Assume that you are responsible for the reliability aspects of a system containi
ID: 3066135 • Letter: A
Question
Assume that you are responsible for the reliability aspects of a system containing both electronic and mechanical elements. The customer for the system requires that a numerical reliability prediction be provided:
a. Describe what is meant by a “reliability prediction” in this context.
b. Identify some sources of data that can be used to assist in quantifying the prediction, and discus the dangers that have to be guarded against in the use of such data
c. Do you expect the prediction to overestimate or underestimate the reliability that could be achieved by the system? Give your reasons.
Explanation / Answer
a)
The reliabilty of a system can be understand as the probabilty that a system will work in a satisfactory manner for specified time interval and under predefined operating standards. Hence the term "relaibility prediction" can be define as the failure rate of such system.
b)
Some traditional example of reliability prediction are field tracking studies and warranty return data.
Field tracking study is used to evaluate the performance of the product under the predefined standards to assist the reliability of a system to predict whether the new product will work in satisfactory manner. Accurately tracking of such data is a serious concern because the product might be failed due to violation of conformities for example inaccurately installation or some environmental effect .
Warranty return data is used by manufacturing companies to predict the reliability of a product design.It helps the companies to understand the causes of failure and to maintain the service or replacement. The data need to be observe carefully because one cannot replace a product for a minor cause of error it may affect the profitability. The failure might be occurs other than the failure of subparts or subsystem because of mishandling or violating the operating condition and so on.
c)
The prediction for reliability may be overestimate or underestimate for example if we are predicting the product failure rate using the simple linear regression analysis, the prediction can be overestimate or underestimate because of the nonlinear relationship between dependent and independent variables. In thi case the result will cause overestimate for one variable and underestimate for other.