Universal Design Spaces

Parametric modeling and design space exploration are increasingly used in early design. While many approaches have been proposed, challenges remain. One challenge is maintaining the right model resolution to track with a natural design process. In this work, we focus on creating “universal design spaces” and corresponding exploration methods for repeated design problems. We are generating large datasets for common design problems (universal design spaces), such as a daylight room on a facade, that can be customized for the needs of multiple projects throughout various design stages. To support customization, we propose methods for sensitivity metrics and prediction on subsets of universal design spaces.

Files

Metadata

Work Title Universal Design Spaces
Subtitle Early design approach + exploration methods
Access
Open Access
Creators
  1. Laura Elizabeth Hinkle
  2. Nathan Brown
  3. Leland Curtis
Keyword
  1. Parametric modeling
  2. Model-Agnostic Meta-Learning (MAML)
  3. Machine learning
  4. Architectural design
  5. Design Space Exploration (DSE)
License CC BY-NC 4.0 (Attribution-NonCommercial)
Work Type Poster
Publication Date September 23, 2021
Source
  1. Fall 2021 Stuckeman Research Open House
Deposited February 22, 2022

Versions

Analytics

Collections

Work History

Version 1
published

  • Created
  • Added Creator Laura Elizabeth Hinkle
  • Added Creator Nathan Brown
  • Added Creator Leland Curtis
  • Added Universal Hinkle.pdf
  • Updated License Show Changes
    License
    • https://creativecommons.org/licenses/by-nc/4.0/
  • Published
  • Updated
  • Updated