Abstract
SUMMARY & CONCLUSIONSThis paper proposes a separable binary decision diagram (BDD)-based analytical (SBA) method to assess mission reliability of a heterogeneous multi-phase drone system considering cascading effects from overload and intra-phase collaboration. Drone systems have been widely applied in critical real-world scenarios, where multiple heterogeneous drones collaborate to collect and communicate information, and subsequently take actions to deliver intended service. The interaction and collaboration among drones introduce interdependencies, facilitating failure propagations and potentially incurring high-impact cascading failures (CFs). Based on the law of total probability, the proposed SBA method adopts a divide-and-conquer strategy to transform a complex problem into simpler subproblems that can be solved in parallel, enabling efficient modeling of CFs. A case study on a two-phase rescue mission system demonstrates how the proposed method captures the spread of failures and their impact. Data results from a comparative analysis confirm that CFs are more detrimental to mission reliability than individual failures. Moreover, as the quality of drones deteriorates, the overall mission becomes increasingly vulnerable to CFs.