No abstract

Over the last two decades, an interesting area of Brazilian military and civil sectors is the Unmanned Aircraft Vehicle (UAV) development. This article tackles the modeling of conflicts resolution of Unmanned Aircraft System (UAS) using a lightweight Duration Calculus (DC) to verify if the temporal specification and design of the system is correct and to ensure formally that the system implementation meets all its requirements. Moreover, the article proposes a formal modeling (using DC) of a conflicts resolutions set of rules, adapted from Free Flight concept in Communications, Navigation and Surveillance/Air Traffic Management (CNS/ATM). In the adapted approach to UAS, each UAV is surrounded by an imaginary space of two cylinders, which form, respectively, the protected zone and the alert zone. The major contribution of this article is structuring a new scenario application of the conflicts resolution to UAS through formal modeling, using the DC technique to confirm that the models could be implemented without deadlocks and unreachable states, as well as with satisfaction of temporal restrictions. Furthermore, this work uses the state-ofthe-art practices in formal methods, including a model checking tool to ensuring correct real-time requirements specification of a real-time critical system.

With the increasing importance of software in the aerospace field, as evidenced by its growing size and complexity, a rigorous and reliable software verification and validation process must be applied to ensure conformance with the strict requirements of this software. Although important, traditional validation activities such as testing and simulation can only provide a partial verification of behavior in critical real-time software systems, and thus, formal verification is an alternative to complement these activities. Two useful formal software verification approaches are deductive verification and abstract interpretation, which analyze programs statically to identify defects. This paper explores abstract interpretation and deductive verification by employing Frama-C's value analysis and Jessie plug-ins to verify embedded aerospace control software. The results indicate that both approaches can be employed in a software verification process to make software more reliable.