• German
German

Main Navigation

B3  Data Mining on Sensor Data of Automated Processes


Deuse.jpg
Prof. Dr. Deuse, Jochen
Morik.JPG
Prof. Dr. Morik, Katharina

Enhancing the control of industrial processes and the quality of products can be supported by learning from sensor data. Project B3 focuses on the investigation of how decentralised data mining can be used for real-time quality predictions and how it can be integrated into production processes.

The tasks of data analytics in production systems have been ordered according to their difficulty: (1) anomaly detection, (2) diagnostic analytics, (3) predictive analytics, and (4) prescriptive analytics. In the first phase, we succeeded in developing a new decentralised anomaly detection and invented learning from label proportions. In the second phase, we investigated diagnostic and predictive analytics. Diagnostic analytics aims at explaining not ok products by process data. Based on time series data from a hot rolling process for steel-bars production, features have been extracted and aggregated that help to distinguish between ok and not ok products. Modeling the overall preprocessing in the software RapidMiner took the toolbox to its extreme, becoming a high-level programming environment. The advantage is the reproducibility of RapidMiner processes and their understandable documentation. Predictive analytics in flexible production processes has contributed to adequate process control in real-time. For distributed data mining, the training of local models from counts (TLMC) for vertically partitioned data has been developed. The anomaly detection has been enhanced for distributed settings as they are given by the Internet of Things. In the third phase, prescriptive analytics will be the focus of the project; i.e., the learned model changes the process in real-time. An example is not only to recognise a deviation of sensor measurements from the normal, but also to adapt the process parameters accordingly.

Project management:

Prof. Dr. Deuse, Jochen
Prof. Dr. Morik, Katharina
Prof. Dr.-Ing. Wiederkehr, Petra

Project members:

Amal, Saadallah
Schmitt, Jacqueline
Wiegand, Mario

Alumni:

Blom, Hendrik
Dr. Bohnen, Fabian
Dr. Erohin, Olga
Dr. Konrad, Benedikt
Dr. Stolpe, Marco

Software:

LLP

Publications:

Saadallah/etal/2018a Saadallah, Amal and Finkeldey, Felix and Morik, Katharina and Wiederkehr, Petra. Stability prediction in milling processes using a simulation-based machine learning approach. In 51st CIRP conference on Manufacturing Systems, Elsevier, 2018.


Schmitt/etal/2018a Schmitt, Jacqueline and Wiegand, Mario and Deuse, Jochen. Qualitätsbasierte Auftragszuordnung - Zuordnung von Zwischenprodukten zu Kundenaufträgen auf Basis von Qualitätsprognosen. In ZWF online, 2018.


Schmitt/etal/2018b Schmitt, Jacqueline and Hahn, Florian and Deuse, Jochen. Mathematical modelling of the quality-based order assignment problem. No. 2, Institute of Production Systems, TU Dortmund University, 2018.


Deuse/etal/2017a Deuse, J. and Schmitt, J. and Stolpe, M. and Wiegand, M. and Morik, K.. Qualitätsprognosen zur Engpassentlastung in der Injektorfertigung unter Einsatz von Data Mining. In Schriftenreihe der Wissenschaftlichen Gesellschaft für Arbeits- und Betriebsorganisation (WGAB) e.V., 2017.


