A lightweight UAV simulation environment for evaluating different levels of global and local trajectory optimization and planning.

1. Local Point Cloud

   • Topic: /drone_<drone_id>_pcl_render_node/cloud
   • Type: sensor_msgs/PointCloud2
   • Description: Local 3D point cloud data around the robot.

2. Odometry

   • Topic: /drone_<drone_id>_odom
   • Type: nav_msgs/Odometry
   • Description: Drone's current odometry in simulation.

1. Trajectory

   • Topic: /drone_<drone_id>_planning/traj
   • Type: traj_msgs/SingleTraj
   • Description: Refers to SingleTraj

2. (Optional) Position Command

   • Topic: /drone_<drone_id>_planning/pos_cmd
   • Type: quadrotor_msgs/PositionCommand
   • Description: Refers to PositionCommand

• Replace <drone_id> with your assigned drone ID (default is 0).

• Use the grid map module used by HKUST Aerial Lab and Fast Lab (take the implementation in the ego planner).
• Provide a local map, local goal, and initial path (if you need it as an initial guess)
• Input the generated trajectories.

1. Local Grid Map

2. Odometry

   • Topic: /drone_<drone_id>_odom
   • Type: nav_msgs/Odometry
   • Description: Drone's current odometry in simulation.

1. Trajectory

   • Topic: /drone_<drone_id>_planning/traj
   • Type: traj_msgs/SingleTraj
   • Description: Refers to SingleTraj

2. (Optional) Position Command

   • Topic: /drone_<drone_id>_planning/pos_cmd
   • Type: quadrotor_msgs/PositionCommand
   • Description: Refers to PositionCommand

• Replace <drone_id> with your assigned drone ID.

1. Global Grid Map

2. Odometry

   • Topic: /drone_<drone_id>_odom
   • Type: nav_msgs/Odometry
   • Description: Drone's current odometry in simulation.

1. Trajectory

   • Topic: /drone_<drone_id>_planning/traj
   • Type: traj_msgs/SingleTraj
   • Description: Refers to SingleTraj

2. (Optional) Position Command

   • Topic: /drone_<drone_id>_planning/pos_cmd
   • Type: quadrotor_msgs/PositionCommand
   • Description: Refers to PositionCommand

• Replace <drone_id> with your assigned drone ID.

# Go to your ROS2 workspace
cd ~/ros2_ws/src

# Clone this repo
git clone --recurse-submodules https://github.com/yuwei-wu/planner_benchmark.git

# Go to workspace root
cd ~/ros2_ws

# Build
colcon build --symlink-install

For single drone:

source install/setup.bash
ros2 launch api_server level0_eval.launch.py

For multiple drones:

source install/setup.bash
ros2 launch api_server level0_eval_swarm.launch.py

• Thanks to the OpenAI GPT for assistance in developing the code and documentation.
• Reference:
  • https://github.com/ZJU-FAST-Lab/ego-planner-swarm/tree/ros2_version