Have you heard of the "bathtub curve"? The graphic representation looks like a very lazy "U." On the x-axis, we plot the entire lifecycle of a component or assembly, and on the y-axis, we plot the number or percentage of failures.
Generally, this curve tells us that, for any given component or assembly, the earliest period -- the steepest part of the curve and known as infancy -- has the highest failure rate. Alternatively, the flat part of our lazy "U," which depicts the lowest failure rate over a protracted period of time, is known as useful life. And moving to the right on our plot where the curve goes back up again, this section represents failures as the part reaches its end of life due to intrinsic material issues and accumulative electrical or mechanical stresses.
The Bathtub Curve
We could use the bathtub curve reliability engineering lingo, or we could just simply recognize that what we are looking at is a cradle-to-grave lifecycle phenomenon. Infant mortality, useful life, and wear-out are experienced by everything in our universe.
One business term for tracking and managing lifecycles in the electronics and high-tech industry is product lifecycle management (PLM). From product conception through design, introduction, production, and eventual obsolescence, we need to understand how PLM tools and analyses allow us to not only understand and track these lifecycles, but in some cases manipulate them in our favor. How do we extend the useful life period such that the dependability, reliability, and longevity of a product or service increases?
Marketing and sales, for example, need to understand the shape of the curve from their approach because, if the flattest part of the sales curve representing healthy sales can be extended without having to go back to R&D to spend more money on a replacement design, then the savings become obvious. Just as our failure rate increased at the end of a component's life, the marketing and sales plot would show a steady increase in loss of sales over a shorter period of time.
Understanding this aspect alone would help a manufacturer know the optimum time to invest in new research and product introductions. New or newer products change manufacturing schemes, require new tools and fixtures, and need modified firmware or software. Additionally, marketing materials and a lot of additional administrative costs are incurred for updating engineering and operations support systems such as the MRP, CAD, after-market support, and document control operations.
The fact is, just as marketing and sales can keep an eye on product health in the field while reacting to real-time trends, some items can be given extended useful lifetimes by reducing the stresses the parts experience. Think of it in terms of a runner. If I ask someone to run at top speed as far as he can go, then wherever the speedster ends up, he got there faster than walking. But that is it. He has expended his available energy, and if I immediately asked the exhausted athlete to run just as fast again, it just could not happen. However, if I say to that same runner, "just jog as far as you can," he will end up much further down the road because he has greatly reduced the stress on his heart, lungs, and muscles.
Similarly, if I can back off on the electrical or physical stresses incurred through temperature, current, voltage, pressure, etc., my components, and thus the entire assembly, will last longer. It will go the distance for which it was designed. If I can't mitigate any of those stresses, I can anticipate them in my design considerations and pick components that have higher stress handling capacities. Not pushing the components to their limits will extend their life.
In the next series of Best-Practices, we will be exploring how to use PLM to its maximum effectiveness by identifying and picking the low-hanging fruit off our product trees. Taking points of stress out of any system or process flow will always create a better end product.