Edge-enabled cloud computing management platform for smart manufacturing
Jeffrey Ying, Jackie Hsieh, Dennis Hou, Janpu Hou, Tuo Liu, Xiaobin Zhang, Yuxi Wang, Yen-Ting Pan
Abstract
The progress on intelligent edge and intelligent cloud has made manufacturing company much more autonomy. The edge device and the public cloud provider become a new hybrid to enable effective collaboration between suppliers, manufacturers and industrial customers. This study proposed a computing management platform required in implementing lean manufacturing system for a manufacturing company. The platform can be used to form an integrated management system based on the quality management system. The model has three modules. The edge metrology module used subspace learning model for defect detection and quality control. The edge production module used stacked recurrent neural network(S-RNN) for continuous process control. The cloud manufacturing management module used convolutional neural network (CNN), recurrent neural network (RNN) and attention model to provide individual customer production and delivery schedule. All the manufacturing issues are managed by edge computing, only the latent representation are uploaded to the cloud for AI-assisted decision making. The proposed platform has been tested on a semiconductor chip production line to provide AI-based defect detection, preventive maintenance scheduling, continuous production control, production optimization and forecasting.