Logo Universität Bayreuth

University of Bayreuth, Presse release No. 033/2023 - 28 March 2023

New DFG research unit develops process mining techniques for the Internet of Things

Prof. Dr. Agnes Koschmider from the University of Bayreuth is the spokeswoman of the new DFG research unit FOR 5495 "SOURCED – Process Mining on Distributed Event Sources". Process mining is a proven technique developed for the discovery, analysis and evaluation of business processes. As of today, the application of process mining struggles to process data from distributed sensor-based systems in the context of the Internet of Things. The goal of the new research unit SOURCED is therefore to provide the methodological foundations of novel process mining techniques for data from the Internet of Things. 

The research unit combines competencies from the fields of process management, data and software engineering, distributed systems and privacy. It comprises six subprojects located at the University of Bayreuth, the Christian-Albrechts-University of Kiel, and the Humboldt-University at Berlin. The German Research Foundation (DFG) is funding SOURCED for an initial period of four years. 

A similar mobile Tiny House will be set up as a real laboratory on the Campus of the University of Bayreuth.

As part of the research unit, a "mobile tiny house" will be set up on the campus of the University of Bayreuth to bridge research to concrete applications. As a real laboratory, it will enable the prompt evaluation and further development of all methods developed by the project partners.

SOURCED addresses a global trend: The "Internet of Things", which connects real and virtual objects and integrates them into new information and communication structures, has experienced a strong upswing. Spatially distributed sensors are often embedded in such intelligent infrastructures, for example sensors for temperature, speed or time measurement. They transmit event-related data that are evaluated in their respective context and thus form a basis for decisions: In logistics, for example, data from ship transponders is used to monitor the loading and unloading of ships. In healthcare, hospitals are installing real-time location systems so that they have the required overview of clinical processes at all times. Smart city initiatives, on the other hand, track information about traffic events and public transit density.

"In all of these scenarios, there is a technical infrastructure with a large number of spatially distributed locations where sensors generate event-related data. This data is essential for accurate reconstruction, analysis, and evaluation of complex operations and thus contributes to efficient control of the overall system. For example, information about incidents in a hospital or delays in city traffic must first be classified into the relevant chains of causes and effects: In this way, they can support efficient decision-making by doctors or by users," says Prof. Dr. Agnes Koschmider, Chair of Business Informatics and Process Analytics at the University of Bayreuth.

The term "process mining" describes a technique to discover, analyze and evaluate processes in complex infrastructures. However, applications on infrastructures in which spatially distributed sensors provide important data are currently still associated with technical and conceptual problems. These relate to the efficient processing of the event data transmitted by sensors, but also to data protection: although the high level of detail of the data is a quality feature, further processing that is legally and ethically appropriate requires a degree of generalization that protects individuals, companies or organizations from recognition and allows protection of privacy. In addition, the communication aspect should not be underestimated: The findings obtained through process mining must be visualized and communicated in their factual and logical contexts so clearly that they represent a real decision-making aid for those responsible in companies and organizations.

"In our new research unit, we are laying the foundation for novel applications of process mining based on event-related data transmitted by distributed sensors in a network of the Bayreuth, Berlin and Kiel sites. We refer to this new generation of process analysis techniques that we want to develop as Sourced Process Mining," says Koschmider.

Prof. Dr. Agnes Koschmider.

Prof. Dr. Agnes Koschmider

Chair of Business & Information Systems Engineering and Process Analytics
University of Bayreuth

Phone: +49 (0)921 / 55-4583
E-mail: agnes.koschmider@uni-bayreuth.de

Christian Wißler, Wissenschaftskommunikation

Christian Wißler

Deputy Press & PR Manager, Research Communication
University of Bayreuth

Phone: +49 (0)921 / 55-5356
E-mail: christian.wissler@uni-bayreuth.de