Applying Object-Centric Process Mining to a Real Business Process: Inventory in SAP ERP Systems
Applying Object-Centric Process Mining to a Real Business Process: Inventory in SAP ERP Systems
Student: Julian Kofferath
Title: Applying Object-Centric Process Mining to a Real Business Process: Inventory in SAP ERP Systems
Supervisor: Gyunam Park
1st Examiner: Prof. Wil M.P. van der Aalst
2nd Examiner: Prof. Dr. Stefan Decker
Summary
Many enterprises use ERP (Enterprise Resource Planning) systems to execute and record their business processes. For all relevant departments of a company, e.g., sales and distribution, accounting, or controlling, ERP systems like SAP ERP provide solutions. However, those business processes are often not executed efficiently. From an economic point of view, it would be negligent not to use the recorded data to optimize the processes and save resources. Using process mining, one can identify bottlenecks and inefficiencies of real-life process executions. To apply existing algorithms to discover process models (process discovery), verify existing process models (conformance checking) and improve those (enhancement), or analyze performance measures, one needs to extract an event log from the ERP system’s database. Since processes from ERP systems often involve several object types and do not provide a unique case notion, Object-Centric Process Mining (OCPM) is more suited for process analysis than traditional process mining since it considers the interaction of objects without the notion of a case id. This thesis conducts a case study of the inventory process in SAP ERP to enable the object-centric process analysis of an ERP process. We simulate the process and populate an SAP ERP system with regular executions. Next, we extract an object-centric event log from the database to discover a process model using OCPM. Finally, we analyze two different scenarios of the process. Results show that even the inventory process as a well-defined ERP process may cause organizational friction.