ALAR: Artificial Life and Adaptive Robotic Lab

School of Computer Science, UNSW@ADFA

Canberra, Australia

Coevolution of Morphology and Mind

The project focuses on evolving behaviours, controllers, and architectures for artificial life-forms in a physically-simulated environment. A physics-based simulation tool is used to bridge the gap between the artificial and real world, thus permitting such controller-morphological designs to be transferred from virtual reality into real-world applications without almost no modifications. The ultimate aims of this work are to develop and implement robust robots in virtual and physical worlds, and better understand the underlying principles governing the relationships between evolution of robot' controllers, architectures, and behaviour that will allow for robust and adaptive artificial life-forms. The research areas include collaborative behaviour, coevolution, distributed/decentralised controllers, evolving strategies for civilian and war games, hierarchical evolution/dynamics, measures of complexities in virtual environments, and self-adaptation.

The project aims at producing low-cost robots for movie productions, science fictions, toys for children and charcter animations.