Introduction

Developed a differential-drive (two-wheel with caster) mobile robot capable of autonomous waypoint navigation and real-time obstacle avoidance. The project bridged the gap between virtual prototyping and physical hardware by first building a complete physics simulation in Gazebo and then deploying the control stack directly onto a physical robot.

Project Details

• Core Tech: ROS2 (Robot Operating System), Gazebo (Physics Simulator), Python/C++.
• Hardware: Differential-drive mobile platform (2 active drive wheels, 1 passive roller/caster wheel), 3x infrared/ultrasonic distance sensors (front-facing).
• Concepts: Robot kinematics, state machines, sensor fusion, obstacle avoidance algorithms.

Key Features & Actions

• Virtual Prototyping (Gazebo): Designed a 3D simulation of the robot in Gazebo, defining its physical properties (mass, friction, wheel torque) and sensor limits to test navigation.
• ROS2 Control Stack: Built modular ROS2 nodes to handle sensor data acquisition, waypoint tracking, and wheel velocity commands ($/cmd_vel$), utilizing ROS2 topics and parameters for clean node communication.
• Reactive Obstacle Avoidance: Programmed an algorithm that analyzed the three front-facing distance sensors to dynamically adjust the robot's heading when obstacles were detected.
• Waypoint Navigation State Machine: Created a coordination system that calculated the distance and heading to a target coordinate, smoothly driving the robot forward while actively overriding the trajectory if an obstacle crossed its path.