1. Theses & HiWi Jobs
    2.  

Implementation and Evaluation of SLAM Algorithms for Mobile Robot Navigation

Institute
Lehrstuhl für Fördertechnik Materialfluss Logistik
Type
Semester Thesis / Master's Thesis /
Content
experimental / constructive /  
Description

Introduction

Autonomous mobile robots require accurate maps of their surroundings to perform useful tasks. These maps are essential for applications ranging from path planning in logistics and inventory management to site monitoring and creating digital floor plans in construction. While numerous open-source Simultaneous Localization and Mapping (SLAM) algorithms are available, their performance varies significantly depending on the robot's sensors, its dynamics, and the structure of the environment.

This thesis aims to address this challenge by implementing, tuning, and systematically comparing leading open-source SLAM solutions. The goal is to identify the robust and accurate system for our robotic platform, providing a reliable foundation for future autonomous navigation research.

Objectives

The primary goal of this thesis is to implement and benchmark 3D SLAM algorithms for our mobile robot by evaluating prominent open-source packages.

The key objectives are:

  • System Integration: To implement and integrate at least two open-source SLAM packages on our mobile robot using ROS.
  • Systematic Tuning: To methodically tune the key parameters of each SLAM system to optimize its performance for our specific hardware and target environment.
  • Comparative Evaluation: To conduct a rigorous evaluation of the selected SLAM systems based on quantitative metrics for map accuracy, localization stability, and computational requirements.

Methodology

1. Sensor Setup and Calibration:

  • Ensure the robot's sensors (3D LiDAR and IMU) are correctly configured and publishing data within the ROS.
  • Verify the accuracy of sensor data and the integrity of the robot's coordinate frame transformations

2. SLAM System Implementation:

  • Install and configure the chosen SLAM packages (e.g., Google Cartographer).
  • Run each SLAM system on the robot and establish baseline performance.

3. Experimental Validation and Performance Analysis:

  • Research the key parameters of each algorithm and their impact on performance.
  • Design and conduct experiments to quantify the effectiveness of the algorithms.

4. Comparative Evaluation:

  • Assess the quality of maps generated by each system based on consistency and alignment accuracy.
  • Evaluate localization accuracy by measuring drift after completing loops.
  • Compare the robustness and computational load of each system under different conditions (e.g., open spaces vs. feature-rich laboratory).

Expected Outcomes

  • A Recommendation for an Optimal SLAM Solution: A data-driven recommendation for the best-performing SLAM package for our specific robot and application.
  • A Set of Benchmarking Datasets and High-Quality Maps: The creation of valuable datasets for future testing and a set of accurate maps of key operational areas.
  • A Comprehensive Thesis Report: A report detailing the implementation of each system, the tuning methodology, and a robust comparative analysis of their performance.
Requirements

Requirements and Qualifications

  • Background in Mechatronics, Robotics, or a related field.
  • Practical experience with the Robot Operating System (ROS/ROS2).
  • Hands-on experience with mobile robots and sensors (e.g., IMU, LiDAR).
  • A theoretical understanding of SLAM algorithms would be beneficial.
  • A proactive and independent approach to problem-solving.
  • Good communication skills in English.

Application

We look forward to your application. If you're interested in writing your thesis with us, please email your CV, proof of grades, and a brief introduction of your interest and experience. Feel free to contact us with any questions.

Possible start
sofort
Contact
Yuan-Jen Huang, M.Sc.
Room: MW 0501
Phone: +49 (89) 289 - 15931
yuan-jen.huangtum.de