Optimizing visual comfort during architectural design: A tool for view and window analysis

Abstract

New building performance simulation tools are advancing the building design process in the Architecture, Engineering, and Construction (AEC) industry to conserve energy, reduce operational costs, and enhance occupants' well-being and welfare. Most of the current tools focus on energy efficiency, thermal comfort, indoor air quality, and daylight analysis. However, visual comfort - especially view analysis and accurate prediction of view quality - remains under-prioritized, despite evidence linking high-quality window views of natural environments with improved psychological well-being, reduced stress and impulsivity, and increased workplace satisfaction and productivity. This research systematically reviewed current building simulation tools, identifying critical gaps and proposing a new tool that leverages satellite imagery and machine learning to assess windows and site surroundings, and predict view quality. The acceptability and appropriateness of the tool were evaluated through a survey that collected insights from design experts, along with a video demonstration that helped participants envision how a view analysis tool would assist their design process in Revit. The findings highlighted the potential for enhancing workflow efficiency, facilitating informed design decisions, optimizing the early-stage design process, and ultimately improving design quality and occupant well-being. This paper presents a proof-of-concept for a prototype tool that evaluates window views by integrating advanced machine learning and computer vision techniques to analyse the satellite imagery and environmental data around a building site with data from building information models (BIM), enabling architects, and designers to visualize the views of a window, optimize window size, and placement during design and ultimately ensure occupants' comfort and well-being.