ROS2/Autoware.universe bridge for CARLA simulator

This ros package enables autonomous driving using Autoware in addition to the basic function of the official ros-bridge package (communication between ros and carla)

Environment

ubuntu	ros	carla	autoware
22.04	humble	0.9.13	universe/master

example

• 11th Gen Intel® Core™ i7-11700F @ 2.50GHz × 16
• GeForce RTX3070
• 32GB mem

Setup

install

• Autoware.Universe

• CARLA Installation

• carla-ros-bridge (this repo)

  mkdir -p colcon_ws/src
  cd colcon_ws/src
  git clone git@github.com:kuriatsu/carla-autoware.git --recursive
  rosdep install -i --from-paths src

• autoware containts

  1. Download maps (y-axis inverted version) to arbitaly location
  2. Change names. (point_cloud/Town01.pcd -> Town01/pointcloud_map.pcd, vector_maps/lanelet2/Town01.osm -> Town01/lanelet2_map.osm)

• Change file path at ros-bridge/carla_spawn_objects/carla_spawn_objects.py L49 as shown in first paragraph of Fix Log.

build

cd colcon_ws
colcon build --symlink-install

Run

1. Run carla, change map, spawn object if you need

/opt/carla/CarlaUE4.sh
python3 /opt/carla/PythonAPI/util/config.py -m Town10
# optional
python3 /opt/carla/PythonAPI/examples/generate_traffic.py -n 10 -w 20

2. Run ros nodes

ros2 launch carla_ros_bridge carla_ros_bridge_with_example_ego_vehicle.launch.py
ros2 launch carla_autoware carla_autoware.xml 
ros2 launch autoware_bridge autoware_launch.xml map_path:=<path-to-map-dir>/Town10 vehicle_model:=sample_vehicle sensor_model:=sample_sensor_kit

3. Set initial pose
4. Set goal position
5. Wait for planning
6. Engage (use web interface or Rviz plugin (AutowareStatePanel))

TODO

• Need to refine vehicle controller at autoware_bridge/src/carla_vehicle_interface.cpp L82