This paper illustrates a practicable approach to reliability evaluation for highly reliable software systems based on the analysis of their operational experience and demonstrates its applicability to the control software of a gearbox system. The investigations were carried out within a cooperation of academia and automotive industry. The article also elaborates on the possibility of assessing software reliability at system level by combination of component-specific software reliability estimates.

The quantitative evaluation of high software reliability measures (e.g. for safety-critical automatic control systems) still represents a serious challenge for both developers and assessors. The use of statistical sampling techniques is often doomed to failure due to the enormous testing effort it usually requires. In order to support the practicality of this well-founded approach by increasing its cost efficiency, the present article examines the potential of exploiting operational experience for the purpose of replacing prohibitively extensive testing phases and reports on a successful experience gained in the context of a real-world automotive application.

Fur den Einsatz komplexer Softwaresysteme in sicherheitskritischen Anwendungsbereichen ist angesichts der schwerwiegenden Folgen von Versagen ein rigoroser Nachweis hoher Zuverlassigkeit angebracht und oft auch vorgeschrieben. Die wahrend einer vorangegangenen Testoder Betriebsphase beobachtete fehlerfreie Funktionsweise deutet zwar auf ein zuverlassiges Produkt hin, jedoch fehlen bisher Methoden um einen quantitativen Nachweis auf Basis der gewonnen Betriebserfahrung zu erbringen.