rk.weibull: Reliability & Life Data Analysis for RKWard
rk.weibull is an external plug-in for RKWard that brings industrial-grade Reliability Engineering tools to the RKWard environment. It interfaces with the open-source WeibullR package to perform Life Data Analysis (Weibull Analysis) and Warranty Prediction, serving as a free alternative to software like ReliaSoft Weibull++ or Minitab's Reliability module.
Features
1. Life Data Analysis (Weibull)
-
2-Parameter Weibull Fitting: Fits the Slope ($\beta$) and Characteristic Life ($\eta$) to failure data.
-
Flexible Data Input: Handles complete data (failures) and right-censored data (suspensions).
-
Analysis Methods:
-
Maximum Likelihood Estimation (MLE): Recommended for larger datasets ($N > 30$). Automatically applies Fisher Matrix confidence bounds.
-
Rank Regression (RR): Standard Least Squares method ($X$ on $Y$ or $Y$ on $X$); automatically applies Pivotal confidence bounds.
-
Visualization: Generates standard Weibull Probability Plots (Weibull Canvas) with 95% confidence intervals.
🌍 Internationalization
The interface is fully localized to match your RKWard language settings:
- 🇺🇸 English (Default)
- 🇪🇸 Spanish (
es)
- 🇫🇷 French (
fr)
- 🇩🇪 German (
de)
- 🇧🇷 Portuguese (Brazil) (
pt_BR)
2. Warranty Prediction (B-Life)
-
B-Life Calculation: Calculates the time by which a specific percentage of units will fail (e.g., $B_{10}$ Life).
-
Integration: Directly uses the fitted model objects from the Life Data Analysis component, ensuring consistency between the plot and the prediction.
-
Robust Parameter Search: Automatically detects and extracts parameters regardless of
WeibullR's internal naming conventions (case-insensitive search for beta/eta or shape/scale).
Dependencies
This plugin relies on the following R packages:
WeibullR
-
rkwarddev (for development/building)
Installation
Installing from GitHub (Recommended)
- You can install the latest version directly from GitHub using the
devtools or remotes package in R:
library(devtools)
install.packages("rk.weibull", repos = c('https://alfcano.r-universe.dev', 'https://cloud.r-project.org'))
- Restart RKWard.
Usage Guide
The plugin adds a new entry to the main menu under:
Analysis > Industrial Stats > Reliability
Example: Analyzing Failure Data
Scenario: You have tested 50 components. Most failed, but some tests were suspended. You want to determine the $B_{10}$ life (when 10% will fail).
-
Generate Data:
library(WeibullR)
times <- rweibull(50, shape=2, scale=100)
event <- rep(1, 50)
-
Step 1: Fit the Model
- Navigate to Analysis > Industrial Stats > Reliability > Life Data Analysis.
-
Time: Select
times.
-
Event: Select
event.
-
Method: Select Maximum Likelihood (MLE).
-
Save Weibull Object: Enter
my_weibull_fit.
- Submit.
-
Output: Check the "Parameter Estimates" table ($\beta \approx 2$, $\eta \approx 100$) and the Probability Plot.
-
Step 2: Predict Warranty Life
- Navigate to Analysis > Industrial Stats > Reliability > Warranty Prediction.
-
Fitted Weibull Object: Select
my_weibull_fit (from the previous step).
-
Probability: Enter
0.1 (for 10% unreliability).
- Submit.
-
Output: The "Predicted Life (Time)" value is your $B_{10}$ life.
Author
Alfonso Cano
Benemérita Universidad Autónoma de Puebla
License: GPL (>= 3)
This plugin was developed with the assistance of Gemini, a large language model by Google.