Blom/Morik/2016a Blom, Hendrik and Morik, Katharina. Resource-Aware Steel Production Through Data Mining. In "Berendt, Bettina and Bringmann, Bj\"orn and Fromont, \'Elisa and Garriga, Gemma and Miettinen, Pauli and Tatti, Nikolaj and Tresp, Volker" (editors), Machine Learning and Knowledge Discovery in Databases, pages 263--266, Springer, 2016. year = {2016},
pages = {263--266},
publisher = {Springer International Publishing},
url = {https://link.springer.com/chapter/10.1007/978-3-319-46131-1_31},
abstract = {Today’s steel industry is characterized by overcapacity and increasing competitive pressure. There is a need for continuously improving processes, with a focus on consistent enhancement of efficiency, improvement of quality and thereby better competitiveness. About 70 % of steel is produced using the BF-BOF (Blast Furnace - Blow Oxygen Furnace) route worldwide. The BOF is the first step of controlling the composition of the steel and has an impact on all further processing steps and the overall quality of the end product. Multiple sources of process-related variance and overall harsh conditions for sensors and automation systems in general lead to a process complexity that is not easy to model with thermodynamic or metallurgical approaches. In this paper we want to give an insight how to improve the output quality with machine learning based modeling and which constraints and requirements are necessary for an online application in real-time.}
}')">


Stolpe/2016a Marco Stolpe. The Internet of Things: Opportunities and Challenges for Distributed Data Analysis. In SIGKDD Explorations, Vol. 18, No. 1, pages 15-34, 2016.


Stolpe/etal/2016a Stolpe, Marco and Blom, Hendrik and Morik, Katharina. Sustainable Industrial Processes by Embedded Real-Time Quality Prediction. In Computational Sustainability, pages 201--243, Springer, 2016.


Stolpe/etal/2016b Marco Stolpe and Kanishka Bhaduri and Kamalika Das. Distributed Support Vector Machines: An Overview. In Michaelis, S. and Piatkowski, N. and Stolpe, M. (editors), Solving Large Scale Learning Tasks: Challenges and Algorithms, Vol. 9580, pages 109--138, Springer, 2016.


Wiegand/etal/2016a Wiegand, Mario and Stolpe, Marco and Deuse, Jochen and Morik, Katharina. Prädiktive Prozessüberwachung auf Basis verteilt erfasster Sensordaten. In at-Automatisierungstechnik, Vol. 64, No. 7, pages 521--533, 2016.


Eickelmann/etal/2015a Eickelmann, Michel and Wiegand, Mario and Konrad, Benedikt and Deuse, Jochen. Die Bedeutung von Data Mining im Kontext von Industrie 4.0. In Zeitschrift für wirtschaftlichen Fabrikbetrieb (ZWF), Vol. 110, No. 11, pages 738-743, 2015.


Liebig/etal/2015a Liebig, Thomas and Stolpe, Marco and Morik, Katharina. Distributed Traffic Flow Prediction with Label Proportions: From in-Network towards High Performance Computation with MPI. In Proceedings of the 2nd International Workshop on Mining Urban Data (MUD2), Vol. 1392, pages 36--43, CEUR-WS, 2015.


Stolpe/etal/2015a Marco Stolpe and Thomas Liebig and Katharina Morik. Communication-efficient learning of traffic flow in a network of wireless presence sensors. In Proceedings of the Workshop on Parallel and Distributed Computing for Knowledge Discovery in Data Bases (PDCKDD 2015), pages (to appear), CEUR-WS, 2015.


Deuse/etal/2014a Deuse, Jochen and Wiegand, Mario and Erohin, Olga and Lieber, Daniel and Klinkenberg, Ralf. Big Data Analytics in Produktion und Instandhaltung. In Biedermann, Hubert (editors), Instandhaltung im Wandel. Herausforderungen und Lösungen im Zeitalter von Industrie 4.0, pages 33-48, 2014.


Deuse/etal/2014b Deuse, Jochen and Erohin, Olga and Lieber, Daniel. Wissensentdeckung in vernetzten, industriellen Datenbeständen. In Lödding, Hermann (editors), Industrie 4.0. Wie intelligente Vernetzung und kognitive Systeme unsere Arbeit verändern, pages 373-395, Gito, 2014.


Bhaduri/Stolpe/2013a Bhaduri, Kanishka and Stolpe, Marco. Distributed Data Mining in Sensor Networks. In Aggarwal, Charu C. (editors), Managing and Mining Sensor Data, Berlin, Heidelberg, Springer, 2013.


