Process Mining vs. BPM - Not the same.

Diana Amador, Ph.D.

Business Process Management (BPM) is a discipline in operations management that combines information technology, management sciences, and industrial engineering to discover, model, analyze, improve, and automate business processes[i]. Process Mining is “to discover, monitor, and improve real processes (i.e., not assumed processes) by extracting knowledge from event logs readily available in today’s systems.”[ii] While both disciplines create process models, I would like to discuss the distinction between those created by BPM tools and those created by a Process Mining tool.

In this discussion, I will consider processes models to be Informal, used for documentation purposes, or Formal, which are used for analysis and are also called Execution process models.

An Informal Process Model can be created using PowerPoint or Visio showing a high-level depiction of the process, whereas Formal or Execution models can go as far as to be built in executable code; even an RPA application such as Blue Prism.

Business Process Management Notation (BPMN) is used to create both types of process models because it describes processes in terms of activities, dependencies, and relationships by modeling decisions and defining roles, rules, priorities, and the use of data. Pegasystems, Bizagi, Appian, iGrafx, Blueprint, IBM BPM can be considered BPMs.

The BPMN uses explicit gateways to model control-flow logic[iii], assuming that there is an uncontested agreement between model and reality. However, there is an intrinsic limitation in the notation. It can only represent an idealized view of the process that only partially aligns with reality. A rigorous analysis of the process often reveals different levels of complexity that are not or cannot be represented by a BPMN diagram simultaneously. These types of process models are called by some: “paper tigers.”

To understand the difference between BPM and Process Mining, we should consider the BPM lifecycle, which describes the business process phases: design, configuration/implementation, enactment/monitoring, adjustment, and diagnosis/requirements. As you can see in the BPM lifecycle (figure below,) the cycle restarts when the process performance or its requirements change. Usually, only critical problems or external changes will trigger a new lifecycle. However, thanks to new technologies, now it is possible to collect factual data logged in the enterprise systems and analyze it by using data mining and machine learning techniques. Process Mining is the discipline born from these two techniques that allow us to dynamically understand the real process and create Formal or Execution models that represent its behavior and deviations from assumed models.

In his book “Process Mining,” Dr. Wil van der Aalst, cites the SAP reference model as an example. Dr. Van der Als explains that “more than 20 percent of the SAP models contain serious flaws (deadlocks, livelocks, etc.) [J. Mendling, 101]. Such models do not align with reality and, thus, have little value for end-users.”[iv]

Dr. van der Aalst also explains that the BPM approach tends to be model-driven without taking into consideration the story that real process data tells. However, he also recognizes a new trend in the BPM community to focus on enactment/monitoring, adjustments, and diagnosis/requirements phases, which are data-driven tasks that can be adequately supported by the use of process mining techniques.

Process Mining connects data-driven and process-driven approaches by leveraging the power of ML and data mining. In summary, BPM and Process Mining overlap and can serve the organization’s process modeling, and improvement needs better when used together because they provide complementary approaches to create well documented and realistic process models.

References:

van der Aalst, Wil M. P. Process Mining (p. 458). Springer Berlin Heidelberg. 2016. Kindle Edition.

J.Mendling, G. Neumann, and W.M.P. van der Aalst. Understanding the Occurrence of Errors in Process Models Based on Metrics. In F. Curbera, F.Leymann, and M.Weske, editors, Proceedings of the OTM Conference on Cooperative Information Systems (CoopIS 2007), volume 4803 of Lecture Notes in Computer Science, pages 113–130. Springer, Berlin, 2007.

[i] https://en.wikipedia.org/wiki/Business_process_management | Jeston, John; Nelis, Johan (21 January 2014). Business Process Management. Routledge. ISBN 9781136172984.

[ii] van der Aalst, Wil M. P. Process Mining (p. 458). Springer Berlin Heidelberg. 2016. Kindle Edition. P. 30

[iii] van der Aalst, Wil M. P. Process Mining (p. 458). Springer Berlin Heidelberg. 2016. Kindle Edition. P. 28

[iv] van der Aalst, Wil M. P. Process Mining (p. 458). Springer Berlin Heidelberg. 2016. Kindle Edition. P. 30