Automatic Fault Detection in Batch Distillation Using Machine Learning

My research is part of DFG Research Group 5359, which aims to apply deep learning methods to chemical process data, particularly to automatically identify operational errors in chemical plants at an early stage. A key challenge here is that deep learning methods require large amounts of training data, but chemical process data has been extremely scarce to date.

To solve the data problem, three strategies are being pursued: First, large amounts of experimental data will be generated by studying a wide variety of material systems in a laboratory-scale batch distillation plant equipped with a comprehensive sensor system. Second, physical process simulations will be developed and utilized in collaboration with Fraunhofer ITWM to generate additional data. Third, in collaboration with the Machine Learning research group, we plan to develop hybrid generative models to significantly expand the datasets synthetically. Furthermore, I am interested in addressing uncertainties and incomplete knowledge in process data, as well as in how previously unused process variables can be incorporated into process simulations.

Information on possible topics for student projects with me or my colleagues are available at:  https://mv.rptu.de/fgs/ltd/jobs/studentische-arbeiten

Lecture Support

• ChemCar-Competition I (WS 2022/2023)
• Machine Learning in Process Engineering (SoSe 2023)
• Process Thermodynamics (WS 2023/2024)
• Electrochemical Energystorage (SoSe2024)

Office Hours

By arrangement.

Publications

• T. Specht, J. Arweiler, J. Stüber, K. Münnemann, H. Hasse, F. Jirasek: Automated Nuclear Magnetic Resonance Fingerprinting of Mixtures, Magn. Reson. Chem.1 (2023), 1-12, doi: 10.1002/mrc.5381.
• N. Hayer, T. Specht, J. Arweiler, D. Gond, H. Hasse, F. Jirasek: Prediction of activity coefficients by similarity-based imputation using quantum-chemical descriptors, Phys. Chem. Chem. Phys. 27 (2025) 4307-4315. doi: 10.1039/D4CP04341C.
• N. Hayer, T. Specht, J. Arweiler, H. Hasse, F. Jirasek: Similarity-Informed Matrix Completion Method for Predicting Activity Coefficients, J. Phys. Chem. 129 (2025) 3141-3147, doi: 10.1021/acs.jpca.4c08360.
• D. Wagner, F. Hartung, J. Arweiler, A. Muraleedharan, I. Jungjohann, A. Nair, S. Reithermann, R. Schulz, M. Bortz, D. Neider, H. Leitte, J. Pfeffinger, S. Mandt, S. Fellenz, T. Katz, F. Jirasek, J. Burger, H. Hasse, M. Kloft: NoBOOM: Chemical Process Datasets for Industrial Anomaly Detection, In: Advances in Neural Information Processing Systems 38 (2025).
• J. Arweiler, I. Jungjohann, A. Muraleedharan, H. Leitte, J. Burger, K. Münnemann, F. Jirasek, H. Hasse: Batch Distillation Data for Developing Machine Learning Anomaly Detection Methods, Sci. Data 13, 513 (2026). doi:10.1038/s41597-026-07124-3.
• D. Gond, J. Arweiler, T. Specht, H. Hasse, F. Jirasek: CHAOS - A Large-scale Database for sigma-Profiles and Other Molecular Descriptors, J. Chem. Inf. Model. 7, 66 (2026) 3969-3975. doi: 10.1021/acs.jcim.6c00058.
• S. Lutz, J. Arweiler, A. Muraleedharan, N. Kahlhoff, F. Hartung, I. Jungjohann, M. Nagda, D. Reinhardt, D. Wagner, J. Werner, J. Will, J. Burger, M. Bortz, H. Hasse, S. Fellenz, F. Jirasek, M. Kloft, H. Leitte, S. Mandt, S. Reithermann, J. Schmid, D. Neider: A Benchmark Suite for Verifying Neural Anomaly Detectors in Distillation Processes, In: Springer Machine Learning and Principles and Practice of Knowledge Discovery in Databases. ECML PKDD 2024. Communications in Computer and Information Science 2559 (2026) 290-306. doi:10.1007/978-3-032-25305-7_20.
• M. Nagda, P. Ostheimer, J. Arweiler, I. Jungjohann, J. Werner, D. Wagner, A. Muraleedharan, P. Jafari, J. Schmid, F. Jirasek, J. Burger, M. Bortz, H. Hasse, S. Mandt, M. Kloft, S. Fellenz: Style Transfer for High-Fidelity Time Series Augmentation, In: Springer Machine Learning and Principles and Practice of Knowledge Discovery in Databases. ECML PKDD 2025. Communications in Computer and Information Science 2841 (2026) 295-316. doi:10.1007/978-3-032-19102-1_18.

