A model of cooperative agent based on imitation and Maslow's Pyramid of needs

Recent works have addressed the problem of imitation in the framework of the interactions between two agents, whether humans or robots. We develop a model aiming at improving the self-organization of population of robots by relying on imitation. Imitations between the robots are regulated by a very simple model of emotional expression. The model is tested in the context of a simple task for the robots: to explore their environment to localize sources needed for their survival. Following a biology-inspired approach, imitation has been introduced within a population of autonomous agents, as bidirectional social needs, in line with the Maslow's pyramid of needs (1943). In our model, imitation is integrated into a global architecture based on artificial neural networks. Running our simple and scalable model resulted in a significant increase of the population's survival rate and a decrease of the global amount of the average necessary movements of each agent.