Meet The Team

Maksim Drapey

  • Instagram

University of Michigan

Master of Architecture

Yubei Song

  • Instagram
  • Grey LinkedIn Icon

University of Michigan

Master of Architecture

Shanghai University

Bachelor of Engineering

Yiying Tang

  • Grey LinkedIn Icon

University of Michigan

Master of Architecture

A message from us:

Artificial Intelligence has three qualities: Intentionality, Intelligence, and Adaptability. This thesis project explores the adaptability of acoustic design in specific architecture. We are proposing a human/AI collaborative design that will focus on creating spaces that alter and accentuate acoustic features in architecture.

Our project is still in the early stages in terms of devising a complete acoustic solution for concert halls through an AI-driven platform. The training of our neural network can be improved with the help of a more rigorous data set which would include more detailed 3D models, material properties, and associated values such as absorption coefficient for each material. As it stands now, the platform exists somewhere on the line between early design and optimization, requiring additional human intervention to aid in the complete design of a concert hall.

We find our AI platform to be particularly useful in the early stages of design, exposing us to a series of unexpected acoustic/formal solutions that we may have never considered. The possible sounds and qualities of these spaces captivate and inspire us and have the potential to connect to a wider culture of music, musical performance, and audience experience.

We were successful in correlating some critical acoustic parameters such as reverberation time, volume, and dimensional proportions together with formal qualities of concert hall interiors. The outcomes we achieved are unique interpretations of acoustic spaces through machine vision. The final proposals which we have chosen to develop and represent opens up a dialogue about creativity, authorship, and collaboration between architects and AI. 

 

As architects, we are often working with multiple collaborators and share authorship with respect to the projects that we design. In the case of this project, we consider our AI-driven platform to be a creative contributor and an important collaborator. Collaboration with an AI platform is unique however in how we interact. We act as incriminators by selecting the data which the Neural Network learns from, and as discriminators by choosing which of Network’s outputs fits our criteria as architectural designers.

Special Thanks

Matias del Campo

University of Michigan

Associate Professor of Architecture

Alexandra Carlson

University of Michigan

Computer Vision and Robotics PhD student

Sandra Manninger

University of Michigan

Assistant Professor of Practice in Architecture

Danish Syed

University of Michigan

Computer Vision

 © 2020 by Taubman College, Thesis Team

  • Grey Facebook Icon
  • Grey Instagram Icon
  • Grey Vimeo Icon