Point Cloud Tectonics: Data-Driven Wall Structures in Design and Architecture

This course offers an exploration of data-driven methodologies in architectural design, emphasizing the potential of point cloud data and AI augmentation to redefine how wall structures are conceived, optimized, and articulated. Building upon contemporary research in digital tectonics, computational geometry, and AI, the course is designed to engage students from all backgrounds—no prior experience in computational design is required. The aim is to demonstrate how emerging technologies can inspire creativity while addressing architectural challenges of efficiency, material economy, and aesthetic expression.

Students will engage with point cloud manifolds as a foundational concept, learning to represent spatial geometries through discrete data points. The course introduces the process of manifold collapsing to simplify complex datasets, optimize structural logics, and aggregate clusters into modular wall components. A structured volumetric workflow underpins the course, providing a clear, accessible method for translating abstract data into tangible architectural structures.

To enhance the aesthetic and expressive potential of these designs, the course incorporates AI-driven style transfer, building on recent advancements in machine learning to augment tectonic expressions without compromising functionality. Through hands-on projects and theoretical discussions, students will explore how computational and creative processes can converge to create innovative architectural solutions.