Using P&IDs and Process Data To Construct Bayesian Networks For Anomaly Detection

Jakob Ø. Hansen, Anders Læsø Madsen, Thomas Dyhre Nielsen, Joseph Mietkiewicz, Gabriele Baldissone, Micaela Demichela
Proceedings of the 13th International Conference on Probabilistic Graphical Models, PMLR 343:363-378, 2026.

Abstract

Industrial processes are often safety-critical and consist of multiple interacting components. The complexity of such systems, combined with the scarcity of faulty-operation data, makes the construction of reliable anomaly detection models challenging. To address this issue, we investigate how qualitative information in technical process documentation - specifically Piping and Instrumentation Diagrams (P&IDs) - can guide the construction of Bayesian networks for anomaly detection. P&IDs describe the instrumentation and piping of industrial systems, including the placement of sensors and actuators and their interconnections. We show how this structural information can be exploited to decompose an industrial process into smaller subsystems for which local models can be specified or learned independently. For anomaly detection, we employ a conflict-based measure that quantifies inconsistencies in observed process data. We adapt existing threshold calibration and smoothing concepts to the conflict-measure setting, introducing three sequential filters that balance false-alarm frequency against detection latency for true anomalies. The proposed methodology has been integrated into PredictIT, a predictive maintenance platform for industrial processes and systems. Experiments on a simulated formaldehyde production plant demonstrate that P&ID-informed models outperform corresponding full-system models, yielding lower false positive rates and reduced detection latency.

Cite this Paper

BibTeX
@InProceedings{pmlr-v343-hansen26a, title = {Using P&IDs and Process Data To Construct {Bayesian} Networks For Anomaly Detection}, author = {Hansen, Jakob {\O}. and Madsen, Anders L{\ae}s{\o} and Nielsen, Thomas Dyhre and Mietkiewicz, Joseph and Baldissone, Gabriele and Demichela, Micaela}, booktitle = {Proceedings of the 13th International Conference on Probabilistic Graphical Models}, pages = {363--378}, year = {2026}, editor = {Camps-Valls, Gustau and Leonelli, Manuele and Varando, Gherardo}, volume = {343}, series = {Proceedings of Machine Learning Research}, month = {09--11 Sep}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v343/main/assets/hansen26a/hansen26a.pdf}, url = {https://proceedings.mlr.press/v343/hansen26a.html}, abstract = {Industrial processes are often safety-critical and consist of multiple interacting components. The complexity of such systems, combined with the scarcity of faulty-operation data, makes the construction of reliable anomaly detection models challenging. To address this issue, we investigate how qualitative information in technical process documentation - specifically Piping and Instrumentation Diagrams (P&IDs) - can guide the construction of Bayesian networks for anomaly detection. P&IDs describe the instrumentation and piping of industrial systems, including the placement of sensors and actuators and their interconnections. We show how this structural information can be exploited to decompose an industrial process into smaller subsystems for which local models can be specified or learned independently. For anomaly detection, we employ a conflict-based measure that quantifies inconsistencies in observed process data. We adapt existing threshold calibration and smoothing concepts to the conflict-measure setting, introducing three sequential filters that balance false-alarm frequency against detection latency for true anomalies. The proposed methodology has been integrated into PredictIT, a predictive maintenance platform for industrial processes and systems. Experiments on a simulated formaldehyde production plant demonstrate that P&ID-informed models outperform corresponding full-system models, yielding lower false positive rates and reduced detection latency.} }
Endnote
%0 Conference Paper %T Using P&IDs and Process Data To Construct Bayesian Networks For Anomaly Detection %A Jakob Ø. Hansen %A Anders Læsø Madsen %A Thomas Dyhre Nielsen %A Joseph Mietkiewicz %A Gabriele Baldissone %A Micaela Demichela %B Proceedings of the 13th International Conference on Probabilistic Graphical Models %C Proceedings of Machine Learning Research %D 2026 %E Gustau Camps-Valls %E Manuele Leonelli %E Gherardo Varando %F pmlr-v343-hansen26a %I PMLR %P 363--378 %U https://proceedings.mlr.press/v343/hansen26a.html %V 343 %X Industrial processes are often safety-critical and consist of multiple interacting components. The complexity of such systems, combined with the scarcity of faulty-operation data, makes the construction of reliable anomaly detection models challenging. To address this issue, we investigate how qualitative information in technical process documentation - specifically Piping and Instrumentation Diagrams (P&IDs) - can guide the construction of Bayesian networks for anomaly detection. P&IDs describe the instrumentation and piping of industrial systems, including the placement of sensors and actuators and their interconnections. We show how this structural information can be exploited to decompose an industrial process into smaller subsystems for which local models can be specified or learned independently. For anomaly detection, we employ a conflict-based measure that quantifies inconsistencies in observed process data. We adapt existing threshold calibration and smoothing concepts to the conflict-measure setting, introducing three sequential filters that balance false-alarm frequency against detection latency for true anomalies. The proposed methodology has been integrated into PredictIT, a predictive maintenance platform for industrial processes and systems. Experiments on a simulated formaldehyde production plant demonstrate that P&ID-informed models outperform corresponding full-system models, yielding lower false positive rates and reduced detection latency.
APA
Hansen, J.Ø., Madsen, A.L., Nielsen, T.D., Mietkiewicz, J., Baldissone, G. & Demichela, M.. (2026). Using P&IDs and Process Data To Construct Bayesian Networks For Anomaly Detection. Proceedings of the 13th International Conference on Probabilistic Graphical Models, in Proceedings of Machine Learning Research 343:363-378 Available from https://proceedings.mlr.press/v343/hansen26a.html.

Related Material