[SystemSafety] Qualifying SW as "proven in use" [Measuring Software]

[Peter Bernard Ladkin]

On 7/2/13 11:50 AM, Ben Bradshaw wrote:
> Regarding the values quoted for 90% confidence and the number of hours of fault free testing, I believe that if you have 3 x 10^X hours of failure-free operation you can be 95% confident that the average failure rate is one failure every 10^X hours.

Yes, that's right, thanks!

PBL

Prof. Peter Bernard Ladkin, Faculty of Technology, University of Bielefeld, 33594 Bielefeld, Germany
Tel+msg +49 (0)521 880 7319  www.rvs.uni-bielefeld.de