Bohnen/etal/2013a Bohnen, Fabian and Stolpe, Marco and Deuse, Jochen and Morik, Katharina. Using a Clustering Approach with Evolutionary Optimized Attribute Weights to Form Product Families for Production Leveling. In Windt, Katja (editors), Robust Manufacturing Control, pages 189--202, Berlin, Heidelberg, Springer, 2013.


Bohnen/etal/2013b Bohnen, Fabian and Buhl, Matthias and Deuse, Jochen. Systematic Procedure for leveling of low volume and high mix production. In CIRP Journal od Manufacturing Science and Technology, Vol. 6, No. 1, pages 53-58, 2013.


Deuse/etal/2013a Deuse, Jochen and Konrad, Benedikt and Bohnen, Fabian. Renaissance of Group Technology: Reducing Variability to Match Lean Production Prerequisites. In Bakhtadze, Natalia and Chernyshov, Kirill and Dolgui, Alexandre and Lototsky, Vladimir (editors), Manufacturing Modelling, Management, and Control, Vol. 7, pages 998-1003, 2013.


Konrad/etal/2013a Konrad, Benedikt and Lieber, Daniel and Deuse, Jochen. Striving for Zero Defect Production: Intelligent Manufacturing Control through Data Mining in Continuous Rolling Mill Processes. In Windt, Katja (editors), Robust Manufacturing Control, pages 215--229, CIRP, Berlin, Heidelberg, Springer, 2013.


Lieber/etal/2013a Lieber, Daniel and Stolpe, Marco and Konrad, Benedikt and Deuse, Jochen and Morik, Katharina. Quality Prediction in Interlinked Manufacturing Processes based on Supervised & Unsupervised Machine Learning. In Procedia CIRP - 46th CIRP Conf. on Manufacturing Systems, Vol. 7, pages 193-198, Elsevier, 2013.


Lieber/etal/2013b Lieber, Daniel and Erohin, Olga and Deuse, Jochen. Wissensentdeckung im industriellen Kontext - Herausforderungen und Anwendungsbeispiele. In Zeitschrift für wirtschaftlichen Fabrikbetrieb (ZWF), Vol. 108, No. 6, pages 388-393, 2013.


Stolpe/etal/2013a Stolpe, M. and Bhaduri, K. and Das, K. and Morik, K.. Anomaly Detection in Vertically Partitioned Data by Distributed Core Vector Machines. In Blockeel, Hendrik and Kersting, Kristian and Nijssen, Siegfried and \vZelezný, Filip (editors), Machine Learning and Knowledge Discovery in Databases: European Conference, ECML PKDD 2013, Prague, Czech Republic, September 23-27, 2013, Proceedings, Part III, pages 321--336, Springer, 2013.


Lee/etal/2012a Lee, S. and Stolpe, M. and Morik, K.. Separable Approximate Optimization of Support Vector Machines for Distributed Sensing. In Flach, Peter A. and De Bie, Tijland and Cristianini, Nello (editors), Machine Learning and Knowledge Discovery in Databases: European Conference, ECML PKDD 2012, Bristol, UK, September 24-28, 2012. Proceedings, Part II, Vol. 7524, pages 387--402, Springer, 2012.


Lieber/etal/2012a Lieber, Daniel and Konrad, Benedikt and Deuse, Jochen and Stolpe, Marco and Morik, Katharina. Sustainable Interlinked Manufacturing Processes through Real-Time Quality Prediction. In Dornfeld, David A. and Linke, Barbara S. (editors), Leveraging Technology for a Sustainable World, pages 393-398, CIRP, Berlin, Heidelberg, Springer, 2012.


Maschek/etal/2011a Maschek,Thomas and Konrad, Benedikt and Deuse, Jochen and Hermanns, Gerhard and Weber, Daniel and Schreckenberg, Michael. Verkehrsforschung in der Produktionsflussanalyse - Übertragung von Modellen der statistischen Physik auf die Analyse von Produktionssystemen. In ZWF - Zeitung für wirtschaftlichen Fabrikbetrieb, Vol. 106, No. 11, pages 833--837, 2011.


