2027 Robot — FRC Team 1405

FRC robot code for the 2027 season. Built on WPILib 2026.2.1 with CTRE Phoenix6, PathPlanner, and PhotonVision.

Project Structure

src/main/java/frc/robot/
├── Robot.java               # Main robot class (extends LoggedRobot)
├── RobotContainer.java      # Subsystem instantiation and button bindings
├── Constants.java           # Tuning constants (PID gains, CAN IDs, etc.)
├── Telemetry.java           # Swerve drivetrain telemetry
├── commands/                # WPILib commands
├── subsystems/              # Robot subsystems (Shooter, Intake, Indexer, …)
│   └── vision/              # PhotonVision integration
├── constants/               # Feature switches, field layout, robot constants
├── generated/               # CTRE Tuner X generated swerve constants
├── lib/                     # Shared utilities (logging, simulation helpers, …)
└── sim/                     # Physics simulation profiles

Building & Deploying

# Build
./gradlew build

# Deploy to roboRIO
./gradlew deploy

# Simulate (desktop)
./gradlew simulateJava

Feature Switches

constants/FeatureSwitches.java contains boolean flags to enable or disable features at compile time (e.g. vision filters, debug logging, mechanical protections). All flags default to the safe/2026-baseline state.

---

AdvantageKit

AdvantageKit (v26.0.2) is integrated for structured telemetry logging and deterministic log replay.

What it does

• Logs all @AutoLogOutput fields from every subsystem automatically each robot cycle, publishing them to NetworkTables and writing them to a .wpilog file on the roboRIO.
• Enables log replay — you can re-run robot code in simulation against a real match log, deterministically reproducing every sensor input and state change.

Viewing logs

Open the .wpilog file created in /home/lvuser/logs/ on the roboRIO with AdvantageScope. During a match or simulation you can also connect AdvantageScope to the robot's live NetworkTables stream.

Logging a new field

Add @AutoLogOutput to any public or package-private instance field in a SubsystemBase subclass. Supported types include boolean, int, double, String, double[], Pose2d, SwerveModuleState[], and other WPILib structs.

import org.littletonrobotics.junction.AutoLogOutput;

public class MySubsystem extends SubsystemBase {

    @AutoLogOutput(key = "MySubsystem/IsActive")
    private boolean isActive = false;

    @AutoLogOutput(key = "MySubsystem/TargetRPS")
    private double targetRPS = 0.0;
}

The key is the path shown in AdvantageScope/NetworkTables under the AdvantageKit root table.

You can also log one-off values anywhere in code with:

Logger.recordOutput("MySubsystem/SomeValue", myValue);

Current annotated fields

Class	Key	Type
Shooter	Shooter/TargetRPS	double
Shooter	Shooter/Locked	boolean
Shooter	Shooter/WasLocked	boolean
Shooter	Shooter/ShotCount	int
Indexer	Indexer/Active	boolean

Replaying a log (advanced)

Note: Log replay is not wired up by default — simulation currently runs the physics sim with live NT telemetry only. To enable replay:

1. Transfer a .wpilog file from the roboRIO to your PC.
2. Open the log in AdvantageScope and use File → Export replay log to set it as the replay source, or edit Robot.java's simulation block:

} else {
    setUseTiming(false); // run as fast as possible during replay
    String logPath = LogFileUtil.findReplayLog();
    Logger.setReplaySource(new WPILOGReader(logPath));
    Logger.addDataReceiver(
        new WPILOGWriter(LogFileUtil.addPathSuffix(logPath, "_sim")));
}

3. Run ./gradlew simulateJava — the robot code will re-execute against the recorded inputs and write a new _sim log with all outputs.

You can also watch for log file changes with the Gradle task:

./gradlew replayWatch

Notes on logging configuration

On the real robot, AdvantageKit writes to /home/lvuser/logs/ (same directory as WPILib's DataLogManager). Both systems run in parallel — AKit writes its own .wpilog file while DataLogManager writes a separate one for CTRE Hoot compatibility. This is intentional; you can remove the DataLogManager calls in Robot.java once you no longer need CTRE Hoot replay.

---

Vision

The robot uses PhotonVision cameras for AprilTag-based pose estimation. Vision filtering parameters (boundary rejection, stddev tuning, tag rankings) are controlled via feature switches in FeatureSwitches.java. See docs/vision-testing-protocol.md for the A/B testing methodology.

Coprocessor (Orange Pi) scripts — metrics publishing and raw-frame vision recording — live in coprocessor/; see coprocessor/README.md.