Current Practice of Reliability Engineering
- Not Good Enough -
Yes, I'm certified as a Reliability Engineer. I've taught the basics to 1000+ engineers. I have applied all the tools myself for over 25 years... and I can make nice powerpoint slides and Weibull plots with the best of them!
BUT THAT IS BORING AND INADEQUATE!
When you have a ton of experience, and you start thinking about the real goal, you realize that it requires a combination of rigorous approaches and unusual, practical approaches to get the best information required to do something great.
My Guiding Principles:
- If there is a low-reliability component or sub-system, get rid of it.
- Incremental Improvement is not good enough and it usually doesn't last.
- Big improvement does not equal big risk.
- Adding complexity & cost to address a problem is the lazy, weak approach.
- Compromises are weak and guaranteed not to meet all requirements.
- The best Preventive Maintenance is NO preventive maintenance. Many say this is too hard... I say "too bad, we are going to do it".
- The best Predictive, or Condition-Based Maintenance is better and more interesting, but it is still just a way to avoid eliminating the problem.
- Subjective, random brainstorming is weak and ineffective. This is common when teams do root cause analysis and FMEA's.
- Not letting data drive understanding and decision-making is weak and lazy.
Rigorous, Systematic Problem Elimination
- When You Are Serious -
As we often say: "Serious Illness Required Serioius Medicine".
Common Six-Sigma & Lean tools are oriented towards continuous improvement, whereas these approaches are focused on creating "elegant" solutions, i.e. solutions that are very simple, very inexpensive and do not compromise on any requirements.
Known as "TRIZ", it is a relatively unknown methodology, but is growing rapidly. I have been studying, practicing and teaching TRIZ for 25+ years. It takes great persistence and determination to learn, and deep-thinking to apply, but it always worth the effort. The approach is based on the learnings from millions of patents and I have personally solved problems with savings on the order $50 million.
The other significant part of this approach is a rigorous method for understanding causes, known as "Physics-Based Causal Analysis"... developed jointly with other engineers at Honeywell Aerospace. It is far more rigorous/complete, without depending on brainstorming. The other major benefit to using this method is that it provides many starting points for utilizing TRIZ to move from merely analyzing, to total generating elegant solutions that eliminate the problems.