Design Analytics: Predicting Human Spatial Experience

SCI 6506 Subject & Catalog Number

Description

How do we measure human experience in space? Can we predict a design’s impact on human comfort, cognitive performance, or preference before it is built? As environments are increasingly interpreted by machines through sensors, computational models, and predictive algorithms, designers need new ways to observe, analyze, and evaluate spatial experience. Understanding where human and machine perceptions converge and diverge is essential to rethinking how environments are designed and experienced.

Drawing from concepts in architecture, cognitive science, and computer vision, this course examines methods for translating subjective human experience into quantifiable insights. Through lectures, critical readings, and conceptually driven projects, students will investigate how biological and artificial vision systems interpret space, and how those interpretations shape design outcomes. The course considers human perceptions of scale, depth, motion, attention, and memory alongside computational approaches including object detection, image classification, and semantic segmentation. These core methods will be complemented with emerging technologies such as vision language models (VLMs) and immersive virtual reality (VR), that offer new ways to capture and analyze experience.

We will interrogate the limits of both human and computational perspectives and examine how perception can function as a design input, an evaluative tool, or even a dataset. Students will learn to collect, interpret, and critique both human-generated and machine-generated data. Case studies in adaptive buildings, urban sensing systems, and AI-driven design workflows will ground discussions of bias, ethics, privacy, and the risks of relying on computational technologies to understand and shape environments.

Hands-on experimentation is central to the course. Students will learn to collect perceptual survey data using validated measures, simulate parametric design iterations with generative AI tools, and evaluate design decisions using pretrained VLMs and machine learning models. The course treats the use of existing computational tools as design frameworks and materials, rather than emphasizing technical development from scratch. Short in-class exercises will build progressively toward a midterm and final project. Together, these exercises and projects will equip students to work critically and creatively at the intersection of human experience, spatial thinking, and emerging technologies.

By the end of the course, students will be able to develop, critique, and deploy analytical tools that bridge subjective human experience and objective spatial data, offering new ways to quantify, evaluate, and inform the design of environments at both the building and urban scales.