8561 is a student-found, school-based (METU DF High School) team in Ankara, Turkey. The community’s journey started in 2020 with the goal of establishing a strong foundation and opportunities for students who strive to develop in STEAM.

8561 first made its debut in FRC during the 2021 Infinite Recharge Season. In the 2022 season. We became the Istanbul Regional Winner and got awarded the Highest Rookie Seed (Ranked 2 out of 48). In the 2023 season, we faced some obstacles that taught us a lot.
2024 Season was where 8561 got seasoned. We have built 2 robots in 2024 season (you can check out our 2024 blog!). Improved our software and tested data-driven method. 8561 won the Creativity Award in Istanbul Regional. For the off-season We launched a summer program and built 2 off-season robots. The workshop got upgraded and new machines were introduced to the shop (More on the next post!)

With 65 members, our goal is to push our boundaries and learn much more in the 2025 season!

We will be competing in:

• Marmara Regional | 7-9 March (Week 2)
• Ankara Regional | 14-16 March (Week 3)

This year we will be the host team of Ankara Regional on week 3! We would love to get any advices on organizing a regional. Feel free to reach to me about it

Team Links

• Github - Cartesian Robotics #8561
• Onshape - You can follow our main design in real-time!
• Instagram - cartesian_8561
• TikTok - cartesianrobotics

We hope to reach as many of you as we can!
Feel free to ask any questions : )
(Our github is currently a bit messy, it will be cleaned out soon!)

[burakdemirelli]

As one of the few Open Alliance members from Turkey, I really enjoy seeing 8561’s growth, not just as a team, but also in the quality of your build blog. Wish you a wonderful season!

[ege_dgny]

Workshop Renovation

With major investments from our sponsors and new school administration we renovated what was a storage room, into Cartesian’s Branded workshop.

Strengthening our in-house manufacturing is our priority hence we welcomed a 1000mm by 700mm CNC router to deal with sheet metal and tube processing.

As an off-season project, one of our rookie members designed a manually operated bending machine to bend our 3mm aluminium sheets. Contraption is attached to our workbench and by using leverage, we can successfully do 90-degree bends. It is still in the process of manufacturing, we plan on adding a digital angle gauge. Here is the drawing:

For prototyping purposes, we decided a small form factor laser cutter is ideal for cutting plywood up to 5mm. Creality’s 40W laser cutter worked just fine for us.

The Design Studio

Another major upgrade is our newly obtained classroom. We have refurbished this unused classroom into our design studio! With 6 computers and a Smart Board, our members are now able to work on CAD, software and Digital Design without needing to leave the school, or coming into the workshop (which is not suitable to host 65 students at once).

Additionally, we have installed a second Smart Board to our workshop and linked it with the one in the Design Studio over the school network. Hence, the software and CAD team can broadcast their work to the workshop!

Lastly, we aim to enhance our semi in-house electronics. Over the off-season, the electronics department dedicated significant effort to R&D (more details in the next post!). As a result, we now have several PCBs ready for assembly for the upcoming season. To support this progress, the electronics department now got a designated workspace!

In the next post we will share our off-season work

[ege_dgny]

Thank you for your kind words Burak! 6838 inspired us to start this journey in 2023. Now we also have a keen eye on 10396’s work/blog this season as well! Wish you an amazing season too

[ege_dgny]

Off-Season Work and How It Will Tie Up to the 2025 Season

Our major development was on electronics and software side. Mechanics department worked on a 2-stage, cascading elevator.

The electronics department worked extensively on R&D on circuits and card models. D As the final product we designed 2 new PCBs: A Power Distribution Board (PDB) and a MIPI to HDMI adapter, inspired by Team 971, for Jetson Orin’s cameras.

We noticed that we were running out of space in our Power Distribution Hub so we decided that the most sustainable and logical course of action was to design our own PDB.

The MIPI to HDMI adapter is the result of needing long and durable cables for our OV9281 cameras. (More on them later) We will be using 4 cameras and 2 Jetson Orins for our field localization system.

We produced 8 of these boards for our 4 cameras as the cameras couldn’t be directly inserted into the HDMI ports. As a work around we decided to do the connection as MIPI-HDMI-MIPI, which required 8 boards.

