Simulating the study design in industrial studies.

Abstract

This study focuses on measuring the uncertainty around the estimate of the influence of temperature on the survival times for technical devices with small sample sizes. Industrial or technical devices are expensive, as a result their failure proves to be destructive and very expensive. Insufficient data arises since engineers cannot wait for large number of devices to fail before measuring the impact of temperature. As a result the estimate of the influence of temperature is sometimes not accurate and has large standard error. The aim of this project is to verify under which conditions the best estimate of the influence of temperature on these technical devices can be obtained. To provide a description of the uncertainty surrounding this estimate under different observational schemes or design. Finally, to analyze the accuracy of this estimate on the survival time in order to provide engineers with the most cost efficient study design.