Presentations

• N. Hayer, T. Specht, J. Arweiler, H. Hasse, F. Jirasek: Prediction of Activitiy Coefficients by Similarity-based Imputation using Quantum-Chemical Descriptors, Thermodynamik-Kolloquium, Chemnitz, 26.-28.09.2022.
• J. Arweiler, N. Hayer, T. Specht, H. Hasse, F. Jirasek: Similarity-based Prediction of Activity Coefficients, Verfahrenstechnisches Doktorandenseminar der RPTU Kaiserslautern, Annweiler am Trifels, 21.-22.08.2023.
• J. Arweiler, N. Hayer, T. Specht, H. Hasse, F. Jirasek: Similarity-based Prediction of Activity Coefficients, European Conference on Thermophysical Properties, Venice, Italy, 10.-13.09.2023.
• J. Arweiler, N. Hayer, T. Specht, H. Hasse, F. Jirasek: Prediction of Activity Coefficients with Similarity-based Imputation using Quantum-chemical Descriptors and Machine Learning, 14th European Congress of Chemical Engineering , Berlin, 17.-21.09.2023.
• A. Muraleedharan, J. Burger, J. Arweiler, F. Hartung, D. Wagner, F. Jirasek, M. Kloft, H. Hasse, M. Bortz: Deep Learning on Sparse Chemical Process Data, Jahrestreffen Prozess., Apparate- und Anlagentechnik (PAAT), Frankfurt a. M., 20.-21.11.2023.
• T. Specht, J. Arweiler, Z. Romero, J. Wagner, K. Münnemann, H. Hasse, F. Jirasek: Thermodynamic Modeling of Poorly Specified Mixtures using NMR Spectroscopy and Machine Learning, 33rd European Symposium on Applied Thermodynamic, Edinburgh, Scotland, 09.-12.06.2024.
• J. Arweiler, H. Hasse, F. Jirasek: Phase-Equilibria of Renewable Fuel Blends, Water and Additives, 22nd Symposium on Thermophysical Properties, Boulder, USA, 23.-28.06.2024.
• J. Wagner, J. Arweiler, Z. Romero, T. Specht, K. Münnemann, H. Hasse, F. Jirasek: Automated NMR Fingerprinting of Poorly Specified Mixtures with Machine Learning, qNMR Summit, Buchs SG, Schweiz, 18.-20.09.2024.
• J. Arweiler, A. Muraleedharan, I. Jungjohann, F. Hartung, D. Wagner, J. Burger, M. Kloft, F. Jirasek, H. Hasse: Machine Learning for Anomaly Detection in Distillation Processes, Internationaler Workshop "Machine Learning in Chemical Processes" der DFG FOR5359, Kaiserslautern, 09.10.2024.
• J. Arweiler, I. Jungjohann, A. Muraleedharan, F. Hartung, D. Wagner, M. Kloft, J. Burger, F. Jirasek, H. Hasse: Machine Learning for Anomaly Detection in Distillation Processes, Annual Meeting of Process Engineering and Materials Technology (PEMT), Frankfurt am Main, 11.11. - 12.11.2024.
• J. Arweiler, I. Jungjohann, J. Werner, J. Schmid, D. Wagner, F. Hartung, H. Leitte, M. Kloft, M. Bortz, F. Jirasek, H. Hasse : A Dataset of Normal and Anomalous Batch Distillation Data for Anomaly Detection, Joint Annual Meeting of the DECHEMA/VDI Specialist Groups on “Fluid Process Engineering” with “Heat and Mass Transfer”, Luzern, Switzerland, 09.03.-11.03.2026.
• J. Arweiler, I. Jungjohann, J. Werner, J. Schmid, D. Wagner, F. Hartung, M. Kloft, H. Leitte, M. Bortz, F. Jirasek, H. Hasse: Anomaly Detection in Batch Distillation with Machine Learning – Influence of Data Preprocessing and Augmentation, European Symposium on Computer Aided Process Engineering (ESCAPE36), Sheffield, UK, 21.06.-24.06.2026.

