Enabling Algorithm-Assisted Architectural Design Exploration for Computational Design Novices

This research introduces a workflow based on a prototype tool that automatically encodes a massing design for exploration using a genetic algorithm. The workflow has been customised for an architecture class attended by computational design novices. It streamlines the process of running an algorithm-assisted design exploration, where students only need to (1) construct a 3D massing model, and (2) specify the geometrical parameters and their value range to be explored by the algorithm. The prototype tool uses the two inputs to automatically encode the design as a genetic problem and explore the design space with respect to density and solar objectives. The workflow is used by the students to complete a simplified massing design exercise. The workflow’s feasibility and effectiveness are evaluated and shown through usage data and feedback collected from the students. As the workflow and the tool are openly available, they have the potential to be adapted for use in other design scenarios.