Autonomous Indoor Mapping and Navigation System

Robot Programming Course Project • 2022 Spring Semester

Kwangwoon University, Division of Robotics

Overview

This project implements a real-time SLAM (Simultaneous Localization and Mapping) system for autonomous exploration and return-to-origin navigation using TurtleBot3. The robot autonomously maps unknown indoor environments and navigates back to its starting position using the shortest path.

<aside> 🎯

Key Achievement: Successfully achieved approximately 80% success rate in autonomous indoor space estimation and return navigation across multiple test environments.

</aside>

Technical Approach

System Architecture

The system integrates three major components:

Cartographer SLAM

• Google-developed SLAM library for real-time 2D/3D environment mapping
• Utilizes TurtleBot3's LiDAR sensor data for spatial mapping
• Provides real-time position estimation and map generation

Navigation2 (Nav2)

• Redesigned navigation stack from ROS1's move_base
• Improved code efficiency and error handling
• Manages optimal path planning and obstacle avoidance
• Behavior tree architecture for modular action execution

Autonomous Exploration