Fail-awareness is the ability of a system to detect an upcoming failure before it actually happens. In this paper, we propose a weakness monitoring approach for observing a complex system during its operation, identifying possible degradation of its behavior, and finally raising an alarm in case of an estimated upcoming failure before the system actually goes out of its specification. Our procedure uses online linear regression to monitor trends over time – it is used to optimize the system service. We evaluate our approach on three case studies from the automotive and avionics domains.
CITATION STYLE
Granig, W., Jakšić, S., Lewitschnig, H., Mateis, C., & Ničković, D. (2020). Weakness Monitors for Fail-Aware Systems. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 12288 LNCS, pp. 283–299). Springer. https://doi.org/10.1007/978-3-030-57628-8_17
Mendeley helps you to discover research relevant for your work.