UdZ 01.21

UdZ 2-2021 / 29 6 PwC 2016, p. 20 7 Sadun et al. 2018, p. 36 et seqq. 8 Sadun et al. 2018, p. 36 et seqq.; Makridakis 1990, p. 35 et seqq. Process Complexity Degree of automation Use of Artificial Intelligence Robotic Desktop Automation Robotic Process Automation Cognitive Process- Automation Semi-automated Approach Fully Automated Approaches Figure 1: Evolutionary stages of process automation (Götzen 2020, p. 3) Various Types of Cognitive Biases Inhibit the Use of RPA Companies face the challenge of ensuring acceptance of RPA deployment while taking increasing AI into account. According to a KPMG study, adoption decreases as AI use increases 5 . In many companies, it can be seen that decisions are made on the basis of subjective criteria, despite the presence of automated insights 6 . Two specific influencing factors cause this circumstance: On the one hand, it is the conscious disregard of the knowledge gained due to a lack of acceptance of the same and unconscious prejudices. On the other hand, there are cognitive distortions, so-called biases, which inhibit the acceptance of automated decisions. Such biases can occur individually or in combination in different forms. Lack of acceptance can be explained, among other things, by managers and employees overestimating themselves (overconfidence bias) or by the complexity of automated decision-making 7 . Decision-makers are often unaware of how the results are determined with the help of complex algorithms. Unconscious influences in the context of RPA also include an excessive preference for the status quo over change (status quo bias), an inability to adapt one’s own thoughts to new evidence and information (conservatism bias), or a selective perception in which one ’ s own experiences and background create ‘tunnel vision ’ 8 . Approach in the Project Overcoming the aforementioned hurdles requires a holistic approach that createsworkforceacceptanceof RPA. For SMEs, this is a particularly challenging task due to limited human and financial resources. For this reason, the RPAcceptance research project is dedicated to the development of a model that encompasses the key influencing factors regarding the acceptance of RPA as well as its Intelligent Technology characteristics. These influencing factors will be empirically tested by a questionnaire study, in a behavior-oriented laboratory experiment and with an online study in order to developavalidatedchangemanagementconcept.Theconcept will include approaches to consider holistic, organizational and cultural transformation in addition to comprehensive training concepts. Through collaboration between the FIR and IPRI institutes and a large number of industry partners, the results of the research project will be developed with the help of a concrete process model . In the first step, the general acceptance of the employees with regard to different technological manifestations of RPA will be determined. The interfaces and interactions between the workforce and the software will then be identified in order to define the requirements for the human- software interaction interfaces. Based on the identified requirements, the acceptance model between human and RPA software will be developed as a key deliverable of the project. A reward-based experiment for validation supports the research process by motivating the addressed test subjects with appropriate incentives. Subsequently, training