Environmental and spatial dynamics in a flexible workspace for hybrid work: A data-driven design framework

• Workspace design in a hybrid working setup • A flexible co-working space in London as the case study • Geo-spatial association between spatial and environmental design metrics and occupants’ space preferences • Data-driven machine-learning-based seat occupancy prediction model • Space with high spatial connectivity and large visual field is preferred • Seats with a relatively higher temperature and daylight level have a higher occupancy level. The evolving landscape of hybrid working necessitates a corresponding change in workspace design, yet the understanding of the spatial and environmental design of flexible workspaces remains relatively limited. This study investigates the association between workspace design and occupants’ behaviour and preferences and builds a data-driven model to predict seat occupancy based on design features. Case study data collection process spans over a year in a hybrid co-working space in London, forming a multi-domain integrated dataset including sensor-detected occupancy, on-site measured and simulated environmental variables, computed spatial metrics and human-voted preferences. The combined effects of thermal comfort (temperature and humidity), visual comfort (artificial light and daylight exposure) and spatial metrics (connectivity and isovist field) are analysed to understand their collective impacts on space occupancy and user preferences. Geo-spatial regression and autocorrelation models are employed to explore associations at a grid level. Through cluster analysis, 61 seats are categorised to four levels based on their occupancy and utilisation rates, and a supervised classification model predicts occupancy levels using design variables. The study aims to present a data-driven analytical framework for evaluating flexible workspace utilisation and demonstrate predictive capabilities by identifying associations between environmental and spatial variables and occupancy patterns, offering a tool to guide evidence-based design and management of flexible workspaces. The analysis underscores the significance of spatial and environmental metrics, revealing that areas or seats with high spatial connectivity and larger visual fields tend to attract more occupants. Additionally, environmental variables, such as temperature and daylight level, exhibit positive effects on increased occupancy.