A Monitoring System for Anomaly Detection in Fog Manufacturing

Fog manufacturing is an advanced manufacturing system that integrates Fog computing with interconnected manufacturing processes. It can provide responsive and robust computation services compared with Cloud manufacturing. However, the performance of Fog manufacturing is still not satisfactory due to various anomalies such as inappropriate architecture designs, cyber-attacks, etc. These problems can lead to anomalies in Fog manufacturing and can impede the wide adoption of Fog manufacturing. To quickly detect the anomalies in Fog manufacturing during operations, a monitoring system is desirable. Therefore, in this paper, we proposed a monitoring system based on statistical process control (SPC) methods to comprehensively monitor different types of anomalies in Fog manufacturing. The exponentially weighted moving average (EWMA) and risk-adjusted (RA) charts are adopted to detect different anomalies based on corresponding monitoring statistics. A simulation study and a real case study based on a Fog manufacturing testbed are implemented to evaluate the performance of the proposed monitoring system. The result shows the proposed monitoring system can quickly and efficiently detect different types of anomalies in Fog manufacturing. This monitoring system can also potentially help to improve architecture designs and offloading strategies by further analyzing the root causes of a specific anomaly in future work.