Building envelopes are extremely significant in providing adequate indoor environment. They have tremendous impact on the energy requirements of buildings. The design methodology associated with building envelopes primarily addresses optimization for improving indoor environmental quality and energy efficiency of the buildings. This process does not account for the variance between occupant preferences and their importance on the various indoor environmental quality domains. The design of a building envelope has been found to significantly impact the well being of building occupants. This research proposes a user-centered design approach that evaluates the factors influencing occupant comfort and preferences. To achieve this, facade user-archetypes are employed to personalize building shading systems for users.

The multi-domain impact of building envelopes and external shades is studied to determine the environmental domains associated with shading systems. A classification scheme is developed for shading systems on the basis of their operation, placement, interaction and permeability. Next, shading system parameters are evaluated through geometry, materiality and control to understand which design parameters have the highest influence on occupant comfort and energy performance. To accurately capture the multi-domain influence of shading systems, the shading systems are simulated within a model space using the EnergyPlus and Radiance engines. The simulation results are stored in a data-set that cross evaluates shading system performance across 8 orientations and for occupants at specific spacing from the window.

A systematic literature review is conducted to identify factors impacting occupant preferences and current clustering methods for user archetypes. Based on this, an occupant preference framework is created and used to design a questionnaire. The questionnaire is distributed to office workers and individuals in different settings to evaluate their preferences. The responses received from the questionnaire are analysed using correlation and ANOVA test is used to evaluate which occupant characteristics show a higher correlation to certain preferences and environmental preferences. Based on the results, feature set iterations are developed which are processed further for dimensionality reduction. The feature set that captures the maximum occupant characteristics with a reliable explained variance is clustered using hierarchical clustering and K-means clustering algorithms.

The clusters resulting from the analysis form the archetypes, which are then utilized in design scenarios. The weights and preferences of the archetypes are incorporated to determine the most suitable shading system for the occupants. Scenarios are developed to use supervised / semi-supervised learning methods to predict the archetype of new users based on existing archetypes formed.

The findings demonstrate a high accuracy of the archetypes in recommending shading systems based on the assigned environmental importance and visual preferences of individual users. The research highlights that each user has unique preferences, which can lead to different design recommendations based on their responses. Furthermore, the research showcases the practical implementation of archetypes in designing spaces and emphasizes their potential application in future facade design and control systems.