cacaroot - stock.adobe.com
The field of process mining started in the late 1990s when Wil van der Aalst, who is now a professor leading the Process and Data Science group at RWTH Aachen University, began looking for ways to combine process science and data science. Much of this early work was theoretical, but the field has started accelerating over the last couple of year with advancements in data gathering and analytics technologies.
"The adoption of process mining has accelerated over the last couple of years," van der Aalst said in an interview. There are now over 30 vendors of commercial process mining tools, including leaders like Celonis, Disco, UiPath (ProcessGold), myInvenio, Minit, Mehrwerk, Lana Labs, StereoLOGIC and Everflow. This has made it easier for large organizations, like Siemens and BMW, to apply process mining at scale with thousands of process mining users.
The techniques are being used in a range of industries, including logistics, manufacturing, finance, healthcare, CRM and communication. "Any organization can use process mining, and the prerequisites are minimal," van der Aalst said.
These initiatives are focused on process discovery and conformance analysis to understand why people are not following a process. More sophisticated initiatives are also looking at using process mining to support conformance checking and more advanced analytics, such as predicting the remaining processing time.
Van der Aalst said the main motivation of working on process mining was his disappointment in the practical use of simulation software and workflow management systems that relied on humans to make process models. These were good at capturing the happy flows but failed to capture the less frequent executions that generate most of the problems.
Most workflow management projects failed at the time. They also found that the process of making the business simulations was often more insightful than the results since simulations rely on too many simplifying assumptions.
"All these experiences showed me that process orientation is important, but one should connect process management to the actual evidence recorded in databases and audit trails," van der Aalst said. He's hopeful that this viewpoint has become mainstream over the last five years.
Enterprises are starting to use a variety of data mining techniques in concert with the rise of cloud applications, better analytics and AI to improve their understanding of the business. But van der Aalst believes these are fundamentally different than process mining. "Traditional machine learning, data mining and [AI] approaches do not support process mining because there is no notion of a process model and event log," he said. For example, deep learning cannot be used to discover process models or to check conformance.
Wil van der Aalst
He sees many vendors and scientists using terms like machine learning to rebrand business intelligence functionalities that have been around for decades. This could lead to disappointment and bad investments.
Although process mining is different, it can be combined with data mining and machine learning. After discovering a process model, one may have identified interesting bottlenecks and deviations that need to be investigated further. Using process mining results, it is possible to generate supervised learning problems that can be answered using approaches ranging from decision trees to deep neural networks. For example, the combination of techniques can be used to explain why delays occur or predict when deviations are more likely.
Going forward, van der Aalst sees a convergence of traditional process modeling and process mining. There is still a gap between the Business Process Model and Notation systems made with tools such as Signavio, Bizagi, ARIS, iGrafx and Camunda and the process models generated using process mining. He believes that industry efforts to close these gaps could help make process mining data useful to a wider base of business users.
Focus on business process hygiene
Process mining reveals performance and compliance problems. By making these transparent, processes can be improved by reducing costs, improving quality and removing delays. Although the advantages should be obvious, management often expects a business case to show the ROI, thus slowing the adoption of process mining, van der Aalst said.
A more appropriate focus for CIOs would be to think about process mining compared to personal hygiene. Although it is not easy to make a clear business case for personal hygiene, not doing it increases the risk of getting and spreading a range of diseases. "Organizations not using process mining are, most likely, not aware of their actual processes and problems," van der Aalst said.
He believes it is important to start process mining on a larger scale and not as a small toy project. He also recommended that it be done continuously and for many processes rather than as a pilot project doing one analysis on one process at a given time.
Since data extraction is time-consuming, one should do this only once but do it in such a way that it can be applied continuously using the latest event data. "CIOs should not ask for a business case without understanding the impact and broad scope of process mining," he said.