SmartGeometry Conference and Workshop

Date: 2018

Team: Zeynep Aksöz, Clemens Preisinger, Erika Stadnik, Natalie Sham, Edward Bruun, Peter Olendzki, Louis Leblanc, Omri Menashe, Xiaolong Li, Michael Lee, Anthony Mattacchione, Andre Pereira, Feng Le

Programme: Machine Minds, AI Strategies for Spaceframe Design

Software: Rhino3D, Grasshopper, Owl (library for machine learning-oriented data processing)

We optimised the structure with a pre-trained Machine Learning model, which only took a few seconds to deliver the solution. This is in contrast with an evolutionary algorithm alternative, which would have taken days to deliver the most optimised solutions.

​

Find more information at https://www.openfields.eu/ML-Strategies-for-Space-Frames

​

Images courtesy of the team. 

 

Stages of prototyping: Global shape, Voxels, Bracing, Bracing optimisation with Machine Learning

 
Group of people working on a geometric design with protractors, angle rulers, and copper wire on a table.
People assembling a structure with pink connectors on a table indoors.
Hands working on a geometric structure made of sticks and pink tape, with measurement papers on a table.
Group of people collaborating on a geometric structure made of pink strings and rods in a modern office setting.
People working on a complex geometric structure made of sticks and pink connectors, with scattered printed plans on the floor.
A spacious indoor room with a large, complex pink geometric structure made of rods and connectors displayed on a white pedestal. Several people are observing the structure, with tables and wall-mounted speakers in the background.