Operationalization of Predictive Process Monitoring for Aircraft MRO

 

Student: Julian Walter-Nußberger

Title: Operationalization of Predictive Process Monitoring for Aircraft MRO

Supervisor: Dr. Sebastiaan J. van Zelst

1st Examiner: Prof.Dr. Wil M.P. van der Aalst

2nd Examiner: Prof.Dr. Martin Grohe

Summary

The aircraft maintenance, repair, and operation (MRO) process at <<Company Not Disclosed>> regularly experiences failures where tasks must either be rescheduled in advance or cannot be completed during the scheduled work shift. These issues result in additional workload on the planning department and potential flight delays or cancellations. To streamline the process, the planning department is interested in finding reasons for failures. The MRO process has multiple possible case notions. Therefore, classic event logs are unable to properly represent the data, which disregards certain perspectives. In this thesis, we utilize the novel data structure Object-Centric Event Log (OCEL) which relates events to multiple case notions. We propose processing methods for OCEL filtering and splitting for which no technique previously existed. We use the OCEL data model to discover process models, diagnose reasons for failures in the MRO process and create a classification model to predict failures. The evaluation shows that Object-Centric Event Logs are a suitable data structure for this problem. We identify multiple parameters that correlate with failures and the classifier is able to predict failures with high accuracy.

Note

This work was performed in collaboration with an industrial partner. Therefore, the final thesis is not shared publicly. For more information, contact Dr. Sebastiaan J. van Zelst