Subscriber of ros topic for saving lidar sensor

• Resource ros2 (foxy) setup file in current terminal source /opt/ros/foxy/setup.bash

• On the root of pkg, check if dependencies are all meet using rosdep rosdep install -i --from-path src --rosdistro foxy -y

• Everytime changes are made, ros2 package must be rebuild using colcon build --packages-select py_subscriber (Need to source ros2 environment )

• Need to resource the created pkg setup file in a new terminal before being able to run the package. source path_to_pkg_root/install/setup.bash

• pkg can run using ros2 run py_subscriber listener [more details below]

• Data will be saved to lidar_data folder in csv format. Contains all data from lidar [x, y, z, intensity in our case.]

We can pass command line arguments to the listener node. We have the option to choose the topic to subscribe the node, the fileprefix for saving file and fileformat (we only support csv for now). We only save lidar data for now (data of type sensor_msgs.msg.PointCloud2)

ros2 run py_subscriber listener -t /carla/ego_vehicle/semantic_lidar -s semantic_lidar_frame_ -f csv

• -t or --ros-topic :- topic to subscribe to , default = /carla/ego_vehicle/lidar

• -s or --save-fileprefix :- prefix of filename to save the file arriving on topic. default = frame_

• -f or --outputfile-format :- Format to save, defaults to csv.

System Setup

• Os - Ubuntu 20.04

• carla 0.9.13 (packaged version)

• ros2 foxy for ubuntu 20.04

• carla-ros-bridge (build from source and sourcing as source ~/carla-ros-bridge/install/setup.bash). We need to source everytime in a terminal if we use carla_ros_bridge

• Running an ego vehicle on the terminal where ros-bridge was sourced using ros2 launch carla_ros_bridge carla_ros_bridge_with_example_ego_vehicle.launch.py

• By default lidar sensor is attached to the vehicle, the lidar sensor publishes data to ros2 topic /carla/ego_vehicle/lidar . We can view in Rviz using ros2 run rviz2 rviz2 and subscribing to the appropriate topic. [Need to fix the Fixed Frame issue in Rviz by subscribing to appropriate topic ego_vehicle/lidar so that we can visualize the current lidar data and not a fixed place data]

• The data can be visualized from the lidar sensor, which can be view using ros2 command ros2 topic echo /carla/ego_vehicle/lidar. Can also use rqt_graph on the terminal to visualize the node.