Verifying Emergence of Bounded Time Properties in Probabilistic Swarm Systems

We introduce a parameterised semantics for reasoning about swarms as unbounded collections of agents in a probabilistic setting. We develop a method for the formal identification of emergent properties, expressed in a fragment of the probabilistic logic PCTL. We introduce algorithms for solving the related decision problems and show their correctness. We present an implementation and evaluate its performance on an ant coverage algorithm.