Demand analysis for an augmented reality based assembly training

The first head-mounted display (HMD) was developed in 1986 by Ivan Sutherland. Since then, augmented reality (AR) applications are largely limited to prototypes. One reason might be the lack of user comprehension regarding user requirements. In order to develop a HMD-based training system for daily use, it is important to understand user demands. This paper aims at presenting a demand analysis for an industrial engine assembly line at the car manufacturer BMW Group. We describe the background of training within industries (TWI) and present extant AR-based training studies. We use the value proposition design method to analyze our customers and classify our requirements with a common quality management method, the Kano Model. We find four "must-have" requirements for a HMD-based training system, which are crucial for the development of a minimal viable solution. In order to familiarize assembly employees with augmented reality applications, we design a simple demonstration for the Microsoft HoloLens. Additionally, we present future research directions focusing on comprehensive user studies in real industrial environments.