• en
  • sr

Natasha Bajc

Research laboratory for Sustainable and Resilient Architecture – SaRA is pleased to invite you to the guest lecture “What is the future of design technology?” by Natasha Bajc.

The lecture will take place online, on Thursday, April 29, 2021, starting at 19h CET.

To attend the lecture via Teams, please send your request to

What is the future of design technology?

Generative design has been a subject of research for architects and technologists since the marriage of design and software. Recent development in Artificial Intelligence (AI) and Machine Learning (ML) have exponentially raised the possibilities of intelligent design by software.

This work (2017) explored the capabilities of ML techniques to “learn” architecture. Learned from dataset of 2.000 floor plans Generative Adversarial Networks (GAN) create a floor plan in 1.7seconds or 15.000 in 8 hours. This “proof of concept” research on 280×280 pixel size floor plans, showed that there is significant potential in development of “Creative AI” and development of ML techniques since 2017 proved that GAN’s are viable tool for design.

A footnote to GAN research there is area for research using technique of the Autoencoders, that fill in missing area of the floor plan generating best guess according to what is on the plan.


Natasha Bajc is an architect and technologist with intertest in Artificial Intelligence and Affective Computing (Emotional AI). Her career bridged running a commercial practice (Urbis Design and CodeDesignStudio) and academic and research: teaching at UCLA, SciArc and Woodbury.

Natasha has an undergraduate degree from University of Belgrade, Master of architecture from Sothern California Institute of Architecture (SciArc) and is very close to a Master of Computer Science from USC. She is a winner of Radical Innovation in Hospitality award as well as European Hotel Design award for Tomorrow’s hotel, her work has been exhibited in USA, Serbia and Russia. Her current projects in artificial intelligence envision smart systems for design and deployment at homes, buildings and neighborhoods.

Imagine that you find genie in the lamp. THE GENIE. She is mighty powerful, but has a bit of temper, after all she spend several centuries locked in the bottle. She is not up to date with things.  The other night you had to instruct her how to make sushi: what type of rice, how you like it flavored, how to roll it, fish, the whole thing… now you want to instruct her to design a perfect house for you. What does she have to learn?

Todays genie comes in the form of artificial intelligence (AI) programs. She learns to construct a floor plan in 1.7 seconds, but is that plan any good?

 My interest is in constructing systems of measurement that would suite your personality, accommodate for your lifestyle and generally be able to evaluate living/working/playing spaces. I’ll present to you my research and demonstrate a Virtual Reality app used as platform for data collection that can begin to provide some results in spatial evaluation.