our test code is at: https://github.com/JustinShih0918/ros2-essentials/tree/feat/kobuki-slam
1. compose up kobuki-ws with rtabmap tag 'YES'
2. bringup kobuki to get /odom

./script/kobuki-bringup.sh

3. open vlp-ws and bringup lidar

ros2 launch vlp_cartographer vlp_driver.launch.py 

4. open realsense within templete-ws
feature from https://github.com/yuhsiang1117/ros2-essentials/tree/feat/realsense

ros2 launch realsense2_camera rs_launch.py align_depth.enable:=true pointcloud.enable:=true

5. back to kobuki-ws, launch rtabmap with

ros2 launch kobuki_navigation real_rtabmap.launch.py

the launch file will first:

- read the config yaml

rtabmap: 
  ros__parameters: 
    frame_id: 'base_footprint'
    use_sim_time: False
    use_action_for_goal: True
    qos_image: 2
    qos_imu: 2
    Reg/Force3DoF: 'true'
    Optimizer/GravitySigma: '0'

    # For LiDAR data subscription
    subscribe_scan: True

    # For RGB-D data subscription
    subscribe_depth: True
    subscribe_rgb: True

    # Data synchronization
    approx_sync: False
    topic_queue_size: 40
    sync_queue_size: 40

- remap topic for realsense

# Setup the remaps for camera topics
    remappings = [
        ('rgb/image', '/camera/color/image_raw'),
        ('rgb/camera_info', '/camera/color/camera_info'),
        ('depth/image', '/camera/aligned_depth_to_color/image_raw')
    ]

- open rtabmap node and rviz

    rtabmap_slam = Node(
        package = 'rtabmap_slam', 
        executable = 'rtabmap',
        parameters = [rtabmap_settings],
        remappings = remappings,
        arguments = ['-d', '--delete_db_on_start']
    )
    
    rtabmapviz = Node(
        package = 'rtabmap_viz',
        executable = 'rtabmap_viz',
        parameters = [rtabmap_settings],
        remappings = remappings
    )

### we have tried
1. approx_sync: False , queue_size = 40

seems that clock time is not synchronizes

[rtabmap-1] [WARN] [1764427540.449959784] [rtabmap]: rtabmap: Did not receive data since 5 seconds! Make sure the input topics are published ("$ ros2 topic hz my_topic") and the timestamps in their header are set. If topics are coming from different computers, make sure the clocks of the computers are synchronized ("ntpdate"). Ajusting topic_queue_size (40) and sync_queue_size (40) can also help for better synchronization if framerates and/or delays are different. Parameter "approx_sync" is false, which means that input topics should have all the exact timestamp for the callback to be called.
[rtabmap-1] rtabmap subscribed to (exact sync):
[rtabmap-1]    /odom \
[rtabmap-1]    /camera/color/image_raw \
[rtabmap-1]    /camera/aligned_depth_to_color/image_raw \
[rtabmap-1]    /camera/color/camera_info \
[rtabmap-1]    /scan

2. approx_sync: True
give us the same warning.

3. check hz of each topic
- realsense
- color/image_raw
- 6.239 hz
- aligned_depth_to_cloud
- 5.319 hz
- after closing the background training model, we receive realsense topic hz at about 30hz
- /odom
- 49.980 hz
- /scan
- 39.160

4. open lidar in rtabmap alone, everything is good, the problem may cause by realsense.