Requirements for an AI driven Extended Reality assistance system for assembly tasks

Nine qualitative expert interviews were conducted to assess the requirements for an intelligent assistance system for virtual assembly in extended reality applications. The research interest lies in understanding current industry and research needs for assistance features of AR, VR, and MR applications that support manual assembly tasks. Additionally, the use of artificial intelligence as a supporting technology was investigated. We gathered expert opinions on relevant features and grouped them into seven categories: Augmentation of information in-situ, identification of components, provision of data, analysis of user action, analysis of environment, optimization of processes, and Usability improvements. The experts feedback on the concept of a software demonstrator was collected and structured. The research shows a need for adaptive assistance systems that are capable of categorizing the user’s actions and intent in order to assist efficiently.