Skip to main content

Celonis Product Documentation

Discover Parallel Processes

One main advantage of the Multi-Event Log technology is the ability to model and work with parallel processes. This refers to Use Case 2 mentioned here. Parallel processes can always be looked at in a hierarchical way. An example of a hierarchical process is this car manufacturing process.

This shows a high-level process on the left and a granular process for each station on the right. The activities on the right are taking place inside the activity 'Station: Initialization Wheel Carrier Mounting' of the high-level process.

50731775.png

The top hierarchy of the process is modeled on station level. Each station is represented by one activity. Inside each station are multiple activities happening. This more granular process is in the Data Model stored in a separate Activity table and represented on the right in a separate Process Explorer.

This way of modeling can be exploited for data exploration of parallel processes. If there are parallel streams on the granular process level, they can now be visualized natively in this granular Process Explorer. As we use the granular Event Log for this Process Explorer, multiple cases can occur in parallel for each high-level case. An example of this is shown below for a review process that includes two reviewers. The high-level process on the left has one case per reviewed paper. The granular reviewing process on the right has one case per reviewer. For one high-level case, multiple granular cases can occur in parallel. The Multi-Event Log technology allows visualization and calculation of KPIs in parallel processes.

50731776.png

To set up such a hierarchical Data Model that allows analyzing parallel processes, the Event Logs of different granularity can be linked through the Case tables. The number of levels of granularity is not limited. It would for example be possible to have four or five levels of granularity. This shows a typical set up to analyze parallel processes inside the granular process; this is the Reviewer process. Each case of the granular process (Reviewer Case) is linked to a case of the high-level process (Paper Case).

50731777.png

Note

Existing components and visualizations in Process Analytics can be used for Multi-Event Log analyses to a great extent already. Some components are more applicable to the single Event Log use case and always access the default Event Log: Throughput Time Search, Case Explorer, Process Overview, Selection Views (except Attribute Selection).

Until then, the best practice for leveraging the existing Process Analytics capabilities for MEL analyses is:

  • Throughout Time Search: can be composed with variables, button drill-downs, single KPI, and histogram to calculate throughput times for a specific or across Event Logs.

  • Process Explorer:

    • Use Case 1: Use one process explorer for each Event Log and define “Custom dimension” accordingly.

    • Use Case 4: Use MERGE_EVENTLOG functionality in “Custom dimension”.

  • Case Explorer: can be composed for each Case & Activity Table on a sheet with two OLAP tables, one holding the case information and one holding the activity information.