Betatesting report on all 8 exercises mentioned in issues

[Kunal-Somani]

Hi everyone,
@jmplaza @javizqh @dpascualhe

Completed betatesting of all 8 exercises from the current betatesting list on the latest RADI with NVIDIA GPU enabled on Ubuntu 22.04. All exercises launched successfully, simulations ran at full speed, and no crashes or startup errors were observed.

Exercises tested:

• Car Junction ([Betatesting] Car junction exercise #3456)
• Labyrinth Escape ([Betatesting] Labyrinth escape exercise #3462)
• Power Tower Inspection ([Betatesting] Power tower inspection exercise #3463)
• Drone Gymkhana ([Betatesting] Drone Gymkhana exercise #3464)
• Object Detection ([Betatesting] Object Detection exercise #3465)
• Image Classification ([Betatesting] Image Classification exercise #3466)
• Follow Road ([Betatesting] Follow road exercise #3467)
• Autoparking using 3D lidar ([Betatesting] Autoparking exercise using the 3d lidar #3468)

One thing worth flagging: Image Classification requires torchvision to be installed separately inside the container. It installs cleanly in a few seconds but might catch first-time users off guard.
Should I continue working on the implementation side of my submitted proposal, or are there specific open tasks from or other than the betatesting issues that would be more useful to pick up right now?

[jmplaza]

Hi everyone, @jmplaza @javizqh @dpascualhe

Completed betatesting of all 8 exercises from the current betatesting list on the latest RADI with NVIDIA GPU enabled on Ubuntu 22.04. All exercises launched successfully, simulations ran at full speed, and no crashes or startup errors were observed.

Exercises tested:

• Car Junction ([Betatesting] Car junction exercise #3456)
• Labyrinth Escape ([Betatesting] Labyrinth escape exercise #3462)
• Power Tower Inspection ([Betatesting] Power tower inspection exercise #3463)
• Drone Gymkhana ([Betatesting] Drone Gymkhana exercise #3464)
• Object Detection ([Betatesting] Object Detection exercise #3465)
• Image Classification ([Betatesting] Image Classification exercise #3466)
• Follow Road ([Betatesting] Follow road exercise #3467)
• Autoparking using 3D lidar ([Betatesting] Autoparking exercise using the 3d lidar #3468)

One thing worth flagging: Image Classification requires torchvision to be installed separately inside the container. It installs cleanly in a few seconds but might catch first-time users off guard. Should I continue working on the implementation side of my submitted proposal, or are there specific open tasks from or other than the betatesting issues that would be more useful to pick up right now?

good betatesting @Kunal-Somani, well done. Create an issue for the torchvision dependency, the RoboticsAcademy user should not be required to install any further dependency in the RADI, all the dependencies should be pre-installed there.

No contribution on the submitted applications is desired now, the required tasks on the proposed projects in the published ideas list will be done by the selected candidates, if any.

[Kunal-Somani]

Thank you @jmplaza! I'll open an issue for the torchvision dependency right away.

[Kunal-Somani]

@jmplaza
Issue opened: #3741