Towards Asynchronous VR Collaboration with Virtual Agents as Placeholder Users

Immersive environments offer seamless 3D interaction, connecting even physically distant users in various domains such as education, entertainment, and healthcare. In immersive environments, users can interact not only with each other but also with computer-controlled virtual agents. Using advances such as natural language processing and artificial intelligence, virtual agents can understand, respond, and adapt to user input. Previous research has suggested using virtual agents as placeholders imitating absent users to avoid scheduling conflicts and promote asynchronous collaboration in multi-user environments. However, the potential of asynchronous collaboration is understudied, and the extent of imitation required by placeholder virtual agents to maintain the interaction quality between users is unclear. This paper explores the concept of virtual agents as placeholders for missing users in a multi-user VR training or guidance scenario, highlighting the motivation and the research gap. It also outlines the planned methodology to address this gap, starting with an ongoing literature review of current virtual agent research, followed by empirical user studies to evaluate interactions with agents.