In our continuously changing world, it is crucial for business processes to be highly adaptive and become more efficient and cost-effective. As a consequence, companies require an increasing degree of process automation to stay competitive in their markets. A promising approach is provided by Robotic Process Automation (RPA), which aims to automate business processes or parts of them using software robots mimicking human interactions. Thus, an increasing number of companies run RPA initiatives. In practice, it is common that RPA projects are implemented by knowledge workers without an IT background. However, there exists little research on applying RPA to knowledge-intensive domains, which are dependent on the employees who perform decision-making tasks. In general, RPA projects often fail, therefore, we aim to sustainably support the project implementation.
In this project, a checklist-based framework for RPA in a knowledge-intensive domain, i.e., automotive, is developed. The research is based on design science methodology. In particular, we conduct interviews and distribute questionnaires to understand the effects that are achieved with RPA projects in engineering. Further, through an exploratory case study, we identify challenges in current RPA projects that lead to undesired effects. The current state-of-the-art is analyzed by a systematic mapping study (SMS). Using the framework resulting from the SMS, we determine challenges that have not been addressed in the literature so far. To overcome challenges, we develop and empirically validate an RPA user acceptance model as well as investigating desired human robot interactions empirically. For both research projects, we include 50 RPA users in the automotive industry as participants in the study. The findings are used to derive the checklist-based framework. The latter is evaluated along the framework for evaluation in design science, within four evaluation steps taking into account users from different industries and with different backgrounds.
To conclude, the artifact developed and evaluated in the CheckRPA project, improves the implementation of RPA projects in knowledge-intensive domains while satisfying the special requirements of knowledge workers and, therefore, represents a valuable contribution to practice and scientific literature.
