Automated vision system to improve inspection lead time for molded components based on neural network

• Main Author: Nandagopal, Darvind Kumar
• Format: Thesis
• Language: English; English
• Published: 2023
• Online Access: http://eprints.utem.edu.my/id/eprint/27359/1/Automated%20vision%20system%20to%20improve%20inspection%20lead%20time%20for%20molded%20components%20based%20on%20neural%20network.pdf; http://eprints.utem.edu.my/id/eprint/27359/2/Automated%20vision%20system%20to%20improve%20inspection%20lead%20time%20for%20molded%20components%20based%20on%20neural%20network.pdf; http://eprints.utem.edu.my/id/eprint/27359/; https://plh.utem.edu.my/cgi-bin/koha/opac-detail.pl?biblionumber=123199

This research studies on the improvement on the traditional visual inspection method where in this thesis the manual visual inspection system will be improved to an automated visual inspection system with the aid of neural-network.This is to avoid molding defects to escape to customers and reduce the production lead time.This is because the manual visual inspection is prone to human errors.Therefore, there is a need of intellingent visual inspection method which reduces the reject escapee and also improves on the lead time.The objective of this thesis is to detect the critical defects that could be inspected using automated visual inspection. Moreover, this developed system should be able to detect defect on the molded package. In addition to that, the performance of this developed system is evaluated based on the reduction on the lead time and the effectiveness of the defect detection.Thus, the neural network method is used for defect detection and also to automated the system which aids in improving the lead time.The developed system achieves the objectives by reducing escapee by eliminating outgoing quality assurance detection which means 0 customer complain on molding defects. This also proves that the developed system have 0 escape rate to customers.This system also reduces the lead time from 57 minuted to 32 minutes which means 43.86% reduction on lead time.Moreover, the cost saving for 5 years for the developed system is RM 157500.00.Therefore, as a conclusion the developed system achieves the objectives set by able to identify critical defects which it is able to detect the defects for molded package which improves the visual system effectiveness and also reduces the lead time.