Anaolog Devices produce electronic components which can fail for a variety of reasons. However, there is some experimental evidence suggesting that the components never fail before some threshold time, i.e., the components are guaranteed to have a lifetime which is greater than this threshold.

Knowledge of a threshold time is useful in practice: it provides insight into the physical properties of the component (especially its failure mechanisms). Therefore, methods for establishing the existence (or non-existence) of such a threshold are useful in practice.

Some questions of interest are as follows:

- What is the best estimation procedure for thresholds?
- How well can we distinguish between distributions with and without thresholds in practice?
- Can progress be made without assuming anything about the distribution a-priori?
- What is the most efficient experimental protocol (sample size, duration, number of observed failures, experimental setup etc.) to establish threshold values?

It is anticipated that the answers to these questions can be found through a mixture of numerical simulations and literature review. Some real experimental data will also be made available.

Aside from statistical considerations described above, a mathematical model / stochastic process for individual failure mechanisms will also be useful to help establish the existence of such a threshold time based on fundamental physical properties.