PUB - Publikationen an der Universität Bielefeld

Artificial Intelligence (AI)-based applications promise great potential benefits for companies. However, an isolated consideration of the technical system is not sufficient for the design. Rather, it is necessary to design the entire work system taking into account the socio-technical system approach. This enables the combination of the strengths of people and intelligent systems. This paper presents an approach for a socio-technical requirements elicitation in the design of AI-based systems by adapting the HTO-analysis. First, a mission statement is developed. Based on a detailed process modelling, existing data and systems are recorded. In addition, all relevant stakeholder groups are included by conducting interviews and surveys. Thus, the procedure enables the derivation of a comprehensive catalogue of requirements. The application of the approach is illustrated by using an example from industrial practice, the design of an intelligent workforce planning system.