Process mining was the one-stop solution to knowing the process “as-is” until process discovery came into the picture. It was superior to mining as it didn’t stop at analyzing the log data; it observed humans as they interacted with the systems in real-time.
As a cognitive tool, process discovery delivers accurate information on the process. It unveils the dark, hidden process through observation. All process variations, nuances, and other process instances become visible, creating a process meta-model or digital twin. Leveraging technologies like Computer Vision, Deep Learning, Machine Intelligence, and Artificial Intelligence, process discovery paints a holistic picture that process mining fell short of.
Let’s find out how.
Where Process Mining Falls Short
Enterprises have several digital systems and software powering their processes. Each of these systems leaves behind a digital trace. Process mining software seeks to dig this digital exhaust through the application event logs.
Log data and its analysis for business analytics are at the baseline of process mining. A process mining tool scans through the recorded event logs and gives a visual representation of the pattern or trend of work. It mines the hard data commits stored with respective activity names, timestamps, and unique Case IDs to get the process picture.
But is a process only about the committed data?
Many actions are taken, changes are made, queries are raised, and deletions are done between the first timestamp and the final timestamp. Apart from that, the process participants transact with other systems between these recorded events. As process mining software works on the committed data states, it ceases to capture all these nuances and variations.
Moving Beyond Hard Data Commits
Process discovery isn’t about the hard data commits and logs. As a true process happens between committed states of data, a process discovery tool observes end-to-end interactions to paint the holistic picture. Process mining, by mining the hard data commits, misses out on the bigger picture. Process discovery, on the other hand, observes work on a large scale unobtrusively. It goes far beyond the traditional way of analyzing recorded events or big data analytics and observes the digital traces of each human-system interaction.
Opting for the superior process discovery means you can focus on observing the work as it happens rather than on the recorded events.
Process discovery software brings together the power of human-level observation and machine-level precision. You can capture each nuance of work that could be missed out on in recorded data or event log analysis.
Here’s an example to understand how process discovery works:
An operator works on a process. As a part of the work, he does some calculations in Excel. Or copies a field from A and pastes it to B in the sheet. Not being a part of the transactional processing system, this work on Excel wouldn’t be recorded in the event logs. Thus, this work, which might be crucial for the process, will go unrecorded.
There are many systems and applications similar to Excel that are beyond the purview of the central transactional systems. Thus, many things might go unrecorded when you rely on transactional ERP systems and their recorded event log for process analysis.
If you are still working with a process mining tool, it would undoubtedly, do the following:
- Record the events in the transactional processing system
- Give you an analysis of the event logs
- Show trends and patterns of work
- Detect process violations
- Monitor process conformance
However, it would refrain from continuously capturing the important process steps outside the scope of the main system.
Process discovery goes beyond hard data commits and event log analysis. It captures what happens even between these recorded events. How the process participants interact with the system and applications beyond the ERP system’s purview.
It observes everything.
Compiling the digital exhaust from each human and machine interaction, it paints a holistic and more accurate picture of the process. Thus, enabling businesses to prepare for the next step in their digital transformation journey.
Preparing For Process Discovery
While stored data in the form of an event log is the foundation of process mining, what holds the base of process discovery is observation using computer vision and analysis using Machine Learning or with the help of a Data Scientist.
It observes the work in real-time to tell a process ‘as-is’. All you require is a digital product/system and process discovery software. It doesn’t even need any back-end or API integration.
A non-intrusive lightweight probe is kept on the desktops of process participants to capture the human-digital interactions. This acts as a fit-bit for your enterprise and captures each nuance to ensure conformance. It can continuously monitor the overall process health, not just from a particular process perspective.
The crowd’s resistance to change could pose a problem. So, you also need a powerful product team ready to change. Before launching any process discovery, you must clear all air around intrusion on privacy. They may have inhibitions around the complete observation of work or screen recording. However, this can be cleared. Here’s how.
A process discovery tool is designed to observe the process metadata instead of what’s on the screen. It also offers selective screen masking to safeguard private data and information. It can even be taught to differentiate between process work and personal work. It would also mask the private data on the screen and read it as boxes without accessing information inside.
By making things clear, you can have a team ready for the digital transformation journey.
As the process lies at the foundation for your re-engineering, automation, or digital transformation endeavors, having an accurate picture is imperative. The methods used before gave incomplete visibility to the process.
Process analysts were limited to drawing boxes and arrows by shadowing employees. Process mining was limited to the event logs. The variants, nuances, and subtleties of actual human work on the system could not be captured in the picture using these techniques.
A cognitive process discovery software can help observe processes in real-time with documentation of every little nuance. It not only observes work but also uses Machine Intelligence to analyze it and create workflows. It creates the enterprise digital process twin or process meta-model after analyzing in-between the committed states of data. And this is the accurate picture any process owner needs. This visualization is verifiable and holistic for any intervention. With this visibility at your disposal, you can aim for automation, precision training, Kaizen, or any other attempt to digitally transform.