Houy, ConstantinHamberg, MaartenFettke, PeterRäckers, MichaelHalsbenning, SebastianRätz, DetlefRichter, DavidSchweighofer, Erich2019-02-222019-02-222019978-3-88579-685-5https://dl.gi.de/handle/20.500.12116/20517Against the background of current activities towards administrative modernization based on the digitalization of processes, the usage and integration of Robotic Process Automation (RPA) software into public administration work processes can significantly improve their efficiency, reduce process costs and provide better services for citizens. This paper presents and investigates the concept of RPA and discusses the particular potential and challenges of RPA in the public administration context. Furthermore, it demonstrates an application example of a new cognitive RPA approach for automated data extraction and processing that is used in a trade tax assessment scenario using deep convolutional neural networks (CNN). Based on the findings it can be concluded that RPA has considerable potential for the improvement of the efficiency of admini­strative work processes and for administrative modernization in general.enRobotic Process AutomationRPACognitive RPABusiness Process ManagementBPMDeep LearningConvolutional Neural NetworksText/Conference Paper1617-5468