ros2_full_sensor_suite

This is a ROS2 package that aims to provide easy integration for simulation of essential sensors for mobile robots.

The sensor suite consists of following components;

Sensor type	Topic Name(s)	Message Type	Update Rate
LIDAR	/velodyne_points	sensor_msgs::msg::PointCloud2	30
COLOR CAMERA	/camera/color/image_raw	sensor_msgs::msg::Image	30
DEPTH CAMERA	/camera/aligned_depth_to_color/image_raw	sensor_msgs::msg::Image	30
IR1 CAMERA	/camera/infra1/image_raw	sensor_msgs::msg::Image	1
IR2 CAMERA	/camera/infra2/image_raw	sensor_msgs::msg::Image	1
WIDE ANGLE CAMERA	/wide_angle_camera/image_raw	sensor_msgs::msg::Image	1
GPS	/gps/fix	sensor_msgs::msg::NavSatFix	30
IMU	/imu	sensor_msgs::msg::Imu	30

All this sensors are composed to one gazebo model named full_sensor_suite. Refer to models folder to find the related files. There is also a Gazebo World plugin, which is used to attach the sensor suite model to any robot model. See the provided test.world for example usage.

Many of this sensor plugins are already available in official gazebo_ros_pkgs, However some sensors such as GPS and IMU were troublesome to get them work with robot_localization.

GPS and IMUs from hector_gazebo were ported to ROS2 to resolve the issues.

Installation requirements

This code was developed under Ubuntu 20.04 with Foxy version of ROS2. Make sure to have a similar setup.

clone this repo to your workspace/src and let the rosdep resolve dependencies with ;

rosdep install -y -r -q --from-paths src --ignore-src --rosdistro foxy

finally build with colcon;

colcon build --symlink-install --cmake-args -DCMAKE_BUILD_TYPE=Release --parallel-workers 4

Example usage

Replace botanbot with you robot name and change the fileds in sensor_base_attacher_plugin accordingly.

<?xml version="1.0"?>
   <sdf version="1.6">
       .
       .
       .
       <world name="test_world">   
           <include>
               <pose>0.0 0 0.5 0.0 0.0 0.0</pose>
               <uri>model://botanbot</uri>
           </include>

           <include>
               <pose>0.0 0 1.0 0.0 0.0 0.0</pose>
               <uri>model://full_sensor_suite</uri>
           </include>

           <plugin name="sensor_base_attacher_plugin" filename="libsensor_base_attacher_plugin.so">
               <robot_model_name>botanbot</robot_model_name>
               <robot_base_link_name>base_link</robot_base_link_name>
               <sensor_suite_model_name>full_sensor_suite</sensor_suite_model_name>
               <sensor_suite_base_link_name>base_sensor_link</sensor_suite_base_link_name>
           </plugin>
           .
           .
           .
       </world>
   </sdf>

Launch the world file with;

export GAZEBO_MODEL_PATH=/home/atas/colcon_ws/src/ros2_full_sensor_suite/models ros2 launch ros2_full_sensor_suite test.launch.py

replace the GAZEBO_MODEL_PATH according to your local path. This is important as otherwise Gazebo would not be able to locate the models.

Open RVIZ2 in another terminal, set the global frame to base_link. You can view various incoming sensor data such as LIDAR, depth cloud, images etc.;

Note that since the depth camera pointcloud is coming from optical frmae, the cloud looks strange, you will need to do transfrom e.g(depth_camera_optical_frame -> base_link) so that the point cloud from depth camera will look more meaningful.

Some sensor data cannot be visualized through RVIZ2, but you can just subscribe to a topic to echo the data.

e.g

ros2 topic echo /imu/data

or the same from gps

ros2 topic echo /gps/fix

NMBURobotics/ros2_full_sensor_suite

ros2_full_sensor_suite

Installation requirements

Example usage