Life Data Analysis & Reliability App
Carles CG of Reliable Dynamics created the Life Data Analysis & Reliability (LDA) set of analysis tools and, in cooperation with Accendo Reliability, is making them available to you. Select “Introduction Material” below to expand app description.
The reliability analysis app contains the basic tools for an introduction to the life data analysis field. Among others, reliability analysis is useful to statistically infer the Time-To-Failure of mechanical components.
The aim is this app is to educate, inform, as well as, give analysis capabilities in order to inspire the user. The aim is to motivate the user to invest time and have a deep dive into the world of reliability analysis. For simplification and information purposes the topic ‘Weibull analysis’ and ‘Life data’ has been chosen. This topic was covered by Dr. Abernethy in his book “The New Weibull Handbook”.
Features:
- Play with parameters to generate artificial data and reliability plots
- Upload your data for analysis & template download.
- Calculates on the fly if your data has one or more failure modes and does the plot accordingly.
- Play with different parameters for the regression fit, the confidence method, and the confidence percentage.
App division
The app is divided into 3 functional tabs:
- Reliability plot generator
In this tab is possible to generate some random data following Weibull distributions and get a density and a reliability plot. Is possible to play with:
– Number of points to generate
– Shape and scale of the random numbers.
– Confidence intervals for the plot.
– Different density kernel and bandwidth adjustment. - Weibull Modeler
- Upload your data functionality
You can upload your own data (CSV file) in order to do a reliability plot, and see how your data look like. In order to do so, is necessary certain formatting is fulfilled. In order to do that you can download the CSV template, available in the tab. - Reliability calculator
By default, there is a fully functional example with two failure modes load in this tab for aesthetics purposes. If you upload your own data in the upload tab, you will see the reliability plot here. There are certain functionalists to play around with the parameters of the plot such as:
– Confidence interval level
– Method to fit the points: here is possible to choose between 3 methods.
– Method to calculate the confidence intervals.
- Upload your data functionality
- Zero Failure Testing
- Time testing complete For Zero test complete test to failure, how much time do I need to test a component to demostrate certain reliability level? Given the failure mode, the life to demonstrate the number of components available will output the amount of time that is needed to test those components wirth zero failure.
- Time testing partial In case that the Zero test can not be completed, what is the level of reliability achievied?
Coming soon
- Forecast Modeler
- Model generator
- Train models
- Compare models
Inspired by Openreliability
“Openreliability is dedicated to the development of reliability engineering and operations research applications software under the open source paradigm. As open source, we invite a vigorous peer review of the technical accuracy of the software code in use. An open dialog on the use and application of this software is also encouraged. We appreciate your involvement on any level.”
“Our first and most active project is Abernethy Reliability Methods based on “The New Weibull Handbook, Fifth Edition”, by Robert B. Abernethy. This reference has been the textbook for a Weibull Analysis course that has successfully trained quality and reliability engineers for over three decades.“ Openreliability.org
How to Participate in the LDA App Beta Program
If you’re not already participating in the LDA beta program, join now for free to gain immediate access to the tools.
As a beta tester, you’ll also become an official Accendo Reliability site member. This not only gives you access to the LDA app toolsets, but you also gain access to exclusive members-only content throughout the Accendo website.
Sign up today for this limited time beta offer and tell us what you think. We’ll take your feedback to make usability improvements and to develop tools to fit your needs.
I’m currently a site guest – Sign me up
Existing Accendo Reliability site members
Accendo Reliability site members will be prompted to sign into their membership account to confirm beta program participation and to access the LDA app toolsets.
Already participating in the LDA beta program and don’t see the tools?
Sign in to your Accendo site membership account to gain instant access.
You currently do not have access to the LDA toolsets. Log in to your Accendo Reliability member account or register for the beta program above to view.
I used the Zero failure test plan feature and it works great. I am actually plotting a graph for a confidence range from 95 – 10% based on 175,000 requirement for testing.
thanks – any improvement opportunities or suggestions? It is a beta after all. cheers, Fred
I already logged in, but can’t find link to download the tool. Where I can find it? thanks.
Hi Jinfeng, the tool is web based thus not downloaded. You interact directly with the tools via your browser. cheers, Fred
Some thoughts:
1) Have you identified a target audience for your software? Target sector? Target industry?
2) The portability of the GUI lends itself to a cell phone. Have you considered doing this as a cell phone app?
3) Could you offer an input-file option that would allow the user to define the statistical rank directly? I was attempting to analyze some CDC life expectancy data with this software (N=10^5). There is a way (sort of) to get around the N=100 limitation by picking data that is close to where statistical rankings would occur.
4) What is your Elevator Pitch? Why would this be better than a fit to data using ln(g(R)) vs ln(ln( f(time)) ?
Hi Johan, thanks for the notes on the tool. good idea on the phone app idea – I wonder how easy it would be to input data there? For the other questions, we’ll have to ask the developer.
Cheers,
Fred
Hi Fred,
Nice tool.
It would be nice to be able to download the ranks after application of MON from the adjusted data set as processed by the algorithms.
I like the event density plot(s) as they help highlight issues.
Philip
I do not see a contour plot available if I upload a data set.
Also – a future nice to have feature would be to overlay several data sets on the same plot.
OK, I have downloaded data and would like to see it on the log-log Weibull plot.
But I see a BLANK space in the Weibull Curve box, as shown below. Any thoughts?
I put my data in the CSV template that was available within the software.