Confidence is used in reliability to give a probabilistic value to the likelihood that a taken measurement will represent a full population.  It is determined by measuring a sample size and then using a selected statistical distribution table to translate to a likelihood. Emotional confidence is how one feels about making a decision based on known information at a specific point in time.   Similar but different. But it is important to connect them.  This is why. [Read more…]

I like the occasional good catchphrase.  I typically tend to speak in analogies and anecdotes and many of my customers and colleagues have suggested documenting them.  So on occasion, I will write them in my notebook.  Here are a few. [Read more…]

One of the most difficult challenges of incorporating reliability activities with program planning is estimating the time it will take to demonstrate reliability.  Demonstrated product reliability is typically measured as a confidence in a product’s capability to achieve a reliability goal. At product release, the business unit may feel comfortable with their target market and warranty goals with a demonstration of a 90% confidence in the products reliability goal of 99.1%. [Read more…]

Here is a common situation that is considered when reliability testing is being planned. Let’s use letters to designate depth in a product component structure. The top level product is level “A” and a component in top sub assembly is “B”. A sub assembly, of a sub assembly, of a sub assembly, is at level “D”.  So is it more efficient to test the component at level “E” or at level “B”?

Question: Is it more efficient to test the component at level “E” or at level “B” to demonstrate it’s individual goal? [Read more…]

Weibull analysis is a tool that is used for characterizing the behavior of a data set.  The value of characterizing a data set is being able to make predictions of failure rate for the same or a similar population. Input to the Weibull analysis is time to fail for all failed units and running time for operating units. The assumptions for the data set are critical to ensuring the right information is included in the analysis.  This may include root cause analysis of the failures so they can be characterized as wear-out or use stress related. It would be beneficial to separate these two data sets because their characterization and behavior will be quite different Weibull factors. [Read more…]

If you had to section the history of reliability into only two periods I believe the dividing point would be around 1950.  The first section, stone age to 1950, the “Robust” era; 1950 to today, the “Margin” era.   In the robust, era reliability concerns were addressed by creating a design that was orders of magnitude stronger than required, “over-engineered”.  This protected against the variability of fabrication, use, and environment.  If it was made sloppily with low tolerances, it still worked;  Used for an unintended purpose, still worked; Left out in the snow, dropped in the river, run over by a horse, buried in the sand; still worked. [Read more…]

What does the reliability engineer do?

Do they design reliable products?    Or       Do they guide the development of reliable products?

It’s some mix of course. It depends on the process for the organization they operate in. I will point out that neither extreme works. I am often surprised with how many organizations I see that still have the 1950’s model of ” Design it and then give it to the reliability team to make it reliable.” I still don’t believe there is a single person in any engineering role in any engineering organization that believes that works.

That is what one of my old mentors used to call a “window dressing program”. No one believes it actually helps the product but you get to take your customers around your facility and show them all the great testing you are doing (SILENT END OF SENTENCE ” on out of rev parts and in a program phase where results can’t be implemented.”). [Read more…]

Get to Know Your Realibility Stakeholders

Understanding the reliability stakeholders in an organization is an interesting exercise.

Who are they?

• Executive management
• Business unit
• Sales
• R&D
• Quality
• Manufacturing

It’s easy to look at each of those and say “They all could easily get by and appear successful even if a product has moderate to poor reliability”. You could also state that ” Reliability is a key factor to success in their role” [Read more…]

This is the bathtub curve we are often shown.

This is what a real bathtub curve looks like. [Read more…]

With all the complaints you hear about products rebooting and software crashing, do companies really practice Software Reliability? In fact, there are some companies that do, but they are mostly in the industries that require products to have high availability, such as telecom, defense, and space, or safety-averse industries, such as medical and industrial plant operation. Most other industries don’t pay as much attention to it. The best method to increase Software Reliability without significant increases to schedules or budgets is to use a Software Design for Reliability (SDFR) approach.  These are the key steps. [Read more…]

The original windmills of Europe are feats of engineering.  I was able to get inside of a few in Holland and see all the mechanisms that have been operating for hundreds of years.  I was lucky that the wind was strong and the windmills had their sails out the days I was visiting.  A few were grinding flax seed and others designed to pump water from a low basin to a high basin to control basin water levels. Being inside of the windmill was like being an insect inside of a grandfather clock.  It was so interesting seeing how they engineered the mechanisms to optimize the balance of the system needs in such a different manner than we do with many of our currently engineered products. [Read more…]