Glossary · Automation fundamentals, platforms and components
Process mining
German: Process Mining
In data analysis, process mining is a family of techniques that reconstruct, check and improve real processes from event logs recorded by IT systems, each event carrying at least a case ID, an activity and a timestamp. It covers process discovery, conformance checking and enhancement.
- Automation components
- AI
In one sentence
Process mining reconstructs and analyzes real process flows from event logs with case ID, activity and timestamp, for discovery and conformance checks.
Example
Event logs from the MES show that 18 % of orders loop back to rework at the testing station, a path that was missing from the documented process.
How it applies
- Operation: In manufacturing, event data from MES, ERP and machine logs show how orders, batches or service cases really flow, where they wait and where they deviate from the planned route.
- Data quality: Results depend on consistent case IDs and synchronized Timestamp values across systems. An Audit trail or event log with gaps produces misleading process maps.
- Documentation: Conformance checking compares logs with the documented process. Documentation teams can use the findings to correct procedures that no longer match practice, or to flag practice that violates a procedure. Anonymize or pseudonymize person-related data in logs before analysis.
IEEE 1849 defines the XES format for exchanging event logs. Process mining shows correlations in data; conclusions about causes still require domain review.