ALife

The AHA Model implements each agent as a whole organism that has a unique genome, simple hormonal system, rudimentary physiology, neurobiology including perception, appraisal, motivations, emotions, complex machinery for adaptive decision making, and a complex behavioural repertoire.

agent class hierarchy

The agents "live" within a stochastic virtual 3D environment that contains food and predators. Here they move, execute various behavioural actions, e.g. eat, grow, reproduce. The "life cycle" of each agent includes many time steps at which it makes the decision which behaviour to execute given the specific unique combination of internal and external stimuli as well as the internal emotional state.

The "best fitting" agents that do best during their life cycle then pass their genes to the next generation through the Genetic Algorithm. Thus, the AHA Model implements evolving intelligent and adaptive agents.

generations in GA

It can be said that these agents are capable of goal-directed behaviour. Their "goal" is therefore produced by the evolutionary process: it is to grow, reproduce and pass genes to the next generations.