Design After Nature Project

You Pretty Little Flocker: 4 x backlit transparencies, computer generated images based on a modified boid flocking algorithm, Alice Eldridge, 2008

The Design After Nature project (DAN) ran from 2007 – 2010 and was funded by an ARC Discovery Project Grant. Our research looked at applying natural principles and processes to creative systems. In addition to research papers, our project created a number of new artworks and performances, several of which were exhibited in the Design After Nature Exhibition and an earlier exhibition, Biotope. You can find detailed information on the project’s research here.

Overview and Summary of Results

Scatter plots of gene values in the population at generation 2000 for spatial models for concave (left), flat (centre) and convex (right) fitness curves. Results are shown for (a) homogenous and (b) heterogenous environments.

Our early experiments were concerned with understanding the dynamic behaviour of specific mechanisms in biological systems. For example, we experimented with the idea of trade-offs between generalist and specialist strategies in simple agent-based simulations [paper pdf]. The idea being that a simple control over the population could force the evolutionary behaviour to shift when becoming stuck in local optima.

The project devised a number of new methods for generative art and design, for example, Filterscape, and audio synthesis ecosystem, was able to generate significant diversity using the metaphor of energy recycling. Energy producers and consumers adapt to different sonic environments by processing and shifting sound energy across different frequency bands. Both cooperative and competitive strategies emerge [paper pdf].

Filterscape visualisation. Circle diameter denotes agent size, the arc represents the transformation interval. The horizontal line through the body of each agents signifies its vision range. This image depicts the population shortly after a population explosion. Inset shows how recycled energy is reflected at boundaries.

We also developed new methods to generate heterogeneity in audiovisual art systems using methods based on multi-level habitat formation by ecosystem-engineer organisms in a virtual environment. This resulted in the spontaneous emergence of heterogeneous, dynamic structures – perviously a difficult result to obtain in evolutionary generative systems.

Feedback plays an important role in both cybernetic and biological systems. Works such as Colourfield use a self-observing feedback process to control the flow of energy back into individual colour agents [paper pdf, presentation slides]. Each agent is contributing to the total energy, so individual colour agents must find strategies to keep the energy flow in balance.

Creativity Search

Illustration of the concept of a creative region. Two hypothetical distributions are shown, S1 (in light blue) and S2 (in medium blue) with their intersection highlighted (dark blue). A single "hump" exists in which r, shown within the orange box, is a creative region.

Towards the end of the project, we also looked at the problem of search in interactive creative evolutionary environments. We developed and tested a new method of “creativity search” that automates the search for novel phenotypes in an IGA system. Initial results have been very promising. The method can be applied to any kind of digital artifact for which measurable phenotypic features can be defined. More details here.

 


CEMA Workspace 2009: From left: Aidan, Alice, Nick, Ollie, Ben

Researchers:

  • Jon McCormack
  • Alan Dorin
  • Troy Innocent
  • Alice Eldridge
  • Taras Kowaliw
  • Aidan Lane

PhD Students:

  • Ben Porter (Completed 2011)
  • Mark Guglielmetti (Completed 2012)

Visiting Artists and Researchers:

  • Palle Dahlstedt
  • Mitchell Whitelaw
  • Rob Saunders
  • Petra Gemeinboeck