Chris and Fred discuss what ‘reliability growth’ means? Sounds simple … right? And it can be. But sometimes not.
Join Chris and Fred as they discuss what reliability growth is. And isn’t.
- What is ‘reliability growth?’ It is usually associated with a ‘build-test-fix’ or ‘test-analyze-and-fix (TAAF)’ approach to design where a prototype is built, tested, and all the issues you find are corrected. But reliability growth can technically refer to any activity that makes reliability better. In practice, reliability growth tends to refer to testing prototypes.
- Then there is ‘military-centric’ reliability growth. This involves ‘build-test-fix’ or TAAF when prototypes are tested in ‘operational’ conditions and models are imposed on the test paradigm to be able to track how reliability improves. This has the benefit of being able to quantify reliability … but takes a long time. In this context, there is no scope for increasing stresses to find issues more quickly. And perhaps the worst characteristic of this type of testing is that we need to have a ‘mature’ or ‘production ready’ prototype to test. Which means we only start ‘growing’ reliability at the end of a production process.
- But reliability growth models actually model your DESIGN TEAM. Not the device. Because the rate at which reliability grows is a measure of your team’s ability to design out design flaws. So if your system has a couple of failure mechanisms that have a high failure rate … great! But if your system has lots of failure mechanisms with low failure rates … not so great … So we need to essentially assume a lot of things (including how good your team is) to model reliability growth.
- … so the result is assumed from the start. Which is a problem.
- BUT … if you have data that shows how much reliability has already grown, then we can do something. Reliability growth models are quite powerful – you just need some inputs. So if your team has already started, then you can look at how much reliability has grown and how it will likely continue to grow. This helps you work out how much time or resources you need to keep designing those defects out of the system. This works particularly well with software systems. But not necessarily military vehicles.
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.
SOR 744 Reliability GrowthChristopher Jackson
John E Kreucher says
Fred and Chris, another fine podcast. I appreciate the perspective on RG and when and how you find it useful and when other approaches may make more sense. I find it similar in some sense to standard-based reliability prediction methods that rely a great deal on assumptions including unknown model parameters.
Fred Schenkelberg says
Thanks for the comment – and don’t get me started on the standard-based predictions…. may be a good podcast topic. cheers, Fred