Implement error tracking

[ErikCald]

Part of the api of CANSparkMax let's us ask it for any fault codes it is experiencing. It would awesome to periodically ask the sparks for any fault codes and log them. It would also be incredibly helpful to present any faults and which device is causing them to the pit crew. That would greatly reduce the chance of something going wrong during a match because it can be found in the pits and solved.

For example, they could see that SparkMax number 4, the arm sparkmax, is reporting a sensor fault which means the encoder is unplugged.

The best tool to display stuff to pit crew would probably be [shuffleboard], where you can store a layout of networktable values to show. (https://docs.wpilib.org/en/stable/docs/software/dashboards/shuffleboard/index.html).

This idea can also extend to other devices like the CANCoder, PCM, PDP, and probably others.

[ErikCald]

There are multiple ways to go about doing this, here is one potential way:

• Create a subsystem for error tracking
• Create a method to allow other subsystems to register their CANSparkMax objects with this subsystem
• Use the periodic() method of the subsystem to periodically call functions on the sparkmax's to check errors.
• Log when a error occurs (probably StringLogEntry)

Other related ideas:

• Probably best to limit the amount of calls that check for errors because it can crowd the can bus. The ideal setup would probably be to call only 1 sparkmax every robot cycle, maybe even calling 1 sparkmax every 2 robot cycles. Aka, ask drive of module 1 on one robot cycle, then ask steer of module 1 on the next cycle, then drive of module 2, etc.
• Possible to use sticky faults and then clearing sticky faults instead of just asking the sparkmax for it's faults because we might miss a fault if we are limiting the amount of calls.

[ErikCald]

Options for interface: Shuffle board or web dev with Frc Web Components (https://github.com/frc-web-components/frc-web-components)

[ErikCald]

I think this works FaultID.fromId(m_driveMotor.getFaults());

This is from last year

// Add CommandScheduler to shuffleboard so we can display what commands are scheduled
    ShuffleboardTab basicDebuggingTab = Shuffleboard.getTab("BasicDebugging");
    basicDebuggingTab
      .add("CommandScheduler", CommandScheduler.getInstance())
      .withPosition(3, 0)
      .withSize(3, 6);

https://docs.wpilib.org/en/stable/docs/software/dashboards/shuffleboard/layouts-with-code/index.html

[ErikCald]

Oh apparently it's a bit more complicated than asking the spark for faults. I found this code that is checking each bit in the short from getFaults()

https://github.com/frcmi/Robot2023/blob/ce8f851f4cd1489022d8e99a2f93ca8b0c925ca8/src/main/java/frc/robot/SparkMaxUtils.java#L73