Abstract
We address the problem of verifying whether unwanted states, characterized as a given state formula, are reachable in a given parameterized multi-agent system (PMAS), i.e., whether the PMAS is unsafe. As the multi-agent system is parameterized, it only describes the finite set of possible agent templates, while the actual number of concrete agent instances for each template is unbounded and cannot be foreseen. However, as safety depends in general on the number of agent instances, the verification result must be correct irrespective of such a number. After having defined two distinct execution semantics of PMASs, in this paper we focus on an implemented approach for checking safety, which is composed of two steps. First, we have implemented a modeling tool, called SAFE, that allows to specify the agent templates in the PMAS and their possible interactions, and to automatically translate this model into a textual encoding of an array-based system (ABS). Second, we check safety via infinite-state model checking based on satisfiability modulo theories (SMT), by using the general purpose SMT-based model checker MCMT, which accepts ABS specifications as input. We show the correctness guarantees of this approach by relying on the theory of ABSs. Finally we discuss how this approach lends itself to richer parameterized and data-aware MAS settings beyond the state-of-the-art solutions in the literature, using SMT-based results now available thanks to this work.