NVIDIA-ISAAC-ROS / NVIDIA-ISAAC-ROS/isaac_ros_manipulation

cumotion_nvblox_pose_to_pose.launch.py doesn't launch planner node, ESDF not generated, and robot segmentor not masking correctly with FR3 + Realsense

Open
#12 0 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Dominant language
C++
Stars
114
Forks
25
PR merge metrics
No merged PRs in 30d

Description

Question

Summary

When launching cumotion_nvblox_pose_to_pose.launch.py using FR3 and a single Realsense camera, I encounter the following issues:

  1. The cumotion_planner node does not launch.
  2. ESDF is not generated by NVBlox by default.
  3. Robot segmentor does not mask the robot – in the depth_seg_image, the robot is still visible (not highlighted in white as expected).

🧪 Setup
  • Robot: FR3 (Franka Emika Panda)
  • Camera: 1x Intel Realsense
  • Launch file used: cumotion_nvblox_pose_to_pose.launch.py

❗ Observed Behavior
  1. cumotion_planner node inactive

    • The node doesn’t launch when using the launch file mentioned above.
  2. ESDF not auto-generated

    • No ESDF topics like /nvblox_node/esdf_map/out are published.
    • However, when I manually call the following service:
      ros2 service call /nvblox_node/get_esdf_and_gradient nvblox_msgs/srv/EsdfAndGradients "update_esdf: true
      visualize_esdf: true
      use_aabb: true
      frame_id: 'base_link'
      aabb_min_m:
        x: -1.0
        y: -1.0
        z: -0.5
      aabb_size_m:
        x: 2.0
        y: 2.0
        z: 2.0
      spheres_to_clear_radius_m: []"
      
      NVBlox successfully returns voxel data for the requested bounding box. So voxel computation appears to be working, but not triggered automatically.
  3. Robot not masked by robot segmentor

    • In the output depth segmentation image, the robot is not masked (not in white).
    • I believe masking should highlight the robot in white if working correctly.

Image Image

After esdf service call the voxels come in rviz

Image
✅ Expected Behavior
  • cumotion_planner node should launch automatically via the launch file.
  • NVBlox should continuously generate and publish ESDF layers.
  • Robot should be masked (white) in depth segmentation images by the robot segmentor.

🔍 Questions
  • Is there a config flag I need to enable (e.g. esdf_enabled: true or similar)?
  • Does NVBlox require certain TFs or topics before publishing ESDF automatically?
  • How should the robot segmentor be configured to correctly identify and mask the robot?

📚 Request for Help

Would appreciate if you could share:

  • Example working configurations for:
    • cumotion planner integration
    • ESDF generation via NVBlox
    • Robot segmentor masking
  • Any troubleshooting references for launch behavior and masking issues

🧾 Additional Context

Happy to provide the following if needed:

  • Full launch output logs
  • RViz visualizations
  • TF tree structure and URDF setup

Thanks for your help!

Contributor guide

Open the contributing guide

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Start with cumotion_nvblox_pose_to_pose.launch.py and reproduce the FR3 plus single Realsense setup. Check the cumotion_planner state, the /nvblox_node/esdf_map/out topics before and after the documented get_esdf_and_gradient service call, and the depth_seg_image output. Done means the planner launches, ESDF is published continuously, and the robot is masked in the segmentation image.

Written by the indexing model from the issue text.

Assessment

Tech stack
python
Domain
computer-vision, robotics
Issue type
Bug
Difficulty
4/5
Estimated time
3-5 days
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
25/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.