Predictive Atmospheres
Project Overview
About
Buildings can influence mood and well-being, but architects have limited ways to assess these effects while developing a design. Predictive Atmospheres uses machine learning to connect features such as room dimensions, windows, doors, floor area and lighting with emotional responses measured through brain activity, heart rhythms and participant feedback. Integrated into 3D design software, it provides a real-time “Neuro-Score” as layouts change. The framework could help architects compare options and create spaces better aligned with psychological comfort and well-being.
Team
Manideep Mamidala
I am Manideep Mamidala, an architect, computational designer, and neuroarchitecture researcher holding a recent master's degree from the Institute for Advanced Architecture of Catalonia (IAAC). My work operates at the intersection of spatial design and cognitive computing. Leveraging Python and advanced parametric workflows, I build machine learning frameworks—such as Predictive Atmospheres—that map objective environmental metrics to human physiological comfort. My core objective is to develop computational tools that predict the emotional impact of unbuilt spaces, enabling the creation of environments rigorously optimized for human well-being and cognitive performance.