Optimizing Quality Control with AI-Powered Machine Vision in Manufacturing : A Study of Machine Vision Technology in Manufacturing
El Fassi, Taoufik (2025)
2025
All rights reserved. This publication is copyrighted. You may download, display and print it for Your own personal use. Commercial use is prohibited.
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:amk-2025052817706
https://urn.fi/URN:NBN:fi:amk-2025052817706
Tiivistelmä
This thesis addresses Valmet Automotive’s Vehicle Contract Manufacturing business and investigates how modern quality control processes can be optimized using AI-powered machine vision systems. Although traditional quality control methods are dependable, they often fall short in meeting the high-speed and precision requirements of today’s automotive production lines.
The study primarily explores the implementation of AI-based camera systems, with a specific focus on their deployment at the windshield inspection station. While machine vision tunnel systems are not yet in place at Valmet Automotive Plc, the research examines their feasibility through relevant industry case studies and theoretical insights. Using internal defect data from 2022 to 2024, quantitative data was collected and systematically analyzed to identify critical door-related defect trends. In parallel, an on-site empirical evaluation of the AI camera system was conducted, assessing its real-time functionality, operational performance, and financial implications. Additionally, qualitative methods including expert interviews and structured questionnaires with three external companies provided practical perspectives on the use, benefits, and challenges of machine vision systems.
The findings demonstrate that AI-driven machine vision leads to significant improvements in defect detection accuracy, production efficiency, and cost-effectiveness. These technologies reduce rework, enhance traceability, and enable real-time quality assurance on the production line. A detailed return-on-investment (ROI) analysis confirms the financial viability of such systems. This thesis provides Valmet Automotive Plc and the broader manufacturing industry with a strategic framework for integrating AI-based vision technologies to support efficient, scalable, and reliable quality control processes.
The study primarily explores the implementation of AI-based camera systems, with a specific focus on their deployment at the windshield inspection station. While machine vision tunnel systems are not yet in place at Valmet Automotive Plc, the research examines their feasibility through relevant industry case studies and theoretical insights. Using internal defect data from 2022 to 2024, quantitative data was collected and systematically analyzed to identify critical door-related defect trends. In parallel, an on-site empirical evaluation of the AI camera system was conducted, assessing its real-time functionality, operational performance, and financial implications. Additionally, qualitative methods including expert interviews and structured questionnaires with three external companies provided practical perspectives on the use, benefits, and challenges of machine vision systems.
The findings demonstrate that AI-driven machine vision leads to significant improvements in defect detection accuracy, production efficiency, and cost-effectiveness. These technologies reduce rework, enhance traceability, and enable real-time quality assurance on the production line. A detailed return-on-investment (ROI) analysis confirms the financial viability of such systems. This thesis provides Valmet Automotive Plc and the broader manufacturing industry with a strategic framework for integrating AI-based vision technologies to support efficient, scalable, and reliable quality control processes.