Software and Thinking

Abstract

Chris and Fred discuss the many and varied different software package that can help you do ‘reliability stuff’ … and how we usually assume everything they do is ‘OK.’ But how do we know the software is giving us the numbers we need?

Key Points

Join Chris and Fred as they discuss how software packages help us do reliability stuff … and how we trust them to do the right thing. Is this trust well placed?

Topics include:

Let’s talk about Weibull Plotting. Weibull plotting is perhaps the most widespread reliability activity that we use software to help us do. But if you put the same data into a number of different commercially available software packages, we will get different outcomes. Why? Because these software plotting tools make 4-5 assumptions to quickly get things like confidence bounds. If you don’t know about these assumptions, then you don’t really know what the software is doing.
Come up with what you ‘expect’ to see in your data analysis. Do you believe that your system has some infant mortality or wear-in? Do you then believe that the majority of systems that survive infant mortality will fail due to (for example) ball bearing wear out? Do you know what both these types of failure look like? Do you also know how you expect this to visualize on a Weibull plot? You really need to answer questions like this before you conduct data analysis. If you don’t have the skills to answer these questions … you are not qualified to conduct analysis on any software package (sorry). If you can answer these questions, and the data analysis meets your expectations … great. If not … then there is a gap in your knowledge that you need to fill in before you use the outputs to make a million-dollar decision.
Do you simply trust your software package(s)? You can’t. They are all useful. They all have their issues. They are only useful if you know what the issues are. This means you know when to trust the outputs (and when to question the rest).

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.