Stolpe/etal/2011a Stolpe, Marco and Morik, Katharina and Konrad, Benedikt and Lieber, Daniel and Deuse, Jochen. Challenges for Data Mining on Sensor Data of Interlinked Processes. In Proceedings of the Next Generation Data Mining Summit (NGDM) 2011, 2011.


Stolpe/Morik/2011a Stolpe, M. and Morik, K.. Learning from Label Proportions by Optimizing Cluster Model Selection. In Gunopulos, Dimitrios and Hofmann, Thomas and Malerba, Donato and Vazirgiannis, Michalis (editors), Machine Learning and Knowledge Discovery in Databases: European Conference, ECML PKDD 2011, Athens, Greece, September 5-9, 2011, Proceedings, Part III, pages 349--364, Springer, 2011.


  • Saadallah/etal/2018a - Stability prediction in milling processes using a simulation-based machine learning approach
  • Schmitt/etal/2018a - Qualitätsbasierte Auftragszuordnung - Zuordnung von Zwischenprodukten zu Kundenaufträgen auf Basis von Qualitätsprognosen
  • Schmitt/etal/2018b - Mathematical modelling of the quality-based order assignment problem
  • Deuse/etal/2017a - Qualitätsprognosen zur Engpassentlastung in der Injektorfertigung unter Einsatz von Data Mining
  • Blom/Morik/2016a - Resource-Aware Steel Production Through Data Mining
  • Stolpe/2016a - The Internet of Things: Opportunities and Challenges for Distributed Data Analysis
  • Stolpe/etal/2016a - Sustainable Industrial Processes by Embedded Real-Time Quality Prediction
  • Stolpe/etal/2016b - Distributed Support Vector Machines: An Overview
  • Wiegand/etal/2016a - Prädiktive Prozessüberwachung auf Basis verteilt erfasster Sensordaten
  • Eickelmann/etal/2015a - Die Bedeutung von Data Mining im Kontext von Industrie 4.0
  • Liebig/etal/2015a - Distributed Traffic Flow Prediction with Label Proportions: From in-Network towards High Performance Computation with MPI
  • Stolpe/etal/2015a - Communication-efficient learning of traffic flow in a network of wireless presence sensors
  • Deuse/etal/2014a - Big Data Analytics in Produktion und Instandhaltung
  • Deuse/etal/2014b - Wissensentdeckung in vernetzten, industriellen Datenbeständen
  • Bhaduri/Stolpe/2013a - Distributed Data Mining in Sensor Networks
  • Bohnen/etal/2013a - Using a Clustering Approach with Evolutionary Optimized Attribute Weights to Form Product Families for Production Leveling
  • Bohnen/etal/2013b - Systematic Procedure for leveling of low volume and high mix production
  • Deuse/etal/2013a - Renaissance of Group Technology: Reducing Variability to Match Lean Production Prerequisites
  • Konrad/etal/2013a - Striving for Zero Defect Production: Intelligent Manufacturing Control through Data Mining in Continuous Rolling Mill Processes
  • Lieber/etal/2013a - Quality Prediction in Interlinked Manufacturing Processes based on Supervised & Unsupervised Machine Learning
  • Lieber/etal/2013b - Wissensentdeckung im industriellen Kontext - Herausforderungen und Anwendungsbeispiele
  • Stolpe/etal/2013a - Anomaly Detection in Vertically Partitioned Data by Distributed Core Vector Machines
  • Lee/etal/2012a - Separable Approximate Optimization of Support Vector Machines for Distributed Sensing
  • Lieber/etal/2012a - Sustainable Interlinked Manufacturing Processes through Real-Time Quality Prediction
  • Maschek/etal/2011a - Verkehrsforschung in der Produktionsflussanalyse - Übertragung von Modellen der statistischen Physik auf die Analyse von Produktionssystemen
  • Stolpe/etal/2011a - Challenges for Data Mining on Sensor Data of Interlinked Processes
  • Stolpe/Morik/2011a - Learning from Label Proportions by Optimizing Cluster Model Selection

