Michael Disch

2020

Robotic Process Automation (RPA) is an innovative approach within digital transformation, which refers to software bots learning manual tasks and thus automatically executing business processes. Due to limited experience and lack of empirical data regarding developing RPA software bots, many organizations struggle to implement them. Moreover, information about the necessary process steps, the required effort and the project staff are not available. Previous research approaches focus on the selection of suitable business processes for automation and the consideration of risks and advantages of RPA. For enabling a successful management of RPA software projects in organizations, including the application of Software Cost Estimation (SCE) methods, it is fundamental to analyze information about the necessary process steps, the project team and the project effort involved. In this context, a case study is conducted within a German automobile manufacturer. The results of the case study lead to a generic process model for the implementation of RPA software projects, derived from a process analysis, expert interviews and a survey. The model contains both sequential as well as iterative incremental process procedures. In the process model, standards and techniques such as PMBOK, RASCI and SCRUM are considered, and effort distribution patterns are mapped for both the process steps and the project staff involved. The results of this case study emphasize a clear differentiation of RPA from other software projects, which is apparent in the calibration of SCE methods. Furthermore, the findings of the case study are appropriate to enable the practical implementation of RPA software projects.