Wonderlab UCL Bartlett

Abundance

Abundance is an example of a robotically extruded microstructure, in the form of lightweight lattice. Due to the high level of noise and extremely high resolution of this fabric, its dazzling deep moire and deeply unstable interference effects can only be experienced by being immersed in this high resolution architectural fabric.

Lead Designer

Alisa Andrasek

Curation

Bruno Juricic

code and design

Madalin Gheorghe

Structure

Arup Engineering

Fabrication

AI Build

Lead Designer

Alisa Andrasek

Curation

Bruno Juricic

code and design

Madalin Gheorghe

fabrication

AI Build

structure

Arup Engineering

Lead Designer

Alisa Andrasek

Design And Code

Internet of things

Hitachi Consulting

Structure

Buro Happold

Directed By:

Alisa Andrasek with Andy Lomas, Daghan Cam and Someen Hahm

students:

Mikaela Psarra, Eleni Chalkiadaki, Jingwen He

Lead Designer:

Alisa Andrasek

Design Code Fabrication:

Ningzhu Wang, Jong Hee Lee, Feng Zhou, Zhong Danli

Design was generated by applying Perlin noise through the voxel cloud. Design search comprised of the interplay of different resolutions of voxel/lattice, and the resolution of applied noise. Here the pattern recognition in such a cloudy noisy structure exceeds human cognition and the designer’s capacity to search. There is a clear need for machine learning assistance. In the meantime, immersive VR representation was explored, to allow the designer to step into this lattice.

In this project, unlike the previous one, information for the robotic printing cannot be contained locally through instructions for individual voxels. It requires a path-finding algorithm in order to find the shortest navigation through this reticulated data cloud. Additionally, sequencing of printing steps needs to be inscribed in this pattern for a printing choreography. This makes it a large data problem that also calls for the future development of intelligence for such a high resolution printing process. In the next stages we envision developing an adaptive feature, where the robot is trained to recognise not only the node it needs to connect to (something we have already trained through deep learning), but also, locally in each layer, to understand how to print, and therefore the best path to take.

It resonates Borges’ story “The garden of forking paths” (Borges and Others 1962), where at any fork in this path-finding process, there is a multitude of possible paths or parallel futures.

Back to projects