Your browser doesn't support javascript. This means that the content or functionality of our website will be limited or unavailable. If you need more information about Vinnova, please contact us.

CreAItor - creative AI for innovative and sustainable architecture

Reference number
Coordinator BELATCHEW ARKITEKTER AB
Funding from Vinnova SEK 500 000
Project duration December 2020 - August 2021
Status Completed
Venture AI - Competence, ability and application
Call Start your AI-journey for businesses - autumn 2020

Important results from the project

The aim of this project has been to develop an AI solution for automation of the initial phase of the design process, when several ideas are developed, to be compared and evaluated against a set of criteria. This initial phase currently consumes about 50% of the time in a typical development project when performed by human architects. Automated solutions (possibly semi-automated) could lower the invested time to a fraction, freeing up architects´ time for the following processing phase.

Expected long term effects

In this project we have developed a GAN-based method that is trained with a set of apartment blueprints, and learns to generate novel blueprints. We have collected a dataset of ~3400 apartments and developed preprocessing to remove text. The result is a proof of concept of what is possible to achieve. Our preliminary finding is that it is indeed possible to design an automatic method that can create novel apartment designs, but that a purely data-driven approach is not sufficient.

Approach and implementation

This initial project has pointed to the need for several different development efforts. We see the need to investigate three main future directions, which are highly interconnected: - Take external constraints into account - Format and parameterization of output - Learning from few examples These findings will be the basis of a larger application for EU Horizon Europe funding submitted during the fall of 2021.

The project description has been provided by the project members themselves and the text has not been looked at by our editors.

Last updated 15 October 2021

Reference number 2020-04084