TY - JOUR
T1 - Deep Learning-based System for Quality Control of Coatings in Recess Punch Manufacturing
AU - Turcsanyi, Balint Newton
AU - Saeed, Faisal
AU - Cooper, Emmett
PY - 2023/8/17
Y1 - 2023/8/17
N2 - Increasing efficiency of the quality inspection process is an on-going pursuit in all manufacturing-related industries. The research was proposed by Tooling international ltd ? a company situated in the UK ? in an attempt to solve a decade-long problem faced when undertaking quality inspection of their coated products. The main objective of this research is to develop a model that detects faulty products with unsatisfactory coating. In this re-search, several convolutional neural network (CNN) architectures were tested in order to find the most suitable one for this particular task. The best performing CNN model delivered 97.68% accuracy which exceed-ed the company?s requirements, providing superior accuracy to when com-pared to current company methods. This study will be used to develop an automated quality inspection machine, thus enhancing the company?s productivity, and will potentially be used as the foundation of further AI-based developments in similar manufacturing-related tasks.
AB - Increasing efficiency of the quality inspection process is an on-going pursuit in all manufacturing-related industries. The research was proposed by Tooling international ltd ? a company situated in the UK ? in an attempt to solve a decade-long problem faced when undertaking quality inspection of their coated products. The main objective of this research is to develop a model that detects faulty products with unsatisfactory coating. In this re-search, several convolutional neural network (CNN) architectures were tested in order to find the most suitable one for this particular task. The best performing CNN model delivered 97.68% accuracy which exceed-ed the company?s requirements, providing superior accuracy to when com-pared to current company methods. This study will be used to develop an automated quality inspection machine, thus enhancing the company?s productivity, and will potentially be used as the foundation of further AI-based developments in similar manufacturing-related tasks.
UR - http://www.open-access.bcu.ac.uk/13964/
U2 - 10.1080/08982112.2021.2001828
DO - 10.1080/08982112.2021.2001828
M3 - Article
SN - 2367-4520
JO - ICACIn 2022
JF - ICACIn 2022
ER -