Overview
IEC 61164:2004, "Reliability growth - Statistical test and estimation methods" is an IEC standard that defines statistical models and numerical procedures for assessing and tracking reliability growth during product development and test programmes. It provides methods to estimate model parameters, compute confidence intervals, perform goodness‑of‑fit tests, and project future reliability based on failure data collected in reliability improvement programmes. The standard complements IEC 61014 and includes worked examples and algorithms suitable for implementation in software tools.
Key Topics
- Reliability growth models - Continuous and discrete models for planning and tracking reliability improvement, including the widely used power law model and additional models introduced in this edition (e.g., modified power law, modified Bayesian IBM‑Rosner model, discrete reliability growth model, fixed‑number‑of‑faults model).
- Statistical estimation - Procedures for parameter estimation from failure-time and accumulated test-time data.
- Growth tests and projections - Tests for detecting reliability growth, projection techniques to forecast future MTBF (mean time between failures) or failure rate.
- Confidence intervals and uncertainty - Methods to compute confidence intervals on shape parameters and on current MTBF or failure intensity.
- Goodness‑of‑fit testing - Statistical tests (e.g., Cramér‑von Mises and related methods) to assess model fit to observed failure data.
- Test types and data requirements - Support for Type I (time‑terminated) and Type II (failure‑terminated) tests; guidance on required inputs (initial reliability, design improvement plans, accumulated test times at failures).
- Worked examples and annexes - Real‑life examples, tables, and background material (