Next to my blogging about my Triathlon career you will find some content about my research I am currently pursuing. My focus is Robotic Process Mining, which, in my opinion, does focus on automating the RPA project lifecycle. Instead of manual discovery of routines by running around and asking users their day-to-day routines and then implementing them manually, RPM tries to capture the behavior in an ad-hoc fashion and emulate bots automatically.

I particularly focus on the further progress of RPM after its early releases around 2020. First tools, such as SmartRPA (which I am currently supporting – GitHub), are already on the market. However, they struggle when there are (1) long time recordings with a lot of unnecessary behavior, i.e., noise, (2) interleaved or unclear delineations of user routines, and (3) different variants with different efficiency factors achieving the same objective.

RPM could increase the adoption of automation by discovering repetitive routines and emulating the automation of the most efficient, most ROI-worthy routine. To further increase the benefits of RPM the methods should look further than just RPA and discover automation possibilities in more technical ways. This can include, again in my opinion, the discovery of API integration solutions, the automatic emulation of Workflow-Management-System processes, e.g., directly generating the workflow for tools like ERP, CRM, ITSM, etc.