Generative Mixed Reality app for the 2023 Pratt Design Show in Brooklyn NY.
KNIT + CODE was a design collaboration by Pratt School of Design faculty Hannah Berkin-Harper and Benjamin Howes, and Tracey Weisman of the the School of Design Knit Lab. It was fabricated in Brooklyn NY in the Spring of 2023 by a dedicated team of Industrial and Fashion Design students, and was the ceterpiece of Pratt Shows: Design 2023. This repo contains the source code for the generative AR part of the project. To learn more about the CNC-knit physical installation hit us up on twitter or IG or something!
The generative 3D geometry is created on the fly in the browser using threejs. Each fiberous iteration is very probably unique - we use Math.Random() all over the place. The generative geometry codebase was cobbled together using a combination of some traditional reseach and design workflows (google it, experiment, repeat), and some experiments with generative AI coding assistants (Chat-GPT, Github co-pilot).
We're using the google model viewer to handle interactive 3D interactions on the web page, and to natively handle Augmented Reality / Mixed Reality modes on AR-enabled mobile and tablet devices.
When the page loads, and when the generate new iteration button is pressed, a new set of fibers is generated, and a .glb file is passed to the model viewer as a Blob.
This is a single page react website that was bootstrapped using vite. To run the site locally, fork or clone this repo, install the npm packages, and use the following commands in a terminal to develop, build, and deploy:
local development startup: yarn run dev
local build: yarn run build
github pages deploy: npm run deploy