Beta Guidelines

[Wade and Tom]

[Originally Posted: 12/4/00--Transferred by ReliaSoft Moderator]

Are there any beta guidelines out there for a mix of electrical and mechanical systems? We are uncomfortable assuming a beta of 1 when we no data. Any feedback would be appreciated.

[RS Support]

Assuming a beta of 1 is equivalent to assuming a constant failure rate, i.e. an exponential distribution. By making this assumption (and assuming that you have an increasing failure rate, i.e., beta>1) you tend to underestimate your reliability early on and overestimate it later on (or vise versa if you have a decreasing failure rate). On your question as to what to assume, well I am afraid I do not have an answer for you. You have to base it on your knowledge of your system and keeping in mind that with beta>1 your failure rate increases, and the larger the beta the fastest the increase.