Disserations:

  • Lieber/2018a - Data Mining in der Gütesicherung der Stabstahlproduktion
  • Stolpe/2017a - Distributed Analysis of Vertically Partitioned Sensor Measurements under Communication Constraints
  • Erohin/2016a - Wissensgewinnung durch Datenanalyse zur prospektiven Zeitermittlung
  • Bohnen/2013a - Eine Methodik zur Produktionsnivellierung auf der Basis von Fertigungsfamilien

Final Theses:

  • Haritz/2017a - Parameterschätzung mit Gütegarantie durch Bandit Models für die Regelung im Industrie 4.0 Kontext
  • Honysz/2017a - Anomalie-Erkennung in Spritzgieß-Prozessdaten
  • Rickhoff/2016a - Zyklische Concept Drifts
  • Gaertner/2014a - Klassifikation von Zeitreihen über die Bestimmung häufger symbolisierter Subsequenzen
  • Roetner/2014a - Behandlung von Concept Drift in zyklischen Prozessen
  • Koscharnyj/2013a - Beitrag zur Optimierung von Qualitätssicherungsmethoden in der Stahlindustrie auf Basis einer Untersuchung und Bewertung gängiger Qualitätssicherungsstrategien in der Automobil- und Prozessindustrie
  • Matuschek/2013a - Symbolisierung und Clustering von Zeitreihen als neue Operatoren im ValueSeries Plugin von Rapidminer
  • Spain/2013a - A Survey on Subspace Clustering
  • Stumpf/2013a - Untersuchung der Eignung ausgewählter Mustererkennungsverfahren zur Identifizierung qualitätsrelevanter Einflussfaktoren in der Stahlindustrie
  • Blom/2011a - Entwicklung von Optimierungsverfahren für das Lösen verschiedener Lernaufgaben mit der Stützvektormethode

Preliminary Work:

Bohnen/Deuse/2010a Bohnen, F. and Deuse, J.. Leveling of Low Volume and High Mix Production based on a Group Technology Approach. In Proceedings of the 43rd CIRP International Conference on Manufacturing Systems, pages 949--956, 2010.


Morik/etal/2010a Morik, Katharina and Stolpe, Marco and Deuse, Jochen and Bohnen, Fabian and Reichel, Ulrich. Prognosemodelle zur Ermittlung der Produkteigenschaften -- Einsatz von Data-Mining-Verfahren im Walzwerk. In stahl und eisen, No. 10, pages 80--82, 2010.


Morik/etal/2010b Morik, Katharina and Deuse, Jochen and Faber, Vanessa and Bohnen, Fabian. Data Mining in Sensordaten verketteter Prozesse. In Zeitschrift für Wirtschaftlichen Fabrikbetrieb (ZWF), Vol. 105, No. 1-2, pages 106--110, Carl Hanser, 2010.


Mierswa/2008a Mierswa, Ingo. Non-Convex and Multi-Objective Optimization in Data Mining. Fachbereich Informatik, Technische Universität Dortmund, 2008.


Mierswa/etal/2008b Mierswa, Ingo and Morik, Katharina and Wurst, Michael. Collaborative Use of Features in a Distributed System for the Organization of Music Collections. In Shen and Shephard and Cui and Liu (editors), Intelligent Music Information Systems: Tools and Methodologies, pages 147--176, Igi Global Publishing, 2008.


Birkmann/Deuse/2007a Birkmann, Stephan and Deuse, Jochen. Using a Group Technology Approach to Level a Low Volume and High Mix Production. In Proceedings of 12th Annual International Conference on Industrial Engineering - Theory, Applications and Practice, pages 265--270, Cancun, Mexico, 2007.