One more small change, this off-season we tested out using a different CAN topology rather than a Daisy Chain. We will be running a Star Configuration for our drivetrain (8 Kraken x60, 4 CANcoder & CANivore)

You can access our Gerbers here: GitHub - team8561/8561_electronics

Software Development

Cartesian Robotics wasn’t able to off-season practices. This season, with our new workshop and participation in FTC we managed to achieve consistency with developing software. We couldn’t dedicate much time to FRC, however we managed to work on our Kraken chassis and vision. Throughout this efforts, we used AdvantageScop simulations and first tested out our programs in the simulation.

Before the season we worked on our FTC robot with a similar command-based code structure and won the Turkey Championship. This provided us with good practice for control and autonomy.
You can check out our FTC work here: Team 25153 Final Portfolio-compressed-1.pdf - Google Drive

The Vision!

We started with vision practices with AprilTag tracking. The new vision team learned to use the Limelight and investigated PhotonVision

apriltag tracking

We also got our hands on a MaixCam, which is a RISC-V powered AI camera. The MaixCAM allows us to quickly deploy object detection and tracking models. The Global Shutter camera it offers allows us to capture video at high-speeds and do inference even faster than our usual webcam. We can access it through a Python API and the SDK allows us to customize it from the ground up. The RISC-V vector instruction set allows the data-driven methods to run efficiently. You read more about it here: MaixCAM — Fast Deployment for AI Vision and Audio Projects - Sipeed Wiki

maixcam

Finally, this year we will keep focusing on our Jetson Nano applications more. Last year we were able to run our “Note Detection” on Jetson Nano and Google Coral combo. This year we have ordered 2 of the new Jetson Orin Nano Supers with 67 TOPS of computing power. When they arrive we want to test Team 971’s field localization implementation with Orins: (971-Robot-Code/frc971/orin at main · frc971/971-Robot-Code · GitHub ). We are currently using an OV9281 120-FOV, Monochromatic Global Shutter Camera, however we will switch to a Raspberry Pi Global Shutter Camera when our shipment arrives.

Current Test Setup

IMPORTANT: Our GitHub is now: team8561 (team8561) / Repositories · GitHub

[ege_dgny]

Kickoff and Initial Strategies (07/01/2025)

Kickoff

This year we have hosted our second exclusive kick-off event. Our aim was to gather many students and parents, and their friends to present the significance of Cartesian Robotics in our community. The event included mentor presentations, student-lead presentations and a demo for both of our FRC and FTC robots.

Strategy Overview

Here’s our breakdown of the strategies and their expected outcomes for this year’s game. Let’s dive in!

Auto RP (Ranking Point):

All robots in the alliance score 1 coral in Auto

Probably will be achieved in a large number of matches based on who we are playing with.

The safest bet is to play an L4 auto, cycling in between the Human Player and Reef.

Top Priority: Coral Scoring

• Our primary focus is on maximizing coral scoring efficiency to gain a competitive edge. Here are some critical observations and challenges:
• Coral placement on L4 is key to achieving higher scores and maximizing RP potential. => Top-Down Approach.
• Scoring on L4 will also grant a hypothetical 4-piece auto, without clearing the algae.
• Driver’s have obstructed visibility on the backside of the reef makes aligning and scoring on certain branches very challenging.
• Aligning corals to visible reef branches is not easy => Automated guidance systems!
• Scoring on L4 branches near algae patches may require additional clearance, which could introduce inefficiencies => Simple and weight efficient algae clearing system.
• The design must prioritize fast, repeatable scoring cycles for both high and mid-level coral placement.
• Ground and HP pickups can be strategically balanced due to the 2 Human Player stations, accommodating multiple robot designs in match strategies.

Algae clearing:

Clearing algae will make it much easier to place the necessary amounts of coral on the reef for a RP therefore it is another high priority goal.

Scoring algae:

We deemed it the most logical course of action to score the algae into the processor, as a full proof and consistent design for scoring algae directly to the net seemed implausible and time inefficient. Additionally we deemed it a must to put 2 algae in the processor for the coopertition point and to make the RP from corals easier.

Climb:

A deep climb is a must as it will guarantee a RP as long as one of our alliance members parks.

Overall Superstructure:

The game demands a tall robot for the L3 and L4. The priority should be on a stable and an accurate elevator system. To avoid tall robot syndrome, the elevator’s payload should be small to keep our CG low during fast cycles.

Due to a delay on receiving our KOP, we have focused on building our field and strategizing, while gathering information from the work of RI3D!

[ege_dgny]

Diving into BlockCAD

While waiting on our KOP, Ri3D was the perfect source of data on what to do, how to do and how pieces behave. Hence we had our inspirations and brainstormed on several robot designs.