Poster

• J. Arweiler, T. Specht, K. Münnemann, H. Hasse, F. Jirasek: SMART NMR Fingerprinting of Mixtures using Benchtop Spectrometers, NMRPM 2023, Kaiserslautern, 24.-26.05.2023.
• J. Arweiler, T. Specht, K. Münnemann, H. Hasse, F. Jirasek: Automated NMR Fingerprinting and Thermodynamic Modeling of Poorly Specified Mixtures, Thermodynamik-Kolloquium, Hannover, 25.-27.09.2023.
• J. Arweiler, A. Muraleedharan, F. Hartung, I. Jungjohann, D. Wagner, M. Kloft, J. Burger, H. Hans, F. Jirasek: Anomaly Detection on Experimental Chemical Process Data, Workshop on Machine Learning for Chemistry and Chemical Enginnering at European Conference on Machine Learning, Vilnius, Lithuania, 09.09.2024.
• D. Reinhardt, D. Wagner, A. Muraleedharan, J. Arweiler, I. Jungjohann, F. Jirasek, J. Burger, H. Hasse, M. Kloft, H. Leitte: cPAX: Comparative Visualization of Known and Novel Anomalies for Monitoring Chemical Plants, Workshop on Machine Learning for Chemistry and Chemical Enginnering at European Conference on Machine Learning, Vilnius, Lithuania, 09.09.2024.
• S. Lutz, J. Arweiler, A. Muraleedharan, N. Kahlhoff, F. Hartung, I. Jungjohann, M. Nagda, D. Reinhardt, D. Wagner, J. Werner, J. Will, J. Burger, M. Bortz, H. Hasse, S. Fellenz, F. Jirasek, M. Kloft, H. Leitte, S. Mandt, S. Reithermann, J. Schmid, D. Neider: A Benchmark Suite for Verifying Neural Anomaly Detectors in Distillation Processes, Workshop on Machine Learning for Chemistry and Chemical Enginnering at European Conference on Machine Learning, Vilnius, Lithuania, 09.09.2024.
• M. J. Peter, N. Ghanooni, F. Hartung, B. J. Franks, D. Wagner, P. Liznerski, A. Muraleedharan, J. Arweiler, D. Reinhardt, I. Jungjohann, S. Reithermann, S. Fellenz, F. Jirasek, M. Rudolph, D. Neider, F. Rhein, H. Leitte, C. Song, S. Mandt, M. Bortz, J. Burger, H. Hasse, M. Kloft: Anomaly Classification of Tennessee-Eastman Process Data, Workshop on Machine Learning for Chemistry and Chemical Enginnering at European Conference on Machine Learning, Vilnius, Lithuania, 09.09.2024.
• J. Will, J. Arweiler, I. Jungjohann, J. Werner, M. Nagda, M. Bortz, H. Hasse, F. Jirasek, J. Schmid, M. Kloft, S. Mandt, S. Fellenz: Enhancing Realism in Batch Distillation Simulations: Data-Efficient Time Series Style Transfer with Transformers, Workshop on Machine Learning for Chemistry and Chemical Enginnering at European Conference on Machine Learning, Vilnius, Lithuania, 09.09.2024.
• J. Arweiler, I. Jungjohann, A. Muraleedharan, F. Hartung, D. Wagner, M. Kloft, J. Burger, F. Jirasek, H. Hasse: Anomaly Detection on Chemical Process Data, Thermodynamik-Kolloqium, Stuttgart, 25.-27.09.2024.
• I. Jungjohann, J. Arweiler, J. Werner, A. Muraleedharan, H. Leitte, J. Burger, M. Bortz, F. Jirasek, H. Hasse: Batch Distillation Data for Deep Anomaly Detection, Thermodynamik-Kolloqium, Hamburg, 22-24.09.2025.
• J. Arweiler, F. Jirasek, H. Hasse: Phase-Equilibria of Renewable Fuel Blends, Water, and Additives using NMR-Fingerprinting, Thermodynamik-Kolloqium, Hamburg, 22-24.09.2025.
• D. Wagner, F. Hartung, J. Arweiler, A. Muraleedharan, I. Jungjohann, A. Nair, S. Reithermann, R. Schulz, M. Bortz, D. Neider, H. Leitte, J. Pfeffinger, S. Mandt, S. Fellenz, T. Katz, F. Jirasek, J. Burger, H. Hasse, M. Kloft: NoBOOM: Chemical Process Datasets for Industrial Anomaly Detection, The Thirty-ninth Annual Conference on Neural Information Processing Systems Datasets and Benchmarks Track, San Diego, USA, 02.-07.12.2025.