Next Level for Robotics
in R&D | Press

Next Level for Robotics

KI.RPA Researches Tools for Fully Autonomous Process Optimization

Darmstadt, February 04, 2019 – Robot-based process automation is currently establishing itself more and more as a central optimization technology for efficient, error-free and transparent company processes. In reality, however, the required configuration of the software robots has so far been associated with considerable manual effort - the actual and, above all, consistent autonomy of such systems has thus had relatively clear limits until now.

The KI.RPA research project, funded by the German Federal Ministry of Research and Educa-tion (BMFB) with a seven-figure sum, has made it its mission to change that: The 2 ½-year project has been exploring a promising solution to the above-mentioned problem for a few months now. The basis of the approach is a currently unique combination of Artificial Intelligence (AI), Robotic Process Automation (RPA) and Process Mining.

The consortium leader of KI.RPA is Servicetrace GmbH, a monitoring software specialist that contributes its expertise in the field of process opti-mization and robotics. Other project partners include the non-profit August Wilhelm Scheer Institute for Digital Products and Processes (AWSi), which is contributing its expertise as a research service provider, Process Analytics Factory GmbH (PAF) as a specialist and solution provider for Process Mining, the Telecooperation Lab (TK) at Darmstadt University of Technology, which is investigating the practical benefits of Process Mining methods, and Deutsche Telekom Service GmbH (DTS) as a practical partner, which has extensive experience with RPA projects already implemen-ted in companies.

On the way to the virtual employee pool

“Our Servicetrace robots execute cross-application and cross-media processes completely error-free and consistently. This, of course, gives employees more time for complex tasks and challenging projects," says Markus Duus, CEO of Servicetrace GmbH. The goal of the project is to develop a tool that will enable companies to create virtual employee pools based on robotics that support human employees in recurring tasks - without the limitations of today’s software robots, such as high configuration effort, rigidly prescribed procedures and low degrees of freedom in the process flow.

To this end, the project team is developing a novel software system that captures the data and work steps of various end-to-end processes directly at the human interface, analyzes them, then learns the process independently - and then even executes it itself in an opti-mized form. A vision of the future could become reality here: The closed system that acts completely autonomously and constantly optimizes itself without outside intervention.

“It is basically the dream of every industry: a technology that learns and acts independently. With our PAFnow Process Mining software, which can be used at any digital workplace, we provide the data-based technological framework for this and thus the system’s thought structure, so to speak," says Tobias Rother, CEO of PAF GmbH, about the role of the Process Mining process within the solution.

The core element of the tool is an AI that is able to examine the manual work steps of human users during the input mask on the basis of Process Mining technologies. The backbone of the software will be algorithms developed specifically by KI.RPA based on Deep and Machine Learning, as well as a Process Robotics Engine.

More efficient companies through better working conditions

In addition to technological innovation, the project also aims to give companies the opportunity to release employees from tedious routine tasks for more demanding activities - and thus simultaneously reduce costs, increase speed and lower the error rate. The potential behind this is enormous: as a rule, a software robot can relieve up to five employees of repetitive processes. As a result, humans can devote themselves to tasks that require their unique experience and decision-making skills. “Artificial Intelligence is revolutionizing classic office work. In KI.RPA, we develop intelligent, self-learning software robots that structure and automate processes. Companies become more efficient, employees can work more freely and creatively," says Dr. Christian Linn, Digitization Professional at the AWS Institute for Digital Products and Processes.

Billions in potential in numerous industries

Current market figures forecast a global market volume of around five billion US dollars for 2020 for process automation in the IT environment alone (source: statista, Transparency Market Research). In addition, there are high-potential application areas such as healthcare, sales, and administrative processes. The solution envisaged by the project partners is aimed not only at large companies, but also at small and medium-sized enterprises, which have so far been denied appropriate RPA solutions due to the cost-intensive need for external consultants.

For more information on KI.RPA, visit the project partners at:

Servicetrace GmbH:


Process Analytics Factory GmbH:

Telekooperation Lab:

The project is partly funded by the German Federal Ministry of Education and Research (BMBF), funding line KMU-innovativ: Informations- und Kommunikationstechnologien.