How Process Mining Can Help Manufacturing and Assembly Lines

As industrial manufacturing companies implement a variety of business software tools, each of these systems generates a historical record of data and error reports that are proving to be a boon for manufacturing analytics professionals. Using an approach known as Process Mining, operations analysts can collect and analyze the massive amount of big data in manufacturing systems to gain insight into existing business processes, identify problems such as bottlenecks, and find ways to improve overall operational workflow.


Interconnected Company Concept


The emergence of Process Mining as a standalone discipline is a fairly recent development in the business operations management realm. In this article, we’ll attempt to answer the frequently asked questions about what process mining is and how it fits into the world of manufacturing and assembly lines. Let’s get started!


What is Process Mining and Where Does It Fit Within the Broader Business Management Landscape?


At its most basic level, process mining is a research discipline that scoops up event data logs produced by a variety of heterogeneous enterprise systems in order to find useful information about the current state of actual business processes within the organization.


From this simple definition, it would be easy to come to the conclusion that Process Mining is the same as what’s commonly referred to as Big Data analytics. Both disciplines seek to identify cause and effect relationships that affect business performance, but there’s a major difference between the two in terms of scope.


Unlike Big Data analytics, which examines a wide variety of datasets — ranging from customer preferences to economic conditions to weather forecasts, Process Mining is by definition limited to the realm of extracting useful knowledge from event logs and other similar data sources within the organization or its supply chain.


Process Mining vs. Data Mining, image by All About Requirements
Process Mining vs. Data Mining, image by All About Requirements

Process Mining’s ability to uncover, monitor and analyze business processes across the organization makes it a welcome new tool for professionals working in Operations Research and Business Intelligence (BI) disciplines. Insights from Process Mining analytics can be used to inform a wide variety of Business Process Improvement (BPI) tools and approaches, including:


  • Business Activity Monitoring (BAM)
  • Business Operations Management (BOM)
  • Complex Event Processing (BEP)
  • Corporate Performance Management (CPM)
  • Continuous Process Improvement (CPI)
  • Executive “Dashboards”
  • Kaizen / Toyota Production System (TPS)
  • Key Performance Indicators (KPIs)
  • Lean Manufacturing
  • Lean Six Sigma
  • Total Quality Management (TQM)
  • SCOR Business Process Modeling from APICS
  • Six Sigma


What are the Primary Goals of Process Mining?


Depending upon the initial state of their business process modeling efforts, businesses can expect to achieve one or more of the following Process Mining goals:


1. Discover Actual Business Processes


By looking at event logs (and taking note of process errors and exceptions), Process Mining can create useful diagrams that document the actual processes taking place within an organization, including the pathways that handle exceptions that fall out of standard processes.


2. Provide Conformance Checking Between Actual Business Processes and Assumptions Made in Process Models


Businesses that have already created business process models of their existing operations (known as the “AS-IS” condition in APICS’ SCOR terminology) can evaluate the accuracy of their models by measuring deviations from the real-world conditions as reported by Process Mining.


3. Make Enhancements and Improvements to Business Processes on an Ongoing Basis


Using Process Mining on an ongoing basis can help identify ways to improve processes across the business, such as repairing production bottlenecks or reducing errors under specific conditions. As new process improvements are introduced, Process Mining can also provide useful feedback on the efficacy of individual process changes by comparing the results to historical data records.


What’s the Relationship between Process Mining and Business Process Modeling (BPM) Frameworks and Approaches, Such as Six Sigma and SCOR?



Rather than viewing Process Mining as a threat that could displace existing process improvement frameworks, most business operations professionals welcome Process Mining as a valuable new addition to their business analysis toolkit.


For example, Process Mining’s ability to derive a current “AS-IS” process model from a wide range of event log data can help frameworks like Six Sigma and SCOR validate their assumptions about process workflow as well as identify areas that need improvement.



Can You Use Process Mining Without First Creating a Business Process Model?


The short answer is “yes,” you can implement a Process Mining system without first creating a theoretical business process model. In fact, Process Mining will build its own internal business process model as an outcome of analyzing the data logs created by enterprise software systems.


This Process Mining-generated business process model can be used as a jumping off point for further analysis by various business process improvement frameworks, such as SCOR from APICS, where it could serve as the working model for the “AS-IS” condition.


How Does Process Mining Interface with ERP, CRM, B2B and Other Business Software Tools?


Process mining, image by Lexmark
Process mining, image by Lexmark


Most enterprise resource planning (ERP) systems (such as those from software giant SAP) generate significant amounts of operational data in the form of transaction logs, event logs, activity and case records, and history audit trails. Other business-to-business (B2B) systems do as well, including inventory management systems, warehouse shipping and receiving systems, customer relationship management tools, customer call center / tech support case management systems, contracts and procurement management tools, accounting and accounts receivable systems, and human resource benefits management and payroll systems.


Process Mining imports the historical data logs generated by these systems and uses this information as raw data for downstream analysis. In some cases, Process Mining software developers have created extensions or plug-ins to read input data directly from popular business software tools.



What Specific Benefits Does Process Mining Offer for Manufacturing Facilities and Assembly Lines?


Thanks to its ability to quickly shift through the large amount of available big data in manufacturing control systems, Process Mining can often provide specific, actionable recommendations that may elude conventional manufacturing analytics methods.


Process Mining’s manufacturing analytics approach can be used in a number of ways. For example, it can objectively compare factory output at a high level, such as between facilities in different locations or between shifts at the same facility.


It can also drill down to analyze what’s happening between any two arbitrary points in the manufacturing process to uncover a range of problems, including:


  • Deviations for target processes
  • Inventory shortages
  • Repeated production errors
  • Procurement “irregularities”
  • Unnecessary detours
  • Bottlenecks
  • Quality variations


What is a Petri Net Diagram and How Do You Interpret It?


As we’ve discussed, Process Mining can help create business process models from the data files produced by business software systems. It’s often a challenge to represent these process models in a comprehensive (yet understandable) way. Compounding the issue is the need to represent processes from multiple perspectives, including:


  • the control-flow perspective, which documents all possible activities in sequence
  • the organizational perspective, which identifies the people, departments or systems involved
  • the case perspective, which documents activities from a production element perspective; and the time perspective, which measures sequencing and frequency of operations.


Over the years, several different mathematical models (each using their own graphical notation) have been developed to represent the complexity of dynamic events that occur in the real-world. The most common representations used to diagram Process Mining control flows are:


BPMN Process
BPMN Process. Image by Wikipedia


  • Business Process Model and Notation (BPMN)
  • EPC (Event-driven Process Chain) diagrams
  • Petri net (Place/transition net) diagrams
  • UML (Unified Modeling Language) activity diagrams


uml activity diagram
UML Activity diagram. Image by Wikipedia


For the uninitiated, the BPMN and EPC notation will seem most understandable at first glance as they bear a strong superficial resemblance to the familiar “decision tree” type of graphs. UML activity diagrams are static representations of a language model that was developed to support complex software development. Petri net diagrams, in contrast, have an appearance that at first glance may remind you of high-level electric diagrams; each of the nodes represents a different set of events (circles = conditions, bars = events, arcs = pre- or post-conditions, arrows = transitions).


petri net example
Petri Net Example. Image by Wikipedia


What are some of the Challenges in Implementing Process Mining at Your Facility?


Like many information technology projects, each Process Mining implementation is different and will face its own unique set of challenges.


However, because the usefulness of Process Mining depends upon successfully interpreting event logs produced by different systems, nearly all implementations will come up against the challenge of identifying, merging, and cleaning event log data. This isn’t necessarily the fault of the systems providing the raw data, as few currently-available systems were designed with Process Mining in mind.


Consequently, implementation engineers will often need to come up with creative solutions. For example, many data logs will track the statistics for specific objects rather than the underlying processes involved. In these situations, implementation engineers will have to look for proxies that indicate process workflows in order to make the data more useful for things like manufacturing analytics.


In response to these challenges, the IEEE Task Force on Process Mining has been working to promote logging format standards, such as XES.


How to Start Process Mining at Your Manufacturing or Assembly Facility


Interested in implementing a Process Mining program at your manufacturing or assembly facility?


There are a variety of open source and commercial software Process Mining tools available.  


One of the leading open source Process Mining software solutions is ProM Tools, which is short for Process Mining framework. This software was developed by Wil van der Aalst and his research group at the Eindhoven University of Technology in The Netherlands, which has been a leading center for Process Mining research. Other open source Process Mining solutions include PMLAB and Apromore.


Commercial software vendors are also developing Process Mining tools, including:


  • ARIS Process Performance Manager (Software AG)
  • Celonis (SAP Process Mining Partner)
  • Comprehend (Open Connect)
  • Discovery Analyst (StereoLOGIC)
  • Flow (Fourspark),
  • Futura Reflect (Futura Process Intelligence)
  • Interstage Automated Process Discovery (Fujitsu)
  • OKT Process Mining suite (Exeura)
  • Process Discovery Focus (Iontas/Verint)
  • ProcessAnalyzer (QPR)
  • Rbminer/Dbminer (UPC)
  • Reflect|one (Pallas Athena)


The Eindhoven University of Technology offers a free online course via its training partner Future Learn. Humboldt University in Berlin also offers a comprehensive Process Mining class.


You can also find online training materials specifically for ProM software available at as well as online training classes for software from Celonis. (Celonis is an SAP Process Mining partner.)



Formaspace is Dedicated to Making Your Manufacturing Processes More Efficient

mobile cart attaches and detaches to workbench
Detachable Mobile Cart Station

As you can see, Process Mining is a powerful way to leverage Big Data in manufacturing facilities.


Want to learn about some other ways to make your manufacturing facility more productive?


Talk to the manufacturing workflow experts at Formaspace. We’re happy to share our expertise gained from years of creating industrial furniture solutions for Fortune 500 companies — custom designed to streamline the manufacturing processes on your factory floor — and beyond.


Ask about our consulting services as well. Our team of efficiency experts can visit your facility to perform a Rapid Plant Assessment to help you improve the workflow at your manufacturing or assembly plant.


It’s easy. Just fill out the simple contact form below, and a Formaspace Design Consultant will be in touch.


We'll never share your email with anyone else.