Towards a Framework to Model Intelligent Avatars in Immersive Virtual Environments for Studying Human Behavior in Building Fire Emergencies

Abstract

Driven by the fast development of virtual reality (VR) technologies, immersive virtual environments (IVEs) have been frequently used to conduct human behavior experiments for studying human behavior in building fire emergencies. Avatars in these IVEs are usually used to provide social influence and improve the sense of presence experienced by participants. However, limited intelligence of avatars in prior studies significantly lowered the level of sense of presence and reality experienced by participants. Improved intelligent avatars (IAs) are needed for developing high-quality building fire IVEs. A framework for modeling IAs to support the investigation of human behavior in building fire emergencies was proposed in this study. A number of levels of IA intelligence were defined based on the characteristics of avatars in VR for studying human behavior in building fire emergencies. This study also proposed a roadmap to achieve each of these levels of intelligence. A case study was presented to demonstrate how the framework could be used to guide the design of IAs for research purpose. It was concluded that applications of IAs in VR experiments could benefit the investigation of human behaviors, crowd simulation in building fires, and even fire safety design of buildings.