The following paper explains relevant fundamentals of the discussed topic with respect to process mining as well as for the understanding of supply chain processes. Consequently, in a second step, a literature review regarding the research field of process mining will be unfolded by retracing its design and conduction after elaborating a well-defined search phrase based on the research question. This paper then advances by summarizing the results of the literature review. The state-of-the-art research with respect to the specific topic of this very paper are summarized and commented subsequently, and if applicable, taken up in later course. In this regard, special challenges and approaches will be mentioned. The third part of this paper deals with the simulation as basis for conducting cross-organizational process mining. Here, the setting, the char acteristics, the simulation model, and process errors will be explained. The resulting event log is going to be analyzed and interpreted by comparing to the input model.
In the last part, the event data of the supply chain process simulation serves as input for cross organizational process mining. The results of the application will be analyzed and viewed at with specifically focusing context-related errors followed by evaluation of the process model considering the input-output relation with the event log from the simulated supply chain process. As a conclusion, the results will be discussed and compared to other research results and questions will be raised for further research.
Table of Contents
1 Introduction
1.1 Methodology
1.2 Theoretical Framework
1.2.1 Process Mining
1.2.1.1 Event Log
1.2.1.2 Techniques
1.2.1.3 Software
1.2.2 Context of Process Mining
1.2.2.1 Data Mining
1.2.2.2 Machine Learning
1.2.2.3 Business Process Management
1.2.3 Supply Chain Processes
1.2.3.1 Supply Chain Operations Reference Model
2 Literature Review
2.1 Methodology
2.1.1 Research Questions and Scope
2.1.2 Automated Search
2.1.3 Inclusion and Exclusion Criteria
2.2 Application of Process Mining in the Field of Supply Chain Management
2.2.1 Comparison and Collaboration
2.2.2 Mining Cross-Organizational Event Data
2.2.3 Process Mining as a Service
2.2.4 Mining Inter-Organizational Event Data
2.2.5 RFID-based Process Mining
2.2.6 Process Unaware Systems
2.2.7 Privacy Preservation in Cooperative Process Mining
2.2.8 Alternative Approaches
2.2.9 Challenges in Supply Chain Process Mining
3 Application of Process Mining in Supply Chain Processes
3.1 Simulation
3.1.1 Supply Chain Process
3.1.2 Model
3.1.3 Process Errors and Failure
3.1.4 Event Log
3.2 Mining Simulated Event Logs
3.2.1 Error Identification
4 Conclusion
4.1 Discussion and Outlook
Research Objectives and Themes
This paper aims to explore the potential of process mining techniques to facilitate error identification within cross-organizational supply chain environments. By combining simulation-based event data with established process mining principles, the study investigates how these techniques can effectively address the complexities inherent in multi-party supply chain processes and identify performance deviations.
- Theoretical foundations of process mining and Business Process Management (BPM).
- Development of a reference model for supply chain simulations to generate event logs.
- Comparative analysis of intra-organizational versus inter-organizational process mining applications.
- Evaluation of error identification techniques using specific failure metrics and disturbance tracking.
- Assessment of technical challenges, including data heterogeneity and privacy in cooperative process mining.
Excerpt from the Book
1.2.1.1 Event Log
As stated above, PAIS store tons of event data. This often happens in an unstructured manner, as it is not the PAIS’ main purpose. Thus, the extraction of event data is an essential step towards further data processing and mining. That is, although referring to event logs as basis, event data are not necessarily stored in ready-to-use files.
Speaking of event logs, we presume it represents the data of a single process in the form of a table. The minimum components of an event log are defined by REINKEMEYER as case identifier (ID), activity, and timestamp.
Though, VAN DER AALST opines that case ID and activity are the least necessary for process mining, he prefers event IDs before timestamps for ordering purposes and timestamps for measuring and analyzing performance metrics. Reason being is that taking timestamps for determining the position might be problematic for discovery algorithms that do not just rely on the timestamp within a case but on the timestamps of all activities performed. The case ID relates to a unique process instance. Each event in the log refers to one process instance only. VAN DER AALST complements events can also be linked to an activity. The sum of activities linked to a process instance reveal the whole process being performed on it.
An event log might provide additional information both authors refer to as attributes as for example information on the activity performed.
Being the single source of truth that it is, event logs heavily influence the output of process mining. That makes data quality a major subject considering process mining. Data quality starts with the construction and design of event logs as potentially structural quality issues but also includes missing and concealed entities and missing, incorrect and imprecise attributes.
Summary of Chapters
1 Introduction: Provides the motivation for the study, highlighting the growing significance of process mining for supply chain efficiency in a digitized global economy.
2 Literature Review: Synthesizes current academic research on applying process mining techniques to supply chain management, focusing on both intra- and inter-organizational contexts.
3 Application of Process Mining in Supply Chain Processes: Describes the development of a simulation-based model to test process mining applications, focusing on error identification and data generation.
4 Conclusion: Discusses the findings of the simulation application and concludes by assessing the trade-offs between model complexity and abstraction in real-world supply chain scenarios.
Keywords
Process Mining, Supply Chain Management, SCOR Model, Event Logs, Simulation, Error Identification, Business Process Management, Data Mining, Machine Learning, RFID, Inter-organizational Processes, Process Discovery, Performance Metrics, Digitalization, Supply Chain Integration.
Frequently Asked Questions
What is the core focus of this research paper?
The paper primarily investigates how process mining can be applied to identify errors within complex, cross-organizational supply chain processes through the use of simulation-based event logs.
What are the main research themes covered in the study?
Key themes include the theoretical framework of process mining, the application of the SCOR reference model, the challenges of mining distributed event data across organizational boundaries, and the practical implementation of error detection using simulation.
What is the primary objective of this work?
The central goal is to evaluate the capability of standard process mining techniques to discover sub-processes and filter specific error types within a multi-party supply chain environment.
Which scientific methodology is employed?
The author employs a semi-systematic literature review followed by a practical simulation approach, using Tecnomatix Plant Simulation to generate event data and Disco for subsequent process mining analysis.
What aspects are discussed in the main body (Chapter 3)?
The main body details the construction of a linear supply chain simulation using the SCOR model, defines types of failures (such as scrap and return processes), and demonstrates how station-specific disturbances are captured in event logs.
Which keywords best describe this work?
Core keywords include Process Mining, Supply Chain Management, SCOR Model, Event Logs, and Simulation-based Error Identification.
How are "Spaghetti processes" addressed in the context of this research?
The author discusses how unstructured, high-frequency activities (Spaghetti processes) complicate process discovery and proposes techniques like clustering and process fragment extraction to resolve complexity.
What role does RFID technology play in this research?
RFID is identified as a vital contributor to process mining in supply chains because it provides a shared resource identifier, allowing for better traceability and the mitigation of IT system heterogeneity.
What are the limitations identified regarding the simulation model?
The author acknowledges that the current model assumes a linear supply chain, which is a simplification of real-world networks, and notes that data generation in the "Supplier" phase occasionally interferes with the resulting process analysis.
- Quote paper
- Max-Leonhard Böttrich (Author), 2022, Identifying Errors in Supply Chain Processes with Process Mining, Munich, GRIN Verlag, https://www.grin.com/document/1254759
