What actually is process mining and how does it help my company?
Summary Process Mining
Process mining is the systematic analysis and evaluation of business processes. To illustrate this, let’s imagine a mechanical engineering company. Typically in a process mining project you would initially try to concentrate on one business area.
This could be production, for example. A classic medium-sized company in the SME sector would probably produce at one location and would be dependent on purchasing.
How do I approach process mining in my company?
The first step in process mining would have to be to record, document and understand all the processes that take place in this department in a lean consulting project in order to gain an overview of the content.
It is generally advisable not to create 200-page documentation and describe every little thing in detail, but rather to concentrate on good visualizations. But the most important thing is: focus on processes that generate data or can/should generate it in the future.
Process mining with PowerBI
Once the overview is there, the technological implementation would begin. Ideally, you need a BI tool such as PowerBI from Microsoft & pipelines / connections to make the process data usable for PowerBI. As we move towards PowerBI, the data in the pipelines must be transformed to ensure that the data from potentially disparate sources is usable in a consistent manner.
This is often where a large part of the effort lies - defining what data you need, finding out how to get to it and implementing the technical integration. Once the process data is in PowerBI and the pipelines are in place, visualizing and analyzing the data in the tool is child's play for an experienced data scientist.
The big advantage here: Tools like PowerBI are often so easy to use that even departments can investigate the causes with a little training. Let’s go back to our SME example from the beginning. After identifying the processes, setting up the pipelines and visualizing the data, we have now achieved the following added value for the company: We know the process duration and effort involved in each step in production, understand where bottlenecks arise and what costs us productivity and money.
Even better: Through our pipelines we are able to process live data and have also created a complete monitoring & alerting system (with minimal extra effort). This means we can react quickly to problems in production, stop them in their tracks and clearly identify their causes. But that's not all... based on the foundation created here, we can now start optimizing. We can change and adapt processes and not only know in gut feeling whether the new version is more efficient, but can prove this based on data!
And the best comes last... with a little programming effort, we can even integrate AI algorithms to continuously optimize our production fully automated and taking into account all historical data.
So does process mining make sense? Yes!
The next expansion stage would be, for example, to integrate other departments and recognize the connections. Purchasing, for example, will be very closely linked to production. There is a lot of further potential for optimization here.