🌽 AgroBot Navigation

Gazebo simulation + navigation nodes for autonomous row traversal in a virtual maize field.
Built with ROS 2 Humble.

---

✅ What You Need

• Docker Desktop
• VS Code with the Dev Containers extension
• An X11 server for GUI windows (VcXsrv on Windows, XQuartz on macOS)
• (Optional) NVIDIA GPU + nvidia-container-toolkit — only needed for the ZED camera container

---

🚀 Quick Start

Step 1 — Open the Dev Container

Open this folder in VS Code. It will prompt you to reopen in a container.

Pick this container	When you want to…
ROS	Run the simulation (no GPU needed)
ZED_ROS	Work with the physical ZED camera (needs NVIDIA GPU)

Linux users: allow X11 access before opening the container:

xhost +local:docker

---

Step 2 — Build

cd /home/vscode/workspace
colcon build --symlink-install
source install/setup.bash

Tip: add the source command to your shell so you don't have to run it every time:

echo "source /home/vscode/workspace/install/setup.bash" >> ~/.bashrc

---

Step 3 — Generate a World

python3 generate_world.py

This creates a maize field world file at ~/.ros/virtual_maize_field/generated.world.

---

Step 4 — Launch the Simulation

Terminal 1 — start Gazebo + navigation:

ros2 launch my_robot_description simulation.launch.py

Terminal 2 — spawn the robot:

ros2 launch my_robot_description spawn_sdf_robot.launch.py

That's it — you should see the robot in the field. 🎉

Optional launch arguments

ros2 launch my_robot_description simulation.launch.py paused:=True headless:=True

Argument	Default	What it does
use_sim_time	True	Use the simulation clock
paused	False	Start Gazebo paused
headless	False	Run without the Gazebo GUI
world_path	~/.ros/virtual_maize_field/	Folder with the .world file
world_name	generated.world	World filename

---

🧭 Running Navigation Nodes

The simulation launch already starts the gap follower by default.
To run other nodes instead, use the commands below.

Wall Follower (LiDAR)

Follows the left wall using a PD controller on raw LiDAR scans.

ros2 launch tasks12 tasks12_launch.py type:=lidar

Gap Follower

Finds the largest open gap in the scan and steers toward it.

ros2 launch tasks12 gap_follower_launch.py

Single Node (no launch file)

ros2 run tasks12 wall_follower_lidar
ros2 run tasks12 gap_follower_node
ros2 run tasks12 demo_lidar          # fake LiDAR data for testing without Gazebo
ros2 run tasks12 explorer            # frontier exploration (needs Nav2 + a map)

---

📦 Node Reference

Node	What it does	Publishes to
wall_follower_lidar	PD wall follower on raw LiDAR	/drive (Ackermann)
gap_follower_node	Reactive gap finder	/cmd_vel (Twist) + /steer_fl, /steer_fr
demo_lidar	Synthetic LiDAR publisher for offline tests	/scan
explorer	Frontier-based exploration via Nav2	Nav2 action goals
wall_follower	Map-based wall follower (WIP — not functional yet)	Nav2 action goals

---

📂 Project Layout

Navigation/
├── .devcontainer/
│   ├── ROS/                  ← simulation-only container
│   └── ZED_ROS/              ← ZED camera + GPU container
├── containers/
│   ├── ROS-Dev.Dockerfile
│   └── scripts/              ← install scripts (ROS 2, cmake, FastLIO2, …)
├── launch/
│   └── spawn_sdf_robot.launch.py
├── models/my_robot/          ← robot SDF model
├── tasks12/                  ← navigation ROS 2 package
│   ├── launch/               ← launch files
│   └── tasks12/              ← Python source (nodes listed above)
├── virtual_maize_field/      ← world generation plugin
├── path-planning/            ← path planning workspace
├── sensor-fusion/            ← sensor fusion workspace
├── generate_world.py         ← world generator script
├── simulation.launch.py      ← main simulation launch
└── package.xml / CMakeLists.txt

---

📡 Key ROS Topics

Topic	Message Type	Used by
/scan	sensor_msgs/LaserScan	wall follower, gap follower, demo lidar
/drive	ackermann_msgs/AckermannDriveStamped	wall follower (LiDAR)
/cmd_vel	geometry_msgs/Twist	gap follower
/steer_fl, /steer_fr	std_msgs/Float64	gap follower
/map	nav_msgs/OccupancyGrid	explorer, map-based wall follower
/odom	nav_msgs/Odometry	bridged from Gazebo
/imu	sensor_msgs/Imu	bridged from Gazebo
/clock	rosgraph_msgs/Clock	simulation time sync