I have a simulated camera in Gazebo and want to use the point cloud of the camera for MoveIt2s Octomap. The points are getting published on the topic Canon_Powershot_G7_MKIII/points with the reference frame being Canon_Powershot_G7_MKIII_virtual_frame_link.

Example message:

header:
  stamp:
    sec: 632
    nanosec: 175000000
  frame_id: Canon_Powershot_G7_MKIII_virtual_frame_link
height: 1
width: 307200
fields:
- name: x
  offset: 0
  datatype: 7
  count: 1
- name: y
  offset: 4
  datatype: 7
  count: 1
- name: z
  offset: 8
  datatype: 7
  count: 1
- name: rgb
  offset: 16
  datatype: 7
  count: 1
is_bigendian: false
point_step: 32
row_step: 9830400
data:
- 82
- '...'
is_dense: true

The TF tree is also set properly.

The octomap_updater_config of the move_group_node is:

# Camera setup
    camera_configuration = load_yaml(
        "modproft_camera_description", "config/canon_powershot_g7_mk3.yaml")

    if prefix:
        # camera_name = ''.join([prefix, str(camera_configuration["camera"]["sensor"]["name"])])
        camera_name = str(camera_configuration["camera"]["sensor"]["name"])
    else:
        camera_name = str(camera_configuration["camera"]["sensor"]["name"])

    octomap_updater_config = {
        "sensors": [camera_name],
        camera_name: {
            "sensor_plugin": "occupancy_map_monitor/PointCloudOctomapUpdater",
            "point_cloud_topic": "/{}".format('/'.join([camera_name, "points"])),
            "max_range": 5.0,
            "point_subsample": 1,
            "padding_offset": 0.1,
            "padding_scale": 1.0,
            "max_update_rate": 60.0,
            "filtered_cloud_topic": "/{}".format('/'.join([camera_name, "filtered_cloud"]))
        }
    }
camera_link_virtual = "{}{}".format(camera_name, '_virtual_frame_link')

    octomap_config = {"octomap_frame": camera_link_virtual,
                      "octomap_resolution": 0.05,
                      "max_range": 5.0}

    # Start the actual move_group node/action server
    move_group_node = Node(
        package="moveit_ros_move_group",
        executable="move_group",
        output="screen",
        parameters=[
            robot_description,
            robot_description_semantic,
            robot_description_kinematics,
            ompl_planning_pipeline_config,
            trajectory_execution,
            moveit_controllers,
            planning_scene_monitor_parameters,
            octomap_config,
            octomap_updater_config,
            {"use_sim_time": use_sim_time,
             "verbose": False}
        ],
        arguments=['--ros-args', '--log-level', 'debug']
    )

tf_monitor also yields a result:

RESULTS: for world to Canon_Powershot_G7_MKIII_virtual_frame_link
Chain is: world -> robot_table_link -> robot_table_tabletop_link -> base_link -> base_link_inertia -> shoulder_link -> upper_arm_link -> forearm_link -> wrist_1_link -> wrist_2_link -> wrist_3_link -> flange -> tool0 -> Canon_Powershot_G7_MKIII_link -> Canon_Powershot_G7_MKIII_virtual_frame_link
Net delay     avg = 0.00105632: max = 0.00300813

Frames:
Frame: Canon_Powershot_G7_MKIII_link, published by <no authority available>, Average Delay: 1.68924e+09, Max Delay: 1.68924e+09
Frame: Canon_Powershot_G7_MKIII_virtual_frame_link, published by <no authority available>, Average Delay: 1.68924e+09, Max Delay: 1.68924e+09
Frame: base_link, published by <no authority available>, Average Delay: 1.68924e+09, Max Delay: 1.68924e+09
Frame: base_link_inertia, published by <no authority available>, Average Delay: 1.68924e+09, Max Delay: 1.68924e+09
Frame: flange, published by <no authority available>, Average Delay: 1.68924e+09, Max Delay: 1.68924e+09
Frame: forearm_link, published by <no authority available>, Average Delay: 0.000539706, Max Delay: 0.0022819
Frame: robot_table_link, published by <no authority available>, Average Delay: 1.68924e+09, Max Delay: 1.68924e+09
Frame: robot_table_tabletop_link, published by <no authority available>, Average Delay: 1.68924e+09, Max Delay: 1.68924e+09
Frame: shoulder_link, published by <no authority available>, Average Delay: 0.000544725, Max Delay: 0.00228477
Frame: tool0, published by <no authority available>, Average Delay: 1.68924e+09, Max Delay: 1.68924e+09
Frame: upper_arm_link, published by <no authority available>, Average Delay: 0.000542384, Max Delay: 0.00228357
Frame: wrist_1_link, published by <no authority available>, Average Delay: 0.000546804, Max Delay: 0.00228572
Frame: wrist_2_link, published by <no authority available>, Average Delay: 0.000548919, Max Delay: 0.00228667
Frame: wrist_3_link, published by <no authority available>, Average Delay: 0.000550851, Max Delay: 0.00228739

All Broadcasters:
Node: <no authority available> 581.328 Hz, Average Delay: 2.43251e+08 Max Delay: 1.68924e+09

I would therefore assume that tf2_ros_message_filter would have all information to transform the point cloud so octomap could use it. This is not the case, when running the code get:

move_group-5] [DEBUG] [1689256158.441029452] [tf2_ros_message_filter]: MessageFilter [target=world ]: Added message in frame Canon_Powershot_G7_MKIII_virtual_frame_link at time 13.425, count now 1
[move_group-5] [DEBUG] [1689256158.441049945] [rcl]: Subscription in wait set is ready
[move_group-5] [DEBUG] [1689256158.441051788] [rcl]: Subscription in wait set is ready
[move_group-5] [DEBUG] [1689256158.441060534] [tf2_ros_message_filter]: MessageFilter [target=world ]: Discarding message in frame Canon_Powershot_G7_MKIII_virtual_frame_link at time 13.425, count now -1
[move_group-5] [DEBUG] [1689256158.441063439] [rcl]: Subscription taking message
[move_group-5] [DEBUG] [1689256158.441058909] [rcl]: Subscription taking message
[move_group-5] [INFO] [1689256158.441064612] [move_group]: Message Filter dropping message: frame 'Canon_Powershot_G7_MKIII_virtual_frame_link' at time 13.425 for reason 'Unknown'

I'm not getting the cause of that issue when looking into message_filter.h.

Are there better ways to debug these non-descriptive error messages?

Could be the timestamps the problem? - TF doesn't give a clue here, but since it isn't complaining about it I would assume that the timestamps are allright.