ROS2 FollowBot

Welcome to the ROS2 FollowBot repository that aligns with our master repository, FollowBot! This repository is dedicate to creating packages for the robotic side of FollowBot. Below, you'll find an overview of the basic ROS2 commands and essential information to get started with developing for FollowBot.

Plans for development:

We will be developing a learning model that enables FollowBot to follow the user using computer vision technologies. By implementing You Only Look Once (YOLO), we will process images as features within our model to allow the device to effectively track and follow the user.

Additionally, we will integrate the slam-toolbox, robot_localization, and nav2 packages. This will enable FollowBot to better understand its surroundings and navigate to target locations on campus.

Current Development Updates:

1. Arduino and Raspberry Pi Communication

Sensor data (e.g., wheel encoders, IMU) is transmitted from the Arduino to the Raspberry Pi over a shared serial port. To avoid data collisions, the system uses:

• Thread synchronization: concurrent access to the serial port is managed via multithreading.
• and Demultiplexing: raw sensor data streams are split into separate topics (e.g., /wheel_odom, /imu/data).

We can visualize how these nodes interact, as well as critical topics, via the rqt_graph:

2. Transform Tree The image below demonstrates the aligned transformations necessary for our mapping to function effectively. Our setup includes the following frame hierarchy: map → odom → base_link → base_laser.

• map: global fixed frame (for SLAM/navigation)
• odom: drift-prone odometry frame (fused by robot_localization)
• base_link: robot's base frame
• base_laser: position and orientation of a LiDAR sensor

3. slam_toolbox Integration

Used for real-time 2D SLAM (Simultaneous Localization and Mapping). It processes LiDAR scans and odometry to generate a static map (/map topic) and updates the robot’s pose (/amcl_pose).

4. robot_localization Integration

Fuses wheel odometry (/wheel_odom), orientation from the IMU (/imu/data), and GPS (/gps/fix) into a unified odometry estimate (/odometry/filtered) using an Extended Kalman Filter (EKF).

5. nav2 Integration

Handles autonomous navigation:

• Accepts goals via /goal_pose
• Plans paths and executes velocity commands via /cmd_vel
• Relies on the /map and /amcl_pose topics for localization

Our next steps include:

• developing the navigation flow for target coordinates to be delivered to /goal_pose
• developing motion controllers that respond to /cmd_vel

More detailed information about our research can be found at ROS2 Command List (we didn't know where else to include it).

Prerequisites

• Raspberry Pi: Ensure you have a Raspberry Pi 4 or better with more than 4GB of RAM.
• ROS2: You must be using the latest version of ROS2, which is Jazzy.
• Languages: Developers can create packages in either C++ or Python.

Basic ROS2 Commands

Here are some fundamental commands to help you get started with ROS2:

Setup and Initialization

• Source ROS2 Setup Script:

source /opt/ros/jazzy/setup.bash

Tip: you can also place this in the .bashrc file so you don't have to always type in this command over and over again.

Important installs that are currently within this project

• Necessary package/library installs needed:

sudo apt update
sudo apt install libserial-dev

sudo apt update
sudo apt install libgeographiclib-dev

sudo apt update
sudo apt-get install nlohmann-json3-dev

Creating a New Package

Before creating a new package, ensure you are in the src directory of the FollowBot_AROS2 workspace.

• C++ Package:

ros2 pkg create --build-type ament_cmake my_cpp_package

• Python Package:

ros2 pkg create --build-type ament_python my_python_package

Building Packages

• Build the Workspace:

colcon build

Also: If you are needing to install an individual package and not all of them at once you can do this command and specifiy which package you want to build

colcon build --packages-select my_package

Running Nodes

• Run a C++ Node:

ros2 run my_cpp_package my_cpp_node

• Run a Python Node:

ros2 run my_python_package my_python_node.py

Listing Nodes

• List Active Nodes:

ros2 node list

Topic Commands

• List Topics:

ros2 topic list

• Echo Topic:

ros2 topic echo /topic_name

Service Commands

• List Services:

ros2 service list

• Call a Service:

ros2 service call /service_name std_srvs/srv/Empty

Extra Documentation

For more documentation in how to use ROS2 please follow the below links:

ROS2 Jazzy Documentation

ROS2 Command List

ROS2 VSCode Development

Developing for FollowBot:

FollowBot is a smart robotic device that utilizes ROS2, computer vision, and machine learning to follow users and avoid obstacles. The following steps outline how you can start developing packages:

1. Clone the Repository:

git clone https://github.com/FrankVanris2/ROS2_FollowBot.git
cd ROS2_FollowBot

2. Navigate to the src Directory:

cd FollowBot_AROS2/src

3. Create a New Package: Follow the commands above to create either a C++ or Python package.
4. Build Your Package: Use the build command to compile your new package.
5. Run Your Nodes: Excute your nodes to test their functionality.

We look forward to your contributions and innovations with FollowBot! If you have any questions or ned further assistance, please refer to the ROS2 documentation or contact our development team.