{AI} Urban Morphology
For almost a century, cities have developed and grown in scale partly to manage and realize the advancements of technology, transforming and evolving both our natural and societal environment.

However, as cities grow in size they become increasingly hard to manage and understand. Our current methods of planning: a mixture of top-down planning and commercially incentivized project development has its limits in addressing the changes needed for our society as a whole to grow and developed sustainably. Other than economic and environmental pressures, the aesthetics of many cities are becoming diluted by dry and ubiquitous planning codes and profit-driven designs.

To realize the design required for today’s world, we need a deeper understanding of the morphology of cities in order to respond to the challenges facing urban development. We need the ability to utilize data and new technologies in a more creative and functional way.

This focus of the project is on creating large scale planning solutions for the city of Xiong An by using Deep Neural Networks for algorithmic style transfers. Aiming to understand how Style Transfer can affect and inform the designer’s abilities, and provide a tool that allows more possibilities to be explored for a given site.

Methods

In this research, I created intermediate styles to explore how the color and form of patterns can affect the Style Transfer process. First I have generated various versions of different patterns, and overlaid them to create more intricate patterns with more details.

These patterns would first be superimposed to generate a base plan for the whole area, then these patterns would also be used as styles to create different versions of the base plan.

Styles prepared for transfer

Base plan

Applying style to base plan

Applying Urban Texture

The city plan stated the desired population density of 10,000 people per square kilometer. To generate results that are similar to Xiong An’s requirements, I choose Barcelona with a 15,000 p/km population density as the texture to be applied.

Full Maps

Different results are produced through using one base pattern and adjusting the intensity of four style intermediates.

Comparing between various results

3D maps

The ability to capture colors and forms in both style and context renders Style Transfer a useful tool that could achieve something people cannot, giving us insight into something unknown.

The wide range of results generated by Deep Neural Style transfers, from a design perspective, touches on the paradox between top-down planning and organic growth. With the speed and adaptability of AI and ML, bottom-up design from utilizing raw data can allow for much more flexible concepts than before.

Thank you !

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s