Skip to content

Conversation

@jaybdub
Copy link
Collaborator

@jaybdub jaybdub commented Apr 10, 2025

  • Adds additional visualization during usage, to display path for path following scenario.
  • Adds performance optimizations related to get_world_pose(). Increases FPS from 20->60FPS for all scenarios
  • Adds performance optimizations to path following scenario. Increases FPS from 15->55FPS for this scenario
  • Adds support for loading pre-generated Occupancy Map
  • Adds instructions for generating occupancy manually using existing Isaac Sim tool to README

Testing

  1. Tested by running through full workflow (data collection, replay rendering) with H1 robot for path following and keyboard teleop scenarios
  2. Ensured that old logs still work with replay-rendering

@jaybdub jaybdub requested a review from yanchangNvidia April 10, 2025 20:12
Copy link
Collaborator

@yanchangNvidia yanchangNvidia left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Overall it looks great to me. Left some comments.

```

### Step 2 - Build a scenario
### Step 2 - Load a stage
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

"scene" seems more commonly used than "stage"?

@@ -0,0 +1,111 @@
import argparse
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Why there is a "05" in the file title?

Copy link
Collaborator Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This is given because this script is currently listed as an example (number 5). The purpose of the numbering is to provide developers with a rough progression to follow. I find this pattern helpful when I see it in other projects, like NVISII.

Maybe we should instead place this under "scripts", which is unordered, and meant more for direct usage, not modification

@yanchangNvidia yanchangNvidia requested a review from nv-liuw April 11, 2025 22:00

if dist_to_target < self.robot.path_following_stop_distance_threshold:
self.set_random_target_path()
self.is_alive = False
Copy link

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This might cause a new data folder will be created every time when the robot reaches a target, and the recording segment will be very short.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

4 participants