The Coral

Our main focus was on the coral scoring mechanism. We sticked to prioritizing a HP intake hence we think HP cycles will be the most critical ones and executing HP cycles with accuracy and speed will determine the course of the game, similar to the 2019 and 2023 game.

We first explored and sketched Alumni of Madison’s RI3D design and implemented it on a double jointed arm. We persisted on a double joint since we had experience programming 2 DOF arms from 2023 season and our FTC robot.

However, as the saying goes; “Every DOF saved is a DOF earned”. We wanted to minimize the weight we will displace during our scoring cycle. Furthermore, according to our observation, a design that will reduce the rotation of the drivetrain will further speed up the cycles. To address all these issues, we came up with a design that will allow us to intake from one side and outtake from the opposite, with little to no motion. Our design consists of a funnel that can rotate from -35 to 35 degrees with respect to ground plane. Leaving all the job to gravity will cost us time and accuracy so we decided to add a conveyor belt to the bottom of the funnel.

Additionally, a V shaped funnel may also introduce a 3rd intaking position, where we can prop up against the HP, and collect the Corals in 90 degrees, however, this introduces a possibility where the coral won’t align with the funnel. Today we will experiment how that proceeds. We will update you soon :))

The Cage

From the second we were introduced to the deep climb, our intuition told us to take the cage to the center of the robot. This will grant a leveled climb if we align the center of gravity with the cage. Additionally, we can use our elevator to lift us up at the end game. Only drawback is the space it costs us.

Further seeing this idea implemented by Penn State RI3D, reinforced our intuitive solution. We will keep exploring this climb style throughout this week.

We settled on building a 3-stage elevator, we will start its manufacturing today. We will be using 2 Kraken X60s to power it, use #-25 chain to drive the first stage and to negate the weight we will be using SH16P40 constant force springs from Vulcan with 47N (10.6 lbf) load. The max extension will be 190cm (78.4 in).

You can follow our CAD as we proceed, we share it in real-time : Onshape

We would love to hear your comments!

[ege_dgny]

Vauge Coral Carrier Test

As we mentioned in the previous post, we are working on a funnel that we think, will allow us to intake from the HP, from 3 sides of the robot. This will quicken our cycles in between the HP and the reef by only requiring an axial motion of our funnel.
To make our system more consistent and reliable, we think of adding a conveyor belt beneath our funnel that will push the coral to the reef, when necessary.

Hence, we did this vauge prototype, where we wanted to see how the corals interact with certain belts/polycords. It seems we recorded only one of our trials, but we will keep you updated as we move on to better executed demos.

Regular HTD Belt:

coral_conveyor

Our first impressions were positive about using a belt to transfer corals on our funnel.

[nuclearnerd]

@ege_dgny

Further seeing this idea implemented by Penn State RI3D, reinforced our intuitive solution. We will keep exploring this climb style throughout this week.

You may want to bias that “cage pocket” a little to the top or bottom of the above picture, or else make sure the robot CG is to the right of the cage in the picture above. That way you can make use of the lexan sides of the pocket to push against the cage and keep you straight. If the CG ends up to the right of the cage, you’re likely to flip to the left as the robot pushes the cage down.

[ege_dgny]

Great point! Its an important factor and as you said it will probably tilt the robot and may even cause the cage to slip. We will definitely implement your suggestion. We also think, rather than using the elevator, we can use 2 pistons to push against the cage, and with a locking mechanism mount the robot on the cage.

[ege_dgny]

First Iterations and Production

During the past 2 days we were busy with our CAD. We have finalized our first iteration for the Coral Scoring Mechanism, which we call the “Funnel”

The conveyor consists of a polyurethane belting and a 2mm Carbon Fiber plate. Conveyor is powered by 2 Neo 550 attached to a 12:1 Rev Ultraplanetary. Neo 550s are encapsulated in the conveyor wheels to save space while maintaining a robust mounting.

Here are some demos of this assembly:

neo_550_assembly

neo_550_test

The Funnel is attached to the 3-stage elevator via a 3:1 gearbox, powered by a Kraken x60. We have used a bevel gear within the gearbox to package everything tightly.

Our OnShape Link: Onshape

Manufacturing the Elevator Bearing Blocks

The CAM team and our mentors Cihan and Mert setup the router to produce the 12 bearing blocks for our 3-stage elevator. After the production of the bearing blocks are done, we will be moving on to process our tubings.

elevator_bracket_production

Looking forward for your replies!