Abstract

PurposeTo avoid the high cost of poor quality (COPQ), there is a constant need for minimizing the formation of defects during manufacturing through defect detection and process parameters optimization. This research aims to develop, design and test a smart system that detects defects, categorizes them and uses this knowledge to enhance the quality of subsequent parts.Design/methodology/approachThe proposed system integrates data collected from the deep learning module with the machine learning module to develop and improve two regression models. One determines if set process parameters would yield a defective product while the second model optimizes them. The deep learning model utilizes final product images to categorize the part as defective or not and determines the type of defect based on image analysis. The developed framework of the system was applied to the forging process to determine its feasibility during actual manufacturing.FindingsResults reveal that implementation of such a smart process would lead to significant contributions in enhancing manufacturing processes through higher production rates of acceptable products and lower scrap rates or rework. The role of machine learning is evident due to numerous benefits which include improving the accuracy of the regression model prediction. This artificial intelligent system enhances itself by learning which process parameters could lead to a defective product and uses this knowledge to adjust the process parameters accordingly overriding any manual setting.Research limitations/implicationsThe proposed system was applied only to the forging process but could be extended to other manufacturing processes.Originality/valueThis paper studies how an artificial intelligent (AI) system can be developed and used to enhance the yield of good products.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call