Safety analysis is used to ensure that critical systems operate within some level of safety when failures are present. As critical systems become more dependent on software components, it becomes more challenging for safety analysts to comprehensively enumerate all possible failure causation paths. Any automated analyses should be sound to sufficiently prove that the system operates within the designated level of safety. This paper presents a compositional approach to the generation of fault forests (sets of fault trees) and minimal cut sets. We use a behavioral fault model to explore how errors may lead to a failure condition. The analysis is performed per layer of the architecture and the results are automatically composed. A complete formalization is given. We implement this by leveraging minimal inductive validity cores produced by an infinite state model checker. This research provides a sound alternative to a monolithic framework. This enables safety analysts to get a comprehensive enumeration of all applicable fault combinations using a compositional approach while generating artifacts required for certification.
|Original language||English (US)|
|Title of host publication||Computer Safety, Reliability, and Security - 40th International Conference, SAFECOMP 2021, Proceedings|
|Editors||Ibrahim Habli, Mark Sujan, Friedemann Bitsch|
|Publisher||Springer Science and Business Media Deutschland GmbH|
|Number of pages||18|
|State||Published - Aug 25 2021|
|Event||40th International Conference on Computer Safety, Reliability and Security, SAFECOMP 2021 - Virtual, Online|
Duration: Sep 8 2021 → Sep 10 2021
|Name||Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)|
|Conference||40th International Conference on Computer Safety, Reliability and Security, SAFECOMP 2021|
|Period||9/8/21 → 9/10/21|
Bibliographical noteFunding Information:
Acknowledgments. This research was funded by NASA contract NNL16AB07T and the University of Minnesota College of Science and Engineering Graduate Fellowship.
© 2021, Springer Nature Switzerland AG.