Control and quality inspection of an autonomous mobile robot for overhead crane operations

Thumbnail Image

Files

master_Hou_Yelin_2025.pdf (7.08 MB)

URL

Journal Title

Journal ISSN

Volume Title

School of Engineering | Master's thesis
Unless otherwise stated, all rights belong to the author. You may download, display and print this publication for Your own personal use. Commercial use is prohibited.

Authors

Date

2024-12-31

Department

Major/Subject

Mechanical Engineering

Mcode

Degree programme

Master's programme in Mechanical Engineering

Language

en

Pages

78

Series

Abstract

Ensuring the safety of overhead crane hooks is essential for maintaining reliable and efficient industrial operations. However, manual inspections are often time-consuming, labor-intensive, and susceptible to human error. What's more, some industrial working environments are harzard for human inspector to get in and complete the inspection task. This thesis proposes an autonomous robotic system designed to inspect the appearance and structural integrity of crane hooks in indoor environments without requiring complete equipment downtime. The system integrates a laser-based positioning sensor mounted on the crane's trolley for accurate indoor localization. Navigation and obstacle avoidance are achieved using the Nav2 stack, a ROS-based framework for autonomous navigation and real-time path planning. Computer vision techniques are employed for defect detection, specifically contour detection using OpenCV, to identify cracks and deformations on crane hooks. The proposed system was validated through both simulation in Gazebo and real-world experiments. Simulation results demonstrated reliable navigation and accurate defect detection, reducing the need for extensive real-world testing. Real-world experiments confirmed the system's ability to operate effectively in complex industrial settings. To quantify inspection results, the Hausdorff distance was employed to measure deviations between the hook's contour before and after operation. Statistical analysis using quantile ranges divided detection results into confidence intervals: data below 14.9437 was categorized as a confidence match, data within the range [14.9437, 76.4136] was considered a normal match, and data above 76.4136 indicated a failed match or an anomaly requiring attention. This research demonstrates that the proposed system can improve inspection efficiency, enhance defect detection accuracy, and reduce operational downtime, contributing to safer and more reliable industrial crane operations.

Description

Supervisor

Tammi, Kari

Thesis advisor

Ala-Laurinaho, Riku

Keywords

ROS, Raspberry Pi, TurtleBot3 Burger, path planning, computer vision, industrial inspection, Hausdorff distance

Other note

Citation

Permanent link to this item

https://urn.fi/URN:NBN:fi:aalto-202501221588

Collections

[dipl] Insinööritieteiden korkeakoulu / ENG