SharkNet Networks Applications in Smart Manufacturing Using IoT and Machine Learning

Abstract

With the advancement of Industry 4.0, 3D printing has become a critical technology in smart manufacturing; however, challenges remain in the integrated management, quality control, and remote monitoring of multiple 3D printers. This study proposes an intelligent cloud monitoring system based on the SharkNet dynamic network, IoT, and artificial neural networks (ANNs). The system utilizes a SharkNet dynamic network to integrate low-cost sensors for environmental monitoring to enable low-latency data transmission and deploys ANN models on the cloud for print quality prediction and process parameter optimization. Next, we experimentally validated the system using the Taguchi design and ANN-based analysis, focusing on optimizing printing process parameters and improving surface quality. The main results show that the designed system has a communication delay of 40–50 ms and 99.8% transmission reliability under moderate load, and the system reduces the surface roughness prediction error to less than 17.2%. In addition, the ANN model outperforms conventional methods in capturing the nonlinear relationships of the variables, and the system can be based on the model to improve print quality and productivity by enabling real-time parameter adjustments. The system retains a high degree of scalability in terms of real-time monitoring and parallel or complex control of multiple devices, which demonstrates its potential for applications in smart manufacturing.