Deuse/etal/2007a Deuse, Jochen and Stausberg, Jan Robert and Wischniewski, Sascha. Leitsätze zur Gestaltung einer verschwendungsarmen Produktion. In ZWF -- Zeitschrift für wirtschaftlichen Fabrikbetrieb, Vol. 102, No. 5, pages 291--294, 2007.


Mierswa/Wurst/2006a Mierswa, Ingo and Wurst, Michael. Information Preserving Multi-Objective Feature Selection for Unsupervised Learning. In Maarten Keijzer and Mike Cattolico and Dirk Arnold and Vladan Babovic and Christian Blum and Peter Bosman and Martin V. Butz and Carlos Coello Coello and Dipankar Dasgupta and Sevan G. Ficici and James Foster and Arturo Hernandez-Aguirre and Greg Hornby and Hod Lipson and Phil McMinn and Jason Moore and Guenther Raidl and Franz Rothlauf and Conor Ryan and Dirk Thierens (editors), GECCO '06: Proceedings of the 8th annual conference on Genetic and evolutionary computation, pages 1545--1552, New York, NY, USA, ACM Press, 2006.


Mierswa/Morik/2005a Mierswa, Ingo and Morik, Katharina. Automatic Feature Extraction for Classifying Audio Data. In Machine Learning Journal, Vol. 58, pages 127--149, 2005.


Morik/Koepcke/2004a Morik, Katharina and Köpcke, Hanna. Analysing Customer Churn in Insurance Data - A Case Study. In Jean-Francois Boulicaut and Floriana Esposito and Fosca Giannotti and Dino Pedreschi (editors), PKDD '04: Proceedings of the 8th European Conference on Principles and Practice of Knowledge Discovery in Databases, Vol. 3202, pages 325--336, New York, NY, USA, Springer, 2004.


Morik/etal/99a Morik, Katharina and Brockhausen, Peter and Joachims, Thorsten. Combining statistical learning with a knowledge-based approach -- A case study in intensive care monitoring. In ICML '99: Proceedings of the Sixteenth International Conference on Machine Learning, pages 268--277, San Francisco, CA, USA, Morgan Kaufmann Publishers Inc., 1999.


Eversheim/Deuse/98a Deuse, Jochen and Eversheim, W.. Teilefamilienbildung auf der Grundlage von Produktmodelldaten. In Krause, Frank-Lothar and Uhlmann, Eckhardt (editors), Innovative Produktionstechnik, pages 141--156, München, Wien, Carl Hanser Verlag, 1998.


  • Bohnen/Deuse/2010a - Leveling of Low Volume and High Mix Production based on a Group Technology Approach
  • Morik/etal/2010a - Prognosemodelle zur Ermittlung der Produkteigenschaften -- Einsatz von Data-Mining-Verfahren im Walzwerk
  • Morik/etal/2010b - Data Mining in Sensordaten verketteter Prozesse
  • Mierswa/2008a - Non-Convex and Multi-Objective Optimization in Data Mining
  • Mierswa/etal/2008b - Collaborative Use of Features in a Distributed System for the Organization of Music Collections
  • Birkmann/Deuse/2007a - Using a Group Technology Approach to Level a Low Volume and High Mix Production
  • Deuse/etal/2007a - Leitsätze zur Gestaltung einer verschwendungsarmen Produktion
  • Mierswa/Wurst/2006a - Information Preserving Multi-Objective Feature Selection for Unsupervised Learning
  • Mierswa/Morik/2005a - Automatic Feature Extraction for Classifying Audio Data
  • Morik/Koepcke/2004a - Analysing Customer Churn in Insurance Data - A Case Study
  • Morik/etal/99a - Combining statistical learning with a knowledge-based approach -- A case study in intensive care monitoring
  • Eversheim/Deuse/98a - Teilefamilienbildung auf der Grundlage von Produktmodelldaten