Week 3: Dataset Collection

For this project I need to collect images of existing bus stops in order to generate new designs. The new designs that the model comes up with will be determined by the types of bus stops that I use in my training data. I’m interested to see how the outcomes changes based on the different types of images I include and would like to play around with different datasets.

I searched for the following terms in Google and was able to download approximately 1,000 images to my dataset:

A sample of the images in my dataset

• “art bus shelter”

• “bus shelter”

• “bus stop”

• “modern bus stop”

• “soviet bus stop”

I spoke with my professors Ayo, Nakita and Yining about the best methods and heard from all of them that using a GAN (Generative adversarial network) would be the best approach. I talked to a classmate, YG, who had recently used a GAN for a project and he pointed me towards stylegan.

My next steps will be to collect more images, clean them, and try to run them through sytlegan!

Week 2: Project Proposal

Based off of my ideation process, I have decided to generate new bus stop designs using machine learning. I would like to use ML to imagine more imaginative designs for our built urban infrastructure and prompt questions about the process of how these objects are designed, by whom and with what priorities in mind. I’d also like to explore questions of how ML and AI - both digital/virtual tools - are changing our build landscape. I’ve been thinking about this topic since I saw a TED talk about how algorithms shape our world. In looking for this again, I also found a similar talk about the internet specifically changing lanscapes.

An example of how LinkNYC is commonly used

My original desire was to re-design the LinkNYC kiosks, but there aren’t enough of similar objects worldwide to create a dataset large enough for machine learning. I was motivated by these structures because I always wondered why they had to be the same boring color and shape as the sidewalk - there are already enough gray and shiny rectangles in this city! There are a bunch in my neighborhood and people drag out crates to sit around them, charging their phones or browsing the internet. Why wasn’t a seat built into the design? Why weren’t they designed to be social? Could they be? These are meant to be internet hubs specifically to “bridge the digital divide,” but because they are outside, they are harder to access in the winter or when it is raining. Why weren’t they designed with this challenge in mind? And they are so uninviting that many (higher income) New Yorkers don’t even know what they are even though they have useful resources for everyone! And why is the portal for people to use a tiny ipad that you have to lean over to read but the advertisements are displayed on two enormous 50 inch screens? People made each of these design decisions - if we start using algorithms in design, how will these priorities be encoded?

I decided to focus on bus stops instead because they are another example of public service structures built for people (the term social infrastructure is also sometimes used) and they are much more common. I was also inspired by the soviet era bus stops that my friend recently showed me. Why can’t/shouldn’t bus stops look like this?

The process is inspired by a podcast I heard about the Autodesk offices being designed using machine learning

Diagram of AutoDesk’s process

and the Chair Project by Phillip Schmitt at Parsons that used machine learning to generate new chair concepts.

Image illustrating the process that the chair project used

Week 1: Design Process

Our first assignment was to follow the design process we learned about in class to come up with concepts we want to work on in this workshop based on our different interests. Here is the framework for this exercise:

I made the below diagram to explore some of the concepts I am interested exploring in this workshop: