Abstract:
Aerospace systems have the most stringent requirements when it comes to safety and reliability. The harsh environment combined with the immense costs of designing and man...Show MoreMetadata
Abstract:
Aerospace systems have the most stringent requirements when it comes to safety and reliability. The harsh environment combined with the immense costs of designing and manufacturing space applications places huge concern over ensuring that applications are dependable and function as desired. Formal verification is a tool to guarantee correctness for software executing on aerospace platforms. The advantage of formal verification is the complete absence of error against a design specification. Many applications leverage formal verification to ensure correctness and even integrate it into the certification process (e.g. DO-178C).Formal verification is a highly complex and computationally intense process that requires validating complex structures and an iterative approach to achieve a final, formally verified design. This process is further complicated in aerospace applications as traditional methods often exhibit behavioural differences between a formally verified model and the final software when deployed onto embedded architectures. Thus, formal verification needs to ensure that the verified design is accurate and representative of the executable on the target platform. Another major limitation of formal verification is that current approaches place formal verification towards the end of a project lifecycle. This creates a waterfall design and validation process that often leads to significant rewrites that prevent iterative verification.This research introduces continuous verification, a methodology consistent with current principles and practices in Agile development. This approach allows the early detection and correction of errors during the early stages of a project lifecycle. Designers can now decompose the system as a whole and implement models iteratively. Continuous verification allows the partial verification of models while the programmer iteratively implements them, allowing the early correction of errors and identification of non-verifiable entities that...
Published in: 2024 IEEE Aerospace Conference
Date of Conference: 02-09 March 2024
Date Added to IEEE Xplore: 13 May 2024
ISBN Information:
Print on Demand(PoD) ISSN: 1095-323X