New paper on contextual drivers of building occupant behavior
© CYBERPINK - stock.adobe.com
Our new paper in Building and Environment, led by ETHOS visiting PhD student Tarun Verma, leverages supervised machine learning to better understand and predict human interaction with window blinds — a key factor influencing building energy consumption and occupant comfort, especially in developing countries.
Blind use patterns significantly impact visual comfort and energy efficiency in office buildings. To quantify this impact on building performance, blind control occupant behaviour models are often integrated into building performance simulation during building design. Existing models mainly focus on environmental factors, and integrating the complexities of these models into simulation generally requires familarity with complex simulation tools.
However, in developing countries, there is less adoption of simulation tools in the industry, which can limit understanding of occupant interaction with building features such as blinds. Moreover, existing occupant behavior models overlook contextual and time-related factors due to data collection and quantification challenges.
To address these gaps, this study proposes a novel approach to modeling blind control occupant behavior grounded solely on contextual and time-related factors — using both mixed effect logistic regression and random forest models. We conducted a six-month longitudinal field study in forty-two offices in Tiruchirappalli, India, collecting 935 blind state observations from 87 participants. The analysis revealed a preference for closed blind (73.8%) and identified relationships between blind/shade use and key contextual and time-related factors, including the size and orientation of windows, the building floor, and the building's ventilation mode. The proposed models exhibit convincing predictive power despite relying only on contextual factors and not including environmental data such as sunlight and outdoor temperature. Our models will help architects and building engineers to make informed early-stage building design decisions without relying on complex simulation tools. This study also contributes significantly to window blind research by identifying important relationships between contextual factors and blind use.
Funding is gratefully acknoweldged from the Federal Commission for Scholarships for Foreign Students (FCS) for awarding the Swiss Government Excellence Fellowship, which enabled this research at Smart Living Lab, EPFL. We also appreciate the support from the National Institute of Technology Tiruchirappalli, MNIT Jaipur, and EPFL, Switzerland.