Design for Manufacturability (DFM) is useful only if the product design is also reliable. It is easy to focus too much on DFM and end up with unreliable products. That can cost the company more money than the savings gained from DFM. Thus, we look at reliability and manufacturability together. Reliability of a product tends to grow over time as the deficiencies of the design come to light and corrective action is taken. A goal of DFM is to identify the reliability issues early so that the reliability growth curve is accelerated.
One way to measure reliability is to look at failure rates. Most products have a “bathtub” shaped failure rate curve. The initial high failure rates reflect the so-called infant mortality period. Following the infant mortality period is the so-called useful life period where the failure rate is low and relatively constant. Finally, there is the wear-out period where the failure rate starts creeping up again. Conceptually, the inverse of the failure rate is the Mean Time Between Failures (MTBF). It is often more convenient to refer to and measure MTBF instead of the failure rate.
We perform both theoretical analyses and empirical determinations of MTBF. Using well established formulae, we estimate the system MTBF from the individual component MTBFs. MTBFs for purchased components are often available from their manufacturers or published literature. MTBFs of custom components are estimated from the components used to build them.