We use digital evolution to directly test ecological and evolutionary hypotheses, and to generate control algorithms for complex robotic behaviors. Avida is uniquely valuable for such work in that it carries all of the benefits of simulations (e.g. rapid generation times and full manipulative control over environmental forces), without incorporating mechanisms to artificially select individuals for reproduction (which can readily bias outcomes). Because both ecological and evolutionary interactions in the system are open-ended and unsupervised, even though it is computational and not biological, Avida uniquely allows for unrestricted evolution via natural selection.
I have expanded Avida to introduce new mechanisms for ecological and biotic interactions, such as new sensory mechanisms, more dynamic resource environments, and predation. In the pages on this site, we highlight some of the related areas of our research.