Unit 4

Individual cities: Utopia

Main Lecturer: Casey Rehm

Unit Instructors: Dongyoon Kim + Jaewon Yi

Image credit: Zane Mechem and Chengyu Zhang at SCI-Arc's 2B studio with Casey Rehm

1. Brief

Covid-19 has not allowed us to do urban research from the site. Site researches such as interviews, field surveys, and engagement design are precluded although this method is effective to discover real city conditions. We ask how to meet the people and collect social structure in ‘un-tact’ society.

In this context, artificial intelligence technology could replace conventional urban research and then generate new methodology. The embracement of automation in aspects of the design such as working with different forms of machine intelligence and synthetic intelligence within the design process allow us to rethink how we would interpret the context and look for novel forms of organization or aesthetic that emerged out of the way non-human perception engages things.

Because conventional architectural design process is rarely working with ‘tabula rasa’ so machine which has the ability to respond to contextual inputs could show us a novel way to read the city. Individual cities: Utopia is a speculative narrative about how these transformations would occur, redesign, and renovate the entire city from non-human.


2. Tools

  • Get data set from Flickr using Python script [input (ie. "#labels, #hashtag ....")]

  • AI training: Train a Neural Network (NN) using the data set

  • Go out and take photos of the city for photogrammetry scan (Photogrammetry is the science and technology of obtaining reliable information about physical objects and the environment through the process of recording, measuring and interpreting photographic images and patterns of electromagnetic radiant imagery and other phenomena.)

  • Use the NN to train the photos -> resulting outcome is newly trained photogrammetry scan

  • Use the newly trained images to produce 3D model (using extra software such as Agisoft(https://www.agisoft.com/) / Zephyr (https://www.3dflow.net/3df-zephyr-photogrammetry-software/) ...)


3. Aim and Scope

  • Discover what factors make style and identity

  • Find keyword to collect images using python script

  • Learn machine learning process and apply it to design proposal


4. Expected Workshop Outcome

  • A design proposal