Integrating Biometric Sensors, VR, and Machine Learning to Classify EEG Signals in Alternative Architecture Designs

Design of office spaces plays an essential role in people’s day-to-day work productivity. Research in environmental psychology and neuroscience indicates distinct architectural design features (e.g., color coding, texture, and space layouts, etc.) impact human performance and motivation to work in office spaces. In the current practice, occupants evaluate work space designs via after-the-fact post-construction surveys subjectively. Limited studies exist in the literature on objectively quantifying motivational impact of space design on occupants. This research stems from the need for having objective ways to assess human experience in the built environment for design improvement. Integration of electro-encephalograph (EEG) and virtual reality (VR) equips researchers with the tools to measure human responses when subjects are immersed in alternative virtual designed spaces. This study proposed a machine learning based method to label subjects’ experience in spaces using their EEG data collected when they were in distinctly designed spaces. Results showed this method provided around 85% classification accuracy, which is comparable to other state-of-the-art EEG classification methods. Practitioners in the architecture engineering and construction (AEC) domain can use this method to identify if proposed design options have positive or negative impacts on future occupants.