Reactive Real-Time Cooperation as a Combination of Altruism and Self Satisfaction

This work proposes a method for integrating cooperative behaviors among distributed autonomous agents based on local communications. The presented architecture is reactive-based to ensure self-organization, reactivity and adaptability of the agents. Animal signal spreading have interesting characteristics such as spatial limited communication, intensity meaning, redundant information (robustness) and reactivity. We investigate this type of communication to improve reactive architecture in order to handle robust communication and reactive cooperation. The underlying idea of our architecture is to use a signal of interactive satisfaction to perform action selection and cooperation. By taking into account the satisfaction of its acquaintances, an agent can have an altruistic behavior thus allowing for intentional cooperation between agents.