SPIE Germany & Central Europe GmbH
Contact at the AI Innovation Center
Maximilien Kintz
KIAFAM (AI-based assistance functions in facility management)
Initial situation
During order processing in facility management, process data is recorded that is suitable for analysis using process mining methods. The process data is available in the IT systems, but has not yet been made accessible for analysis. The potential benefits of this data for process analysis and long-term process optimization are difficult to assess.
AI solution ideas
Process mining methods can be used to examine process data and thus gain knowledge about the actual process flows, which can serve as a starting point for future process optimizations. To do this, the data must first be extracted from the original system and prepared for analysis. The analysis with a process mining tool provides valid findings about the actual process flows.
During the initial analysis, the focus is on mastering the tool chain, the basic understanding of the analysis results and the assessment of the potential benefits.
Figure 1: Preparation of the data
Figure 2: Analysis (Celonis Snap)
Benefit
The data extracted as an example from an IT system must first be prepared and converted into a form suitable for process mining. During the initial exemplary analysis, a tool chain was set up that enables further analyses with little adaptation effort. The analysis carried out with the Celonis Snap process mining tool provides interesting insights into the process flows ("happy path", deviations, variants) and initial process KPIs (e.g. throughput times). The high-quality visual presentation of the results enables discussions with the relevant departments. The easy transferability of the results to other systems and processes results in a high potential benefit.
Implementation of the AI application
Open source tools such as Python and Pandas Profiling were used to prepare the data. The actual analysis was carried out with the process mining software Celonis Snap. The interactive analysis can be used at SPIE and, if required, expanded and used as a template for further analyses.