Session:

LBP - Late Breaking Papers

Title:

Investigating the evolvability of biologically inspired CA

Authors:

David Basanta
Mark Miodownik
Peter Bentley
Elizabeth Holm

Abstract:

The developmental processes studied by biologists are emergent self organised processes that are the result of natural evolution. Developmental biology can be a good inspiration for anyone interested in evolving self organised discrete systems like Cellular Automata: nature has proved that evolving self organisation is possible. In this paper, we will describe EmbryoCA, a model of 3D Cellular Automata for pattern generation. Developmental biology has been used as an inspiration to design EmbryoCA and make the model more gradual in the hope of having a more evolvable type of CA. Also, experiments comparing the evolvability of different setups of EmbryoCA with a conventional CA model are shown.