Causal network inference and functional decomposition for decentralized statistical process monitoring: Detection and diagnosis

•Two new causal-based approaches for conducting multi-level decentralized monitoring are proposed: CNET-C and CNET-D.•Both macro-causality (between communities) and micro-causality (within communities) are explored for fault diagnosis.•Markov-blankets allowed for the analysis of macro-causality.•CNET-C and CNET-D always presented higher sensitivity than the benchmark methodologies.•CNET-C tends to perform better in process and correlation faults, whereas CNET-D is generally better for sensor faults. We propose a new systematic approach for conducting decentralized SPM based on the functional decomposition of the system’s causal network. The methodology consists of first inferring the causal network from normal operating data of the system under study, after which the functional modules are identified by exploring the graph topology and finding the strongly connected “communities”. The interaction between functional modules is also taken into account (macro-causality), by extending the original communities with the Markov-blankets of the connection nodes, giving rise to “extended communities”. Two hierarchical monitoring schemes are proposed for distributed monitoring: CNET-C (Causal Network-Centralized) and CNET-D (Causal Network-Distributed). Results demonstrate the increased sensitivity in fault detection of the proposed methodologies compared to conventional non-causal methods and centralized causal methods that monitor the complete network. The proposed approaches also lead to a more effective, unambiguous, and conclusive fault diagnosis activity.