ros2-lidar-obstacle-avoidance

Reactive obstacle avoidance for a differential drive robot using LiDAR sector filtering in ROS2 Humble, simulated in Gazebo.

Comment

ROS2 Lidar-Based A* Navigation with Reactive Obstacle Avoidance

Overview

This project implements a complete navigation stack in ROS2 (Humble) including:

  • Custom A* global planner
  • Path publishing using nav_msgs/Path
  • Waypoint-based path following
  • FSM-based reactive obstacle avoidance using sensor_msgs/LaserScan
  • Simulation in Gazebo with TurtleBot3

Architecture

Global Planner (A*)

  • Path Publisher
  • Path Follower (Odometry-based)
  • Reactive Avoidance (FSM + hysteresis)

Features

  • Standalone A* grid planner
  • Finite-state machine for obstacle handling
  • Yaw-based avoidance commitment
  • Oscillation prevention logic
  • Gazebo-tested in dynamic scenarios

How to Run

cd ~/ros2_ws colcon build source install/setup.bash

Terminal 1

ros2 launch turtlebot3_gazebo turtlebot3_world.launch.py

Terminal 2

ros2 run astar_nav_cpp astar_planner_node

Terminal 3

ros2 run astar_nav_cpp path_follower_node

Simulation Environment

  • ROS2 Humble
  • Gazebo
  • TurtleBot3 Burger

Future Improvements

  • Dynamic costmap integration
  • Replanning during deviation
  • Nav2 comparison benchmark
  • RViz visualization markers

Built With

Share this project:

Updates