Mobile Robot Simulation using ROS 2

Repository: https://github.com/Jpst01/mobile_robot_ros2
Author: Jayeshwar Pratap Singh Tanwar
Project Type: Individual / Solo Project
Development Status: Actively under development

---

1. Project Overview

This repository contains a ROS 2–based mobile robot simulation developed as an individual project, focusing on low-level understanding of robot description, TF frames, sensor integration, and simulation behavior. The project uses ROS 2 Jazzy Jalisco with Gazebo Harmonic and RViz2 to simulate a differential-drive mobile robot equipped with common sensors.

---

2. Motivation and Learning Objectives

The primary goals of this project are:

• To gain a solid understanding of ROS 2 system architecture and middleware behavior.
• To design and simulate a mobile robot with a clean and correct TF tree.
• To integrate multiple sensors and validate their data in simulation.
• To implement SLAM-based environment mapping using Nav2 and slam_toolbox.

This project is used as a learning and validation platform for robotics fundamentals relevant to real-world embedded and robotic systems.

---

3. System Architecture

The system follows a standard ROS 2 architecture:

• Robot description published via robot_state_publisher
• Simulation handled by Gazebo Harmonic
• Sensor data published through Gazebo ROS plugins
• Visualization and debugging performed in RViz2
• TF tree connecting map, odom, and robot frames
• Motion commands published to the robot controller in simulation

All components are integrated using ROS 2 launch files and built using colcon.

---

4. Robot Design and Sensor Configuration

Mechanical Structure

• Mobile robot with a rigid chassis composed of primitive geometries:
  • Box: main body
  • Cylinders: left and right wheels
  • Sphere: support / caster element
• Differential-drive configuration

Sensors

• IMU

  • Mounted on the front of the robot
  • Publishes orientation and angular velocity data

• RGB-D Camera

  • Mounted on the front of the robot
  • Provides color and depth streams

• 2D LiDAR

  • Mounted on the top of the robot
  • Used for mapping and localization

  ---

5. Software Stack and Versions

• Operating System: Ubuntu 24.04
• ROS 2: Jazzy Jalisco
• Simulator: Gazebo Harmonic
• Visualization: RViz2
• Programming Languages: C++, Python
• Robot Description: URDF with XACRO
• Build System: colcon

---

6. Package Structure

├── mobile_description
│ ├── CMakeLists.txt
│ ├── config
│ │ └── mobile_robot
│ │ ├── ros2_controllers_template.yaml
│ │ └── ros2_controllers.yaml
│ ├── launch
│ │ └── robot_state_publisher.launch.py
│ ├── LICENSE
│ ├── package.xml
│ ├── rviz
│ │ └── mobile_robot_description.rviz
│ └── urdf
│ ├── control
│ │ ├── gazebo_sim_ros2_control.urdf.xacro
│ │ └── mobile_robot_ros2_control.urdf.xacro
│ ├── mech
│ │ └── mobile_robot_base.urdf.xacro
│ ├── robots
│ │ └── mobile_robot.urdf.xacro
│ └── sensors
│ ├── imu.urdf.xacro
│ ├── lidar.urdf.xacro
│ └── rgbd_camera.urdf.xacro
├── mobile_robot
│ ├── CMakeLists.txt
│ ├── LICENSE
│ └── package.xml
├── mobile_robot_bringup
│ ├── CMakeLists.txt
│ ├── launch
│ │ ├── load_ros2_controllers.launch.py
│ │ └── mobile_robot_navigation.launch.py
│ ├── LICENSE
│ └── package.xml
├── mobile_robot_gazebo
│ ├── CMakeLists.txt
│ ├── config
│ │ └── ros_gz_bridge.yaml
│ ├── launch
│ │ └── mobile_robot.gazebo.launch.py
│ ├── LICENSE
│ ├── models
│ │ └── depot
│ │ ├── materials
│ │ ├── meshes
│ │ ├── model.config
│ │ └── model.sdf
│ ├── package.xml
│ ├── rviz
│ │ └── mobile_robot_gazebo_sim.rviz
│ └── worlds
│ ├── depot.world
│ └── empty.world
├── mobile_robot_localization
│ ├── CMakeLists.txt
│ ├── config
│ │ └── ekf.yaml
│ ├── launch
│ │ └── ekf_gazebo.launch.py
│ ├── LICENSE
│ └── package.xml
├── mobile_robot_navigation
│ ├── CMakeLists.txt
│ ├── config
│ │ └── mobile_robot_nav2_default_params.yaml
│ ├── LICENSE
│ ├── maps
│ ├── package.xml
│ └── rviz
│ └── nav2_default_view.rviz
├── mobile_robot_system_tests
│ ├── CMakeLists.txt
│ ├── LICENSE
│ └── package.xml
└── README.md

• urdf/: Robot model and sensor definitions
• launch/: Launch files for simulation and visualization
• config/: .yaml files
• rviz/: rviz configuration files

---

7. Simulation Details (Gazebo + RViz)

• Gazebo Harmonic used for physics simulation
• Robot spawns correctly with proper collision and visual geometry
• All sensors publish valid data streams
• RViz visualizes:
  • Robot model
  • TF tree
  • LiDAR scans
  • Camera topics
  • IMU data
• Robot motion is functional and controllable in simulation

---

8. SLAM and Mapping

• SLAM integration is complete using slam_toolbox (online async mode)
• The robot can be driven via keyboard teleop to map the environment
• EKF sensor fusion (wheel odometry + IMU) provides accurate localization
• 2D occupancy grid maps can be generated and saved
• Saved maps are stored in mobile_robot_navigation/maps/

---

9. Current Limitations

• Autonomous waypoint navigation is not yet implemented
• Focus is on correctness of SLAM and sensor fusion rather than performance optimization

---

10. Future Improvements

• Add Nav2-based autonomous waypoint navigation
• Implement obstacle avoidance using the saved map
• Improve simulation realism (noise, friction, sensor parameters)

---

11. How to Build and Run

Build

cd ~/mobile_robot_ws
colcon build
source install/setup.bash

Launch SLAM Mapping

ros2 launch mobile_robot_bringup mobile_robot_navigation.launch.py use_composition:=False

Drive the Robot (in a separate terminal)

# Forward
ros2 topic pub --rate 10 /diff_drive_controller/cmd_vel geometry_msgs/msg/TwistStamped "{twist: {linear: {x: 0.2}, angular: {z: 0.0}}}"

# Rotate
ros2 topic pub --rate 10 /diff_drive_controller/cmd_vel geometry_msgs/msg/TwistStamped "{twist: {linear: {x: 0.0}, angular: {z: 0.2}}}"

Save the Map

ros2 run nav2_map_server map_saver_cli -f ~/mobile_robot_ws/src/mobile_robot/mobile_robot_navigation/maps/my_map