I was given the opportunity by my university (University of East London) to exhibit a site specific installation in a group show curated by Richard Wilson called Red Square Yellow Square. The art was exhibited in the courtyards of the new Knowledge Dock building. The piece I created for the show was derived from my Self Organising Map work. I wanted to develop the work further to create a net that responded to a given image, but not with the tightness of a neural net. What I did instead was create a grid of spots that grew and shrunk based on their distance from each other. Depending as to whether the spots sat on white or black they would effect gravity on the other spots. This system allows one to feed an image decribing what is behind a wall and allowing the spots to react to it. The vinyl cut spots for this installation were supplied to order by Omnicolour.

The layout of the spots was guided by a plan that was created by an application in Processing. I fed an approximation of where the walls and stairs might be based on architectural plans as a bitmap to it. This application used the AIExport library and an additional parsing class to load information about the spots to memory so they could be printed in a compact file for vinyl cutting. I will not be uploading the source for the AIExport commands, suffice to say they were integrated with some experimental Swing components and if I return to this project I will use the PDF library to export the data - the code would simply be easier to read.

The demo of the application below can be activated by clicking on it. The individual spots can be dragged about.

To view this content, you need to install Java from java.com

Source Code: spotApplet, netSpot, parser (listed for documentation purposes)

Built with Processing