Anomaly detection in shop floor material flow: A network theory approach

Anomaly detection in material flow refers to the detection of irregular patterns in shop floor behaviour. As anomalies often translate to harmful effects on a larger scale, it is important to detect them in due time. In the paper, the manufacturing shop floor is considered as a dynamic network with work systems acting as its nodes and with material flow creating connections between them. Regularities and irregularities are defined as specific patterns within the network. The approach is studied through simulation and applied to an industrial case study. The results suggest potential use for decision making in production management.