Evolving behavioural animation systems

Abstract

One of the features of Artificial Life (AL), is its ability to cross boundaries between traditionally separate disciplines. While its fo- undations are in biological research and computing, its visual nature has implications for the fields of art and entertainment. Using a continuous genetic algorithm, adaptive autonomous agents can explore user crea- ted environments. If these agents have pressures of 'natural' selection imposed on them, they can exploit the environment and create simple solutions to survive. When the environment becomes complex enough, the emergent solutions can, in turn gain in complexity, leading to un-expected and visually pleasing results. We produce animation sequences whose content/aesthetic is defined by the foraging and mating behaviour of simulated agent colonies.