RPA and Process Mining – The perfect combination
in Learning | PAFnow | Partner

Process Mining & RPA

A roadmap for bots

Author: Gerrit Kohrs, Head of Business Unit BPM, celver AG

The hype surrounding Robotic Process Automation (RPA) continues unabated. A number of companies have already achieved major efficiency gains by automating repetitive tasks in their business processes. Bots take over the rather simple activities of humans, allowing them to focus on higher-value tasks.

pafnow process mining software with bot

Challenge: Targeted implementation

It is still difficult to transform these initial successes into sustainable value-creating solutions. On the one hand, this is due to the fact that especially in the initial phase of implementation and in RPA pilot projects, the activities that are automated are those that can be mapped comparatively easily by a bot. On the other hand, companies face the challenge that it is only through the use of the bot that data on the process is collected in the first place. Users only find out afterwards whether the use of the respective bot is worthwhile.

We can only use RPA technology in a truly efficient and targeted manner if we understand our processes in detail, know their maturity level, and can in advance identify activities that can be executed well by a bot. For this purpose, reading system and process documentation as well as workshops with the business departments are only suitable to a very limited extent. Without IT support, it is almost impossible to analyze the thousands of process variants and examine their automation potential. This is where Process Mining comes into play.

Process Analysis - The Roadmap for Bots

Process Mining uses already existing log files from different systems to illustrate the overall process in all its characteristics on the basis of each individual transaction, even across system boundaries. Various process KPIs allow the classification of the maturity level. The identification of bottlenecks, errors and particularly frequently recurring activities allows conclusions to be drawn about the optimization potential - including activities that can be effectively automated. It is not without reason that around 44 percent of companies see Process Mining as “a tool to prepare for process automation and the use of Robotic Process Automation”.

idg studie gründe für process-mining
Reasons for using Process Mining: 1) Optimization/Standardization of as-is processes; 2) Increased visibility of as-is processes; 3) Preparation for (robotic) process automation; 4) Monitoring/control after (robotic) process automation; 5) Reducing costs, Source: Study Robotic Process Automation 2020, published by IDG Business Media GmbH, industry-independent survey in the D-A-CH region

The successful introduction of RPA depends above all on these factors:

  • Building capacities and know-how (keyword Competence Center) and/or selection of a suitable external partner.
  • Identification of a suitable process
  • Standardization of the process / reduction of process variants
  • Prioritization of the activities to be automated
  • Monitoring of results, ROI tracking

Companies should clarify the first point before starting an RPA project. Whether an interdisciplinary competence center or an IT- or department-driven project organization is the right thing to do cannot be answered in general terms. However, it is important to provide sufficient in-house capacity from IT and the business unit within the company and most importantly to establish the most detailed possible knowledge of the company’s own processes in addition to a basic understanding of the technologies. If these prerequisites are met, external partners can be integrated more efficiently. According to the IDG study, companies rely on both their own experts and experts from consulting firms in roughly equal proportions (50 percent each).

The other points sound harmless at first, but they also have their own challenges. The positive news is that all these issues can be managed well with the help of Process Mining. The extent to which a process is suitable for automation in the first place becomes clear relatively quickly via the analysis of homogeneity (number of variants), conformity (compliance with standards) and frequency. (Sub-) processes that run only very infrequently and/or very individually may contain activities that a bot could take on from a technical point of view. However, the leverage for the organization is then quite small. In the same way, there are activities that occur relatively frequently and could be automated but only with a small impact on capacities and costs. This means that a large number of candidates are eliminated at the latest in the ROI analysis. Finding the right prioritization is therefore extremely important, but fortunately it is also very simple using suitable Process Mining analyses. The standardization and monitoring of processes, in turn, is one of the core competencies of Process Mining. In summary, Process Mining can have a great influence on all of the above factors and is therefore ideally suited for the preparation and support of RPA projects like no other tool. In short: Process Mining gives the bots the orientation they need to become successful.

Platform instead of isolated software

From a user perspective, process analysis and automation - including planning, design, training and monitoring of bots - ideally form a single unit or are mapped via the same platform. This prevents duplicate data storage and gives users the opportunity to work in a uniform, familiar environment. Tool manufacturers have recognized this and expanded their portfolios accordingly through acquisitions or based their products on an established platform from the outset.

Dr. Timo Nolle, CTO of Process Analytics Factory GmbH, explains: “Due to the close interlocking of Process Mining with RPA in terms of content, it is becoming increasingly important that both technologies are available to users without system interruptions. The decision to develop our product on the basis of the Microsoft Power Platform from the outset was therefore spot on."

In summary

RPA projects promise quick wins by freeing up capacity for more demanding activities and absorbing high workloads, reducing errors, and ensuring compliance. Cost savings can also be realized with RPA. However, how well these potentials can be exploited depends less on the chosen tool or technical expertise but stands and falls with the selection of the right processes and activities. This selection can hardly be made correctly without data-based decision support. Process Mining as a key technology closes exactly this gap and should therefore be an integral part of all RPA efforts from the very beginning. This way, we get a roadmap for our RPA-projects and are not among those who drop out mid-project because they were headed in the wrong direction.

About celver AG

celver AG supports customers regarding the topics of planning, analytics and smart data and assists them from technical consulting to process definition and complete integration into the system landscape. With innovative methods and technologies, they build custom-fit solutions together with their customers, which form the basis for intelligent and data-driven decisions.

Find out more at celver.com