Karthik Patanjali

An experiment on how physical products can learn from how users use them, to adapt and evolve its form. This is explored by evaluating how one sits (using machine learning) and generating the design of a chair algorithmically to best fit the user.

http://itp.nyu.edu/thesis/thesis2016/knp275/thesis-summary/

Description

This project is an experiment on how physical products can learn from its users, to adapt and evolve its form to better serve them. This is done in the context of work chair design – how a user sits is evaluated (using sensor data) with data analysis/machine learning and a design of the chair is algorithmically generated based on this analysis. The generated form slowly evolves real-time as the system learns more from the user. When materials that change shape are widely available, this method could be used to adapt and evolve the form of the chair with time/use. Imagine a work chair of the future that is uniquely yours and gets better by learning from you!

Classes

Thesis