Leveraging Machine Learning (cGANs) to Develop Real-Time Inverse-Simulation based High-Performance-Building-Design Optimization Workflow

MASTER OF SCIENCE IN SUSTAINABLE DESIGN THESIS RESEARCH AT CARNEGIE MELLON UNIVERSITY, 2021

“…I have not seen this done before and I thought that it seems really original…you showed us how you can work the maze backwards - And I thought that was really interesting that it was both an original idea, and that it worked! So bravo!!..” - Holly Samuelson (Harvard University)

“…I have never seen pix2pix and solar radiation analysis or grid-based analysis that way, brought together before. I really like the idea I think it's conceptually challenging. I love the idea of reverse engineering that analysis, and more importantly I think that it encourages design freedom and opens up what you design and what design intuition can mean, so I really liked that aspect of it…”  - Putu Marantha Dawkins (Carnegie Mellon University)

Publication: Carnegie Mellon University Kilthub

Timestamps for the video presentation:
0:00 - Thesis Presentation
24:52 - Panel Critique [Panelists: Holly Samuelson (Harvard University), Putu Marantha Dawkins (Carnegie Mellon University) and Joshua Bard (Carnegie Mellon University)]