Design Strategy of Multimodal Perception System for Smart Environment

The construction of a perceptual system in a smart environment is fundamental for the building space to understand the state of its surroundings and the need of its users, as well as provide personalised services to them. A properly constructed perceptual system can effectively improve the intelligence level of the smart environment and optimise resource utilisation, minimising unnecessary waste by reducing the number of sensors. This paper will analyse how to build an optimal intelligent perceptual system for architectural spaces with the help of knowledge in human physiology, cognition, and information science. It will propose possible technologies and design strategies to improve the efficiency and accuracy of information perception in a built space through eight case studies and that the perception system in a smart environment consists of two levels: sensory and cognitive. The first level is the sensory level, that is, the information of the external environment is acquired through the sensory organs and is mapped from the external stimuli to the subject. To construct sensory-level perception in a smart environment, this paper adopts concepts from human physiology to build four categories of perceptual systems for architecture—visual, auditory, haptic, and olfactory-gustatory—and explores the possible ways to implement each system design with specific case studies. The second level is the cognitive level, that is, the sensory information obtained in the first level is analysed and processed by the brain to form meaningful signals or visual images, which are cognitively comprehensible to the subject. To better understand the users' needs in the space, this paper advocates the construction of multidimensional information perceived on a multimodal sensory system as a way to enhance the cognitive ability of the built environment. We also introduce systematic methods for establishing multimodal sensory systems, design strategies for the spatial layout of various sensory systems, and applications of machine learning algorithms to understand environmental information of the users' real-time demands.