Table of Contents

• Folder Structure
• ROS Topics
• Getting Started

Folder Structure

• scripts/

  • Utility scripts for working with CARLA.
  • Example: carla_cleanup.py — cleans up the CARLA simulator before launching ROS 2 nodes.

• src/

  • carla_vehicle_manager/

    • Manages vehicle control and sensor interfaces in CARLA.
    • Sensors:
      • IMU Node
      • GNSS Node
    • Topics:
      • /carla/imu_sensor
      • /carla/gnss
      • /carla/vehicle/*

  • hydrakon_launch/

    • Unified launch file that launches all core modules.
      • Has EKF Parameters directly baked into the launch file

  • lidar_cluster/

    • Processes incoming LiDAR point clouds to detect and cluster potential cones using DBSCAN.
    • Topics:
      • /perception/lidar_cluster
      • /perception/cone_markers

  • lidar_handler/

    • Simulates the Robosense Helios 16 LiDAR using a Carla sensor wrapper.
    • Sensor Node:
      • LiDAR Node
    • Topics:
      • /carla/lidar

  • perception_module/

    • Handles perception-related transformations, localization, and sensor fusion (e.g., navsat transform).
    • Topics:
      • /gps/filtered
      • /gps/pose
      • /odometry/gps
      • /odometry/filtered

  • planning_module/

    • (To be implemented) Will handle path planning algorithms such as Pure Pursuit and MPC.

  • zed2i_camera_sim/

    • Simulates the ZED 2i stereo camera in Carla.
    • Topics:
      • /zed2i/camera_info
      • /zed2i/depth/image
      • /zed2i/rgb/image

---

ROS2 Topics

🛰 Sensors & Environment

• /carla/imu_sensor
  IMU data from Carla’s onboard IMU
  ‣ Type: sensor_msgs/msg/Imu

• /carla/gnss
  GNSS data from Carla’s onboard GPS sensor
  ‣ Type: sensor_msgs/msg/NavSatFix

• /carla/lidar
  Raw LiDAR point cloud
  ‣ Type: sensor_msgs/msg/PointCloud2

• /zed2i/camera_info
  Intrinsic camera parameters from the ZED2i simulation
  ‣ Type: sensor_msgs/msg/CameraInfo

• /zed2i/depth/image
  Depth image from the ZED2i simulation
  ‣ Type: sensor_msgs/msg/Image

• /zed2i/rgb/image
  RGB image from the ZED2i simulation
  ‣ Type: sensor_msgs/msg/Image

---

🚘 Vehicle Control

• /carla/vehicle/*
  Set of vehicle control topics (e.g., throttle, brake, reverse)
  ‣ Type: Carla-specific control interfaces
  ‣ Driven via keyboard_control_node.py

---

🧠 Perception & Localization

• /perception/lidar_cluster
  Clustered cone candidates from LiDAR
  ‣ Type: sensor_msgs/msg/PointCloud2

• /perception/cone_markers
  RViz markers representing clustered cones
  ‣ Type: visualization_msgs/msg/MarkerArray

• /gps/filtered
  GPS data after initial fusion/filtering
  ‣ Type: sensor_msgs/msg/NavSatFix

• /gps/pose
  GPS fused pose as geometry_msgs/PoseStamped
  ‣ Type: geometry_msgs/msg/PoseStamped

• /odometry/gps
  Odometry based on raw GNSS + IMU
  ‣ Type: nav_msgs/msg/Odometry

• /odometry/filtered
  Contains odometry output from robot_localization EKF
  ‣ Type: nav_msgs/msg/Odometry

---

🧭 Transforms & Logs

• /tf
  Dynamic transform tree (e.g., map → base_link)
  ‣ Type: tf2_msgs/msg/TFMessage

• /tf_static
  Static transforms (e.g., base_link → imu_link)
  ‣ Type: tf2_msgs/msg/TFMessage

• /rosout
  Internal ROS 2 logging messages (for rqt_console, ros2 log)
  ‣ Type: rcl_interfaces/msg/Log

Getting Started

1. ✅ Prerequisites

   • Ensure the following are installed:
     • Ubuntu 22.04 (Preferred) or later
     • ROS2 Humble
     • CARLA Simulator
     • Python 3.10
     • colcon, rosdep, and other ROS dev tools
     • NVIDIA GPU with drivers for Carla rendering

2. 📁 Clone the Workspace

   cd ~
   git clone https://github.com/<your-username>/hydrakon_ws.git
   cd hydrakon_ws

3. 📦 Build the Workspace

   source /opt/ros/humble/setup.bash
   rosdep install --from-paths src --ignore-src -r -y
   colcon build --symlink-install

4. 🔧 Source the Setup

   source install/setup.bash

5. 🧪 Run Carla Simulator

   cd ~/<location-to-your-Carla-sim>
   cd Linux
   ./CarlaUnreal.sh 

   Use the -quality-level=Low flag if needed

   ./CarlaUnreal.sh -quality-level=Low

6. 🚗 Launch the Full System

   ros2 launch hydrakon_launch hydrakon_launch.py

7. 📊 Visualize in RViz

   rviz2
   

   Open this in a new terminal, and make sure base_link is selected as the correct fixed frame, then go ahead and choose your desired topics to visualize