History of Reliability Part 2
Chris and Carl continue the discussion on the history of reliability engineering. The discussion that started with Chris and Adam (you can click here to listen to it). In the short time they had to talk about it, they came up with 6 reliability engineering epochs. Now Carl is going to see if there is anything to add – and importantly – anything to learn.
Join Chris and Carl continue a discussion on the history of reliability. Or more correctly – reliability engineering. Chris and Adam came up with 6 reliability engineering epochs: over-engineering; rapid technological change (World War I); formalizing reliability engineering as a discipline (World War II); the Japanese Economic Miracle; western consumers demanding more; and the computing power of the 1980s and 1990s. What can Carl add to this discussion?
It turns out a fair bit. So we have replaced our last epoch (which was admittedly fairly broad) and replaced it with the following really useful additions to our history.
- New Epoch #6 – 1980s and Test and Fix or Build-Test-Fix. This happened a lot in the 1980s with more complex machinery and systems being manufactured. Even though computing power was increasing, there was a slow uptake of its potential. So instead many designs focused on building prototypes, trying to see if they worked, and then redesign the next prototype. Without trying to identify potential failures from a scientific perspective at the blueprint stage. Which leads to long and expensive production lifecycles.
- Epoch #7 – 1990s the ‘Vital Few’ and Design for Reliability – Understanding the Physics of Failure (PoF). Market pressures always drive change. Developing products faster and for shorter service lives meant we needed to have a better understanding of the vital few failure modes that really matter. There are typically only a couple of failure modes that drive failure – which is great! Using a scientific approach to work out which are the weak points of your design gave manufacturers a key advantage over competitors who were slower to evolve. So this is where things like Failure Mode and Effect Analysis (FMEAs) started to become really powerful.
- Epoch #8 – Modelling and Using the PoF. Things like fault trees and reliability block diagrams have been around for a while. But they have not really been used as much as they could have. A 1975 study on nuclear reactors was perhaps the first attempt to do this – but it wasn’t commonplace. So using reliability modelling started to become more common because it drove design effort in a more targeted way. But there was a catch – a thing called reliability prediction started to happen. Tables and handbooks of historical ‘estimates’ for reliabilities of different types (like MIL-STD 217) of components started to be used to get reliability models. But these offered little value … how could they? Using tables of generic reliability estimates means that a reliable Toyota has the same predicted reliability as a Yugo (… Google that if you haven’t heard of it).
What are your thoughts?
Enjoy an episode of Speaking of Reliability. Where you can join friends as they discuss reliability topics. Join us as we discuss topics ranging from design for reliability techniques to field data analysis approaches.