---
title: "Event-Driven Business Process Management (ED-BPM)"
type: concept
tags: [ed-bpm, cep, complex-event-processing, edpa, event-driven-predictive-analytics, prescriptive-process-monitoring, real-time]
sources:
  - "[[sources/2015-krumeich-prescriptive-control-business-processes]]"
created: 2026-05-11
updated: 2026-05-11
---

# Event-Driven Business Process Management (ED-BPM)

A BPM paradigm in which **business processes are monitored and controlled in real time on the basis of complex event patterns detected across heterogeneous event streams** — typically sensor data, transactional records, and process-execution events — using [Complex Event Processing (CEP)](https://en.wikipedia.org/wiki/Complex_event_processing) as the technical substrate. ED-BPM connects BPM's classical post-hoc analytics to operational sense-and-respond architectures.

The term was popularised in the German BPM research community in the early 2010s; [[sources/2015-krumeich-prescriptive-control-business-processes|Krumeich, Werth & Loos 2015]] develop the paradigm into a full system architecture for prescriptive process control in process manufacturing.

## Conceptual layers

The ED-BPM paradigm is typically organised around three event-abstraction layers (Krumeich et al. Fig. 1):

| Layer | Examples | Quantity | Business impact per event |
|---|---|---|---|
| **Complex events** | Combinations of multiple aggregated events indicating critical situations | Few | High |
| **Aggregated events** | Enterprise-system events, event correlations | Many | Medium |
| **Simple events** | Single sensor signals, atomic process-system records | Very many | Low (individually) |

A CEP engine continuously detects complex event patterns from streams of simple/aggregated events and triggers downstream actions — typically in a connected process engine, prediction layer, or notification system.

## Event-Driven Predictive Analytics (EDPA)

A subcomponent of ED-BPM that extends real-time CEP detection with **predictive analytics**: forecasting complex events *before* they occur on the basis of preceding simple/aggregated events. Coined and developed by Krumeich et al. (citing Janiesch et al. 2012 and Redlich & Gilani 2012 as precursors). EDPA is what makes prescriptive process control feasible — the time gap between event prediction and proactive action determines business value (Krumeich Fig. 2).

## Position within prescriptive process monitoring

ED-BPM and CEP provide one architectural realisation of [[concepts/prescriptive-process-monitoring|prescriptive process monitoring (PrPM)]], distinct from the data-warehouse / association-rule approach in [[sources/2014-groger-prescriptive-analytics-bpo|Gröger et al. 2014]] and the data-stream / classifier approach in later work. The three approaches differ in:

- **Latency** — CEP/EDPA operates on streams (sub-second); data-warehouse approaches operate on batch windows (hours/days).
- **Data substrate** — CEP/EDPA fuses sensor + transactional + process events; data-warehouse approaches assume cleaned, aggregated process data.
- **Prediction model** — CEP/EDPA leans on rule-induction + Bayesian networks; later approaches use deep learning.
- **Industry fit** — ED-BPM was originally motivated by *process manufacturing* (steel, chemicals, pharmaceuticals) where high-frequency sensor data + non-linear material flows make CEP a natural fit. Less obviously useful for purely transactional white-collar processes.

## Connections to APM

- **Recommend role** — CEP-detected complex event patterns can drive APM-agent recommendation in real time.
- **Self-modification** — ED-BPM architectures already include feedback loops from prediction outputs to process-engine control; the [[concepts/self-modification|self-modifying ABPS]] research line extends this with adaptation logic.
- **Process-awareness** — ED-BPM's emphasis on intra-process cross-event mining is one form of process-awareness an APM agent needs.

## Related

[[concepts/prescriptive-process-monitoring]] · [[concepts/predictive-process-monitoring]] · [[concepts/agentic-bpm]] · [[concepts/self-modification]] · [[concepts/business-process-simulation]] · [[sources/2015-krumeich-prescriptive-control-business-processes]] · [[sources/2014-groger-prescriptive-analytics-bpo]] · [[sources/2022-kubrak-prescriptive-ppm-slr]]