SmartDashboard data doesn't save to log file

[AniKashyap]

We (1678) are currently competing at Madtown Throwdown. We have encountered a very weird issue during this event.

From what we can see, there seems to be an issue with logging data. Specifically, data relating to the positions and setpoints of swerve modules.

Here is a timeline of what happened:

Quals 3: No issue with data, stuff logged as expected

System Check before Quals 11: During our checks no issues, robot functioned as expected. When looking at the logs post-match we discovered the issue was present here as well.

Quals 11: For the most part, robot functioned as expected during the match. We pulled the log to look at a separate issue.

When we tried to look at the log from Quals 11, this is we discovered the issue.

The robot(fused) pose and the vision pose (from the limelight) stay in the default position the entire log. Other things that are related like the module positions and states also are nonexistent.

I ran additional tests on the robot before turning power off. These values update when connected to the robot and viewing data live via AdvantageScope. When I went and pulled this log (new section of Q11 log) the same data that was logged live was nonexistent in the log file.

Seems like all these fields are not getting initialized.

We swapped battery and power cycled and ran these two tests again. Same issue, the data is visible live but nonexistent in the log.

Our code for reference: https://github.com/frc1678/C2025-SystemCore

Storage also is nowhere near full, Internal is ~35% full and we are putting all logs on a thumb drive which is ~20% full

@corsetto pinging for notifications

[rzblue]

Looking inside your Telemetry class, there are multiple topics being published to SmartDashboard/Drive/ModuleStates
An array of SwerveModuleState structs is published here:
https://github.com/frc1678/C2025-SystemCore/blob/alpha2/src/main/java/frc/robot/Telemetry.java#L40-L42
and updated here:
https://github.com/frc1678/C2025-SystemCore/blob/alpha2/src/main/java/frc/robot/Telemetry.java#L89

The same topic name is used to publish the data, now as a double array, here:
https://github.com/frc1678/C2025-SystemCore/blob/alpha2/src/main/java/frc/robot/Telemetry.java#L102
Likewise for SmartDashboard/Drive/ModuleTargets.

This should have a fairly defined behavior- the second publish (in this case, this should be the double array) should be ignored, and value updates will not have any effect. There will also be a message printed to stdout/DS console:

NT: local publish to '/SmartDashboard/Drive/ModuleStates' disabled due to type mismatch (wanted 'double[]', currently 'struct:SwerveModuleState[]')
NT: local publish to '/SmartDashboard/Drive/ModuleTargets' disabled due to type mismatch (wanted 'double[]', currently 'struct:SwerveModuleState[]')

Could you upload a "good" log (data present) and a "bad" log (data missing), as well as the driver station log for the corresponding matches?

[AniKashyap]

Sure!

Quals 3 Wpilog: https://drive.google.com/file/d/1y8RrkLtbBjVIVFSbZf8OdHH3HTTmmTk2/view?usp=sharing
Quals 3 DSLog: https://drive.google.com/file/d/1I7WYLsn5BTKn8vBbFoSXLVQTnv6l1jkj/view?usp=sharing

Quals 11 Wpilog: https://drive.google.com/file/d/1u-zJ0i3FYepNQ9nZQhKHhUUbBaOgFQDP/view?usp=sharing
Quals 11 DSLog: https://drive.google.com/file/d/1rNSveNeg7pbZX2WEACZXq3RxKtv9Wp6P/view?usp=sharing

[AniKashyap]

Another potentially related thing that we saw in our data for quals 38 is that subsystem data appears in the middle of the match. See pictures below.

First data from the our Algae Deploy (Algae Intake) mechanism This same timestamp in the match is about 75% through it. The mechanism had been running before this point in the match, and at least we should have seen a flat line if it didn't run before this.

Here is the full log as well: https://drive.google.com/file/d/1uXMZKkqJ48ClrT4zKjtdrZHVCSLMca3z/view?usp=sharing

[TheTripleV]

Can you share

• all .wpilog files
• all .dslog files
• all .dsevent files

from the event?

[AniKashyap]

Yeah sure! I have timelined the entire event as much as I can as well if that helps.

I put everything I have/was able to get here: https://drive.google.com/drive/folders/16QoQwmHol0-gCWo-8Bkf2P2-EVjzpmvU

Prac Day: Dont have any logs from this currently, will try to get
Deployed once: frc1678/C2025-SystemCore@fd88abc
Played more practice matches after this and everything was fine

Quals 3: No issues to worry about

Quals 11: Swerve Log issue and Mechanism Data issue were present here

Quals 23: Need to grab wpilog

Quals 31: Need to grab wpilog

Quals 38: Swerve Log issue and Mechanism Data issue were present here

Quals 47: Swerve Log issue and Mechanism Data issue were present here

Quals 54: Swerve Log issue and Mechanism Data issue were present here

Deployed, commit: frc1678/C2025-SystemCore@8c98b11

Quals 67: No issues to worry about

Elims 3: Swerve Log issue and Mechanism Data issue were present here

Deployed, commit: frc1678/C2025-SystemCore@cdb9685

Elims 8: Need to grab wpilog

Elims 10: Swerve Log issue and Mechanism Data issue were present here

I investigated the logs further and found something interesting. Our "Elastic Field 2D" logger seems to be updating for the most part. Its value is derived in robot code from the drive pose. Look at Drive.java Lines 72-73.