Abstract – Over the past decade, the Department of Defense has placed a great amount of attention on the advancements of unmanned aerial vehicles (UAVs), and more specifically on employing a large number of autonomous UAVs into “swarms.” These swarms form an organized cluster of vehicles to act out multifaceted operations as a group. Despite the benefits offered by UAV swarms, there are hurdles that engineering teams must grapple with while designing a UAV swarm system. One key area is creating and understanding the swarming behavior and revealing all potential failure scenarios that may impact the desired mission. This research uses Monterey Phoenix (MP) to model system behaviors by grouping them into distinct, reusable agent-like models of possible actor behaviors and modeling actor interactions as separate constraints. This approach affords the ability to compute every possible variation of actor behaviors with every other possible actor behavior from these models, which generates an exhaustive set of possible scenarios or event traces. Through manual inspection or semi-automated assertion checking of these event traces, the discovery of unwanted and undesirable behaviors and failure modes is achievable, which allows mission planners to then counteract these unsolicited instances with necessary failsafe behaviors.