Discovering Process Disarray with Causal Process Mining

Hey there, friends! Welcome to Noreja Highlights. This month, we’re excited to showcase a key feature of Noreja Process Intelligence – disarray discovery via our graph-based process model map. Let's explore how it simplifies process discovery and the benefits it offers.

What’s Process Mining Again?

Think of process mining as a time machine for your company’s workflow. It uses your existing data to create a step-by-step journey of your business tasks, like a digital blueprint. Unlike the existing approaches, where every activity type must be defined, our causal approach to process mining starts with a basic map and then uncovers the actual workflow based on your company’s activity.

?

What is Disarray Discovery?

Remember we mentioned how this helps in checking the health of an ERP system? Let’s have a look at what makes this possible – and try not to get too technical while doing it!

Here’s a map of the happy path of a process our tool generated based on the dataset of a German Medical Supplies company. For a more focused overview, we have hidden a lot of additional information on the process graphs for you:

Looks simple, doesn't it? Well, the trick to discovering your business’s real processes is that a lot of complexity lurks beneath the surface. If we now increase the variant coverage of the process, we see additional activities popping up on the process map making the process more complex:

In this case, we see this process has three levels of payment escalation, of which each can be the final level before receiving payment. There’s still more that we can find, however. If we turn on some additional path options, for example, we can get the following view:

Here you see a bunch of red arrows going back and forth. This denotes disarray, cases where the process steps were executed in the wrong order. For example, we see that in some cases, the payment was received (Zahlung erhalten) before the invoice was sent, as well as a high number of other order infractions. If we now filter to only those instances with this exact relationship, we see this disarray in even more detail and at a very granular level. Dashed arrows – so called hypothetical paths – indicate that there is an intended process path that is, according to the real data, not being followed:

Causal Process Mining detects this by mapping the real-world execution of each instance to a causal template defined by domain knowledge, e.g., based on explicit documentation or implicit expert experience. By the way, this approach also helps identifying entire process variants that deviate from the intended or assumed process – we’ll have to wait until a future Feature Highlights edition to talk more about this function, though!

?

Practical Uses

So why do I care? We’ve talked about how this helps in QCing digitalisation, but there’s a pile of other use cases:

• Certifications: For those in fields like healthcare or pharma, these insights are crucial for staying within rules and regulations.
• Efficiency: For businesses like food suppliers, this approach can squeeze out extra efficiency to stay ahead.
• Custom Orders: For businesses often dealing in custom orders or a wide range of products, like fabric makers, causal maps can guide you to price your work smartly.

?

A Little Wisdom

While causal process mining sheds light on many aspects of your organisation, wise human judgment remains key. The main advantage of using these technologies is the clearer picture it gives your leadership to make smart, sustainable decisions. Used responsibly and with the right context, it’s a powerful way to make your business run smoother and more effectively.

Next time, we’ll look at variants and how we can use them to understand how people interact with our business processes. See you then!

Best,

Evgeny,

Head of Product @ noreja

Original Link: Noreja Blog - auch auf Deutsch!