Knowledge Sharing in Swarms Realized by Artificially Empathetic Communication

In this paper, we present a novel method of communicating knowledge between units in a swarm, applying artificial empathy and imprecise cognition to improve cooperation and adaptation. This research focuses on building a swarming agent's individual knowledge base, relying on knowledge imprecisely shared by another unit and one's own experience. The fuzzy similarity is used to fill the missing parameters, a subjective subset of which describes each agent. The obtained knowledge is then implemented as a particular set of actions, distinctive for each swarming agent. The actions bring the swarming unit closer to the desired state, appointed with the use of uncertainty combined with imprecision. Each agent compares their future possibilities with other's prospects and decides whether to help, continue with their own action, or learn new knowledge from others.

Joanna Siwek, Patryk Żywica, Konrad Pierzyński, Przemysław Siwek, Adrian Wójcik