Links

Kudan Integration

This guide shows you how to integrate the Kudan packages into Seirios RNS.
Kudan Inc is a Japan-headquartered company focusing on artificial perception computer software by providing artificial perception solutions, such as localization and mapping software.
Kudan offers commercial-grade localization and mapping software based on SLAM technology (Simultaneous localization and mapping). This technology enables machines and devices to understand where they are, how they are moving, and the structure of their environment.
For more information, please refer to Kudan’s website, or if you want to use a Kudan as your SLAM solution for your 3D lidar, you can contact [email protected].

How to Integration the Kudan Packages into Seirios RNS

First, you have to download the Kudan ROS Packages, If you haven't installed the Seirios RNS yet, you can download the kudan-movel easy deploy package using this link (only for x86_64 architecture). But if you already installed the Seirios RNS yet, you can contact [email protected] for the link to download the Kudan Packages. Then please make sure your seirios-ros image version is at least 2.48.9 or newer. You can check it by executing this command: docker ps | grep seirios-ros And see the version on this line:
registry.gitlab.com/movelai/client_deploy/seirios_ros/master_x86:<version>
If your version is older than 2.48.9, please contact [email protected] for updating the version.

1. Kudan Setup

  1. 1.
    Copy the kdlidar_ros package into the ~/catkin_ws/movel_ai/ folder
  2. 2.
    Change the permission of kdlidar_ros, so it can be executed inside the docker chmod -R +x ~/catkin_ws/movel_ai/kdlidar_ros/
  3. 3.
    Configure tf/transform parameters in these launch files:
    1. 1.
      ~/movel_ws/kdlidar_ros/install/share/kdlidar_ros/launch/kdlidar_ros_pcl_localise.launch , Change the platform_to_lidar0_rot and platform_to_lidar0_trans based on your tf from lidar link to base_link.
      <rosparam param="platform_to_lidar0_rot">
      [1, 0, 0,
      0, 1, 0,
      0, 0, 1]
      </rosparam>
      <rosparam param="platform_to_lidar0_tran">
      [0.032, 0.000, -0.178]
      <!-- [0, 0, 1.73] -->
      </rosparam>
    2. 2.
      ~/movel_ws/kdlidar_ros/install/share/kdlidar_ros/launch/kdlidar_ros_pcl_mapping.launch, Change the platform_to_lidar0_rot and platform_to_lidar0_trans based on your tf from lidar link to base_link.
      <rosparam param="platform_to_lidar0_rot">
      [1, 0, 0,
      0, 1, 0,
      0, 0, 1]
      </rosparam>
      <rosparam param="platform_to_lidar0_tran">
      [0.032, 0.000, -0.178]
      <!-- [0, 0, 1.73] -->
      </rosparam>

2. Docker Setup

  1. 1.
    Create the ros_entrypoint.sh in the ~/catkin_ws/movel_ai/ folder.
    #!/bin/bash
    # setup ros environment
    source "/opt/ros/$ROS_DISTRO/setup.bash"
    source /home/movel/kdlidar_ros/install/setup.bash
    exec "$@"
    Also, change the permission of the ros_entry_point.sh by executing this command: sudo chmod +x ~/catkin_ws/movel_ai/ros_entrypoint.sh
  2. 2.
    Edit your docker-compose.yaml file in the ~/catkin_ws/movel_ai/ folder. Under the seirios-ros volumes, add the following line:
    seirios-ros:
    ...
    volumes:
    ...
    -/home/$USER/catkin_ws/movel_ai/kdlidar_ros:/home/movel/kdlidar_ros:rw
    -/home/$USER/catkin_ws/movel_ai/ros_entrypoint.sh:/ros_entrypoint.sh:rw
    ...
  3. 3.
    Add the following variables under the seirios-backend and seirios-frontend environment:
    seirios-frontend:
    ...
    environment:
    THREE_D_MAPPING: "true"
    SHOW_3D_VIEW: "true"
    ...
    seirios-backend:
    ...
    environment:
    SHOW_3D_VIEW: "true"
    ...

3. task_supervisor Config Setup

Edit the task_supervisor.yaml (at ~/catkin_ws/movel_ai/config/task_supervisor/config/task_supervisor.yaml)
  1. 1.
    Add kudan_localization_handler to the plugins section with the following lines and comment if there's any plugin with the same type number. Example:
    - {name: kudan_localization_handler, type: 31, class: 'task_supervisor::KudanLocalizationHandler'}
    # - {name: pcl_localization_handler, type: 31, class: 'task_supervisor::PCLLocalizationHandler'}
  2. 2.
    Add a kudan_localization_handler section with the following parameters at the end of the file:
    kudan_localization_handler:
    watchdog_rate: 2.0
    watchdog_timeout: 0
    kudan_localization_launch_package: "kdlidar_ros"
    kudan_localization_launch_file: "kdlidar_ros_pcl_localise.launch"
    move_base_launch_package: "movel"
    move_base_launch_file: "navigation_common.launch"
    localization_map_dir: "/home/movel/.config/movel/maps"
    navigation_map_dir: "/home/movel/.config/movel/maps/nav"
    kudan_localization_launch_nodes: "/obs_cloud_to_scan /kdlidar_ros_pcl /velocity_limiter /anti_shin_buster_node /rgbd_to_base /velocity_setter_node /plan_inspector"
  3. 3.
    Add kudan_slam_handler to the plugins section with the following lines and comment if there's any plugin with the same type number. Example:
    - {name: kudan_slam_handler, type: 32, class: 'task_supervisor::KudanSlamHandler'}
    # - {name: pcl_slam_handler, type: 32, class: 'task_supervisor::PCLSlamHandler'}
  4. 4.
    Add a kudan_slam_handler section with the following parameters at the end of the file:
    kudan_slam_handler:
    watchdog_rate: 2.0
    watchdog_timeout: 0.0
    save_timeout: 20.0
    map_topic: "/map"
    kudan_slam_launch_package: "kdlidar_ros"
    kudan_slam_launch: "kdlidar_ros_pcl_mapping.launch"
    kudan_map_saver_package: "task_supervisor"
    kudan_map_saver_launch: "map_saver.launch"
    three_to_two_package: "movel_octomap_server"
    three_to_two_launch: "pointcloud_grid.launch"
  5. 5.
    After you edit the task_supervisor.yaml, please restart all the docker containers by executing the following commands:
    cd ~/catkin_ws/movel_ai
    docker-compose down && docker-compose up -d
Now you can do the 3D Mapping using Kudan Package by clicking the "Mapping" feature on Seirios RNS UI. If you are facing any issues and have questions, please feel free to ask by reaching us at [email protected]