Nearly all engineers have some knowledge of the well-established reliability curve for most components. There's the initial, relatively high “infant-mortality” rate, usually followed by a long period of low failure rates, and then an uptick as “wear out” sets in and the rate increases. There is lots of solid data to support this bathtub-shaped curve.
But over the years, I have seen a different sort of reliability bathtub curve, which I will attribute to design and manufacturing, rather than infant mortality and wear out. What really made this evident to me was the repeated experience with basic cordless phones at home: the keyboards of the most recent ones I bought only lasted a few months.
Here's what I think happens: when a new product category starts to develop, there are initial failures due to the inexperience of the vendors with the design and manufacturing, and overall newness. Not good, but technically understandable, especially since many failure modes can only be seen when large numbers of the product get out in the field; long-term testing of a small sample size is not the same thing.
Then, as the problems develop, are understood, and overcome, product reliability increases; it's a process may take a year or two. All is well and we're all happy, but only for a while.
Then trouble sets in, as the market for the product becomes ultracompetitive. Vendors slash their design quality, their assembly quality, their materials, everything. Soon, the product's quality is back in the dumps, despite the experience and history that the vendor and manufacturers now have with the design and assembly. So we come full circle, from poor reliability to higher reliability and back to lower reliability, but for a different set of reasons than the normal reliability bathtub curve.
Maybe it's just me, since I like to keep products—and keep them going—or quite a while, far beyond what vendors consider their normal replacement cycle. For example, I've taken cordless phones and calculators apart just to get under the keyboards and clean out the deteriorating conductive plastic, or whatever they use for the keys and their contacts (some are easy to do, some are hard to open). And as I do so, I can see the changes, for better and worse, of the design and materials. It's this cheapening that makes inexpensive products available to more people, for sure, but also diminishes the reputation of engineers as a professionals, which is jot a good long-term thing.♦