Speaking Of Reliability: Friends Discussing Reliability Engineering Topics| Warranty| Plant Maintenance   /     Doing Calculations

Description

Ever wondered if the number a calculation (or software package giving you that number) is right? You might be surprised to find out that the same data that is fed into different equations or software packages supposed to evaluate the same thing ... give different results! How can this be? The post SOR 1008 Doing Calculations appeared first on Accendo Reliability.

Summary

Ever wondered if the number a calculation (or software package giving you that number) is right? You might be surprised to find out that the same data that is fed into different equations or software packages supposed to evaluate the same thing ... give different results! How can this be?

Subtitle
Duration
0:00
Publishing date
2024-10-11 10:50
Link
https://accendoreliability.com/podcast/sor/sor-1008-doing-calculations/
Contributors
  Christopher Jackson
author  
Enclosures
http://episodes.reliability.fm/sor/sor-1008-doing-calculations.mp3
audio/mpeg

Shownotes

Doing Calculations

Abstract

Chris and Fred discuss how to calculate ‘confidence’ and other things regarding probability and statistics … particularly as it applies to reliability!

Key Points

Join Chris and Fred as they discuss how we go about calculating things like the ‘confidence’ in the outputs … that other equations give us! Wow. What are we talking about here?

Topics include:

  • Equations give us numbers. Like the reliability estimate of a product at two years based on a combination of test and field data. And there are some issues here. With like to think that these equations we find in textbooks are objective. But … there are equations based on different approaches. So they aren’t objective (such as regression on ‘x’ versus Maximum Likelihood Estimate – MLE). Both those horrible terms in the brackets are different approaches to fitting the ‘line of best fit.’ So you need to choose the approach you like … making everything subjective.
  • Confidence is (an emotion) that is a measure of you. But we often talk about it as if it is an objective number. So you might use one of the two approaches in the point above … but how ‘confident’ are you that the output from one approach is correct versus the other? If the data we are trying to fit that line to is from a random process, there will always be some uncertainty (lack of confidence). Some software packages give ‘confidence intervals’ … like the 90 % confidence region that the true line lies. But how does this work?
  • So what do you need to do? Understand. If you are just wanting a number so you can move on … you are not the listener we are after. Otherwise, take the time to understand where the calculation comes from, what competing approaches there are, and what might go wrong. But most of all, you need to understand the decision you are trying to make.
  • And if you understand the decision … then you might not need a lot of ‘confidence’ in some of your outputs to make the right decision. Or you might not need to use calculate anything at all!

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.


Download Audio RSS

Show Notes

The post SOR 1008 Doing Calculations appeared first on Accendo Reliability.