Object Detection using Convolutional Neural Networks for Smart Manufacturing Vision Systems in the Medical Devices Sector

Abstract

Industry 4.0 has opened the doors for Deep Learning to enter into the manufacturing arena with a bid to improve efficiency and quality check process. In many assembly lines Vision Systems are applied that can identify anomalies, read labels, count components and such like. However these systems are sensitive to lighting and setup conditions, and in many cases the technology is unable to read or classify, leaving gaps in the assembly process where human validation is a necessity. A typically manufacturing response is to add further quality control check layers onto the backend of the process. An ideal Industry 4.0 Smart Manufacturing vision system would keep track of components being used, identify anomalies and identify processes successfully during the production stage providing efficient quality checks in real-time, thus creating a more efficient Quality Control process, and move closer to Zero-Defect scenario.

One area in which Vision Systems are rarely used is the medical technology sector, due to the high standards required to approve a line. Because current Vision Systems can fail in different setup conditions, this makes them a risk and so, Quality Control is not in any way aided or improved upon. This study examines the application of Deep Learning with neural networks on components from a medical technology company, to demonstrate how they can be used as a more reliable and less prone to error vision system, that can track the components in real time regardless of lighting conditions and other constraints and perform other Quality Control checks.