SAASquatch Robot Code for Reefscape

This project contains robot code for 8032's robot for FRC 2025 game Reefscape.

Software Features

Drive Train

• Custom swerve drive code for our modules, with kinematics and odometry
• All on-controller feedback loops to control precise speed and module angle
• Path following and pathfinding using PathplannerLib, can follow paths created in GUI
• High-frequency odometry (Inspired by team 6328)
• Full simulation using real swerve drive and module logic allowing for assurance of drive code and pathing logic

Driving

• Two controller scheme, drive and operator
• Strategic controller rumble for drivers to understand robot
• Elastic dashboard for key information
• Dashboard alerts for disconnected devices and other errors
• One button auto align for driver, automatically finds nearest scoring location

Superstructure

• State based control to automatically reach setpoints while avoiding collisions
• Detection for when Coral is held, triggering different states
• All on-controller feedback loops, enhanced with feed forward to counter gravity and friction
• All logic fully simulated, including Coral holding allowing for easy debugging and auto creation

Vision

• PhotonVision based full field localization, robot always new where it was
• Smart camera placement to view relevant April Tags
• Custom advanced filtering and standard deviations calculations to throw out bad results and trust good results more
• Kalman filter to integrate vision results with odometry data
• Allowed for smart auto alignment system and consistent autos when localization was used in pathing

Code structure

• AdvantageKit IO layers to separate control logic with hardware implantations, allowed for logic to be simulated on fake simulated hardware
• Full logging and real time monitoring on AdvantageScope to view robot state, whether simulated or real
• WPILib style command based programming with subsystems and command for clean reusable code

Diagram of code structure

Robot Abilities

• 4 MK4i Swerve Modules using NEO Brushless Motor for swerve drive
• Pigion2 gyroscope for heading data, allowing for field relative drive
• Controlled by two Xbox controllers with intuitive controls
• Two stage elevator with pivoting wrist able to reach all 4 scoring levels
• End effector with beam break sensor able to automatically obtain and release Coral
• Two OrangePi 5s with two low distortion, global shuttle Arducam OV9281s, allowing for localization and smart auto alignment
• Reliable deep hang using unique pivoting design

Dependencies

• WPILib
• AdvantageKit
• Pathplanner
• PhotonVision
• REVLib