IGVC Simulation - Autonomous Robot Navigation

Autonomous navigation simulation for the IGVC (Intelligent Ground Vehicle Competition) course using ROS 2 Humble, Gazebo, and Navigation2. Features lane detection, LiDAR obstacle avoidance, and waypoint-based path following.

Clone from GitHub → git clone https://github.com/PixelPepper/IGVC_2.git → Works from any directory (no hardcoded paths).

---

Features

• Gazebo simulation – Orange robot in IGVC-style world with lanes, cones, and obstacles
• Lane detection – HSV-based white/yellow lane segmentation via OAK-D Pro camera
• Navigation2 – Path planning and obstacle avoidance with fused LiDAR + lane boundaries
• Waypoint navigation – Sequential waypoint following through the course
• Robot localization – EKF fusion of wheel odometry and IMU
• RViz visualization – Costmaps, paths, LiDAR, and camera feeds

---

System Requirements

• OS: Ubuntu 22.04 (Jammy)
• ROS 2: Humble
• Gazebo: Gazebo Classic (gazebo11)
• Python: 3.10+
• Display: For Gazebo and RViz (X11 or headless with virtual display)

---

Dependencies

ROS 2 Packages

Install ROS 2 Humble and common packages:

sudo apt update
sudo apt install -y ros-humble-desktop ros-humble-gazebo-ros-pkgs
sudo apt install -y ros-humble-nav2-bringup ros-humble-navigation2
sudo apt install -y ros-humble-robot-localization
sudo apt install -y ros-humble-cv-bridge ros-humble-vision-opencv
sudo apt install -y ros-humble-pcl-ros ros-humble-pcl-conversions
sudo apt install -y ros-humble-pointcloud-to-laserscan
sudo apt install -y ros-humble-robot-state-publisher ros-humble-joint-state-publisher
sudo apt install -y ros-humble-xacro ros-humble-urdf
sudo apt install -y ros-humble-gazebo-plugins
sudo apt install -y ros-humble-tf2 ros-humble-tf2-ros ros-humble-tf2-geometry-msgs
sudo apt install -y ros-humble-laser-geometry
sudo apt install -y ros-humble-sensor-msgs ros-humble-geometry-msgs ros-humble-nav-msgs

Python / System

sudo apt install -y python3-opencv python3-numpy python3-pip
sudo apt install -y libpcl-dev pcl-tools

Optional (if using additional packages)

# Linefit ground segmentation (orange_sensor_tools)
sudo apt install -y ros-humble-linefit-ground-segmentation-ros

# Velodyne (if using 3D LiDAR)
sudo apt install -y ros-humble-velodyne

---

Installation

1. Clone the repository

git clone https://github.com/PixelPepper/IGVC_2.git
cd IGVC_2

The repository is self-contained (no submodules).

2. Workspace path

setup_orange.sh auto-detects the workspace from its location—no path edits needed. Clone as IGVC_2 (default) or any name; just cd into your clone directory before sourcing.

3. Build the workspace

source /opt/ros/humble/setup.bash
colcon build --symlink-install
source install/setup.bash

4. Source the setup script

source setup_orange.sh

---

Quick Start

Launch simulation (4 terminals)

Terminal 1 – Gazebo

cd IGVC_2   # or your workspace path
source setup_orange.sh
ros2 launch orange_gazebo orange_igvc_simple.launch.xml

Terminal 2 – Navigation + Perception

cd IGVC_2
source setup_orange.sh
ros2 launch orange_gazebo igvc_perception_full.launch.xml

Terminal 3 – RViz (optional)

cd IGVC_2
source setup_orange.sh
rviz2 -d igvc_perception.rviz

Terminal 4 – Run course

cd IGVC_2
source setup_orange.sh
python3 navigate_igvc_course.py
# or: ./run_igvc_course.sh

---

Architecture

The system uses a layered ROS 2 architecture:

• Simulation (Gazebo) → Localization (EKF) → Perception (lane detection, fusion) → Navigation2 → cmd_vel → Gazebo

Key topics: /odom, /imu, /hokuyo_scan, /oak/rgb/image_raw → EKF and perception → /fused_scan, /lane_cloud → Nav2 costmaps → /cmd_vel.

Documentation:

• docs/ARCHITECTURE.md – In-depth architecture, layers, TF tree, config files
• docs/TOPIC_GRAPH.md – Mermaid diagram and topic tables
• docs/SETUP_GUIDE.md – Step-by-step setup for new users

---

Project Structure

IGVC_2/
├── src/
│   ├── orange_ros2/           # Core Orange robot packages
│   │   ├── orange_description # Robot URDF, sensors (OAK-D Pro, Hokuyo)
│   │   ├── orange_gazebo      # Gazebo worlds, spawn, launch
│   │   ├── orange_perception  # Lane detection, point cloud fusion
│   │   ├── orange_sensor_tools
│   │   ├── orange_navigation
│   │   └── ...
│   ├── linefit_ground_segmentation_ros2/
│   ├── velodyne/
│   └── ...
├── docs/                      # Architecture and setup documentation
│   ├── ARCHITECTURE.md        # System design, layers, TF tree
│   ├── TOPIC_GRAPH.md         # Topic diagram and tables
│   └── SETUP_GUIDE.md         # New user setup guide
├── navigate_igvc_course.py    # Waypoint navigation script
├── nav2_params_fused.yaml     # (legacy; primary: orange_gazebo/config/)
├── igvc_perception.rviz       # RViz config
├── setup_orange.sh            # Workspace setup
├── FINAL_SETUP_GUIDE.md       # Detailed setup instructions
└── README.md

---

Configuration

• Waypoints: Edit navigate_igvc_course.py (self.waypoints)
• Nav2 params: orange_gazebo/config/nav2_params_fused.yaml (costmaps, inflation, goal tolerance)
• Lane detection: orange_perception (HSV thresholds, morphological ops)
• Robot spawn: orange_igvc_simple.launch.xml (position, yaw)

---

Troubleshooting

Issue	Solution
Gazebo window doesn't open	Run DISPLAY=:0 gzclient & in a separate terminal
Lane detection shows 0 points	Check camera topic /oak/rgb/image_raw, adjust camera pitch/position
Robot stuck at waypoint	Increase xy_goal_tolerance in nav2_params_fused.yaml
Build errors	Run rosdep install --from-paths src --ignore-src -r -y

---

License

Apache-2.0 (see package package.xml files)

---

Acknowledgments

• KBKN Autonomous Robotics Lab – Original Orange robot repository
• ROS 2 Navigation2, Gazebo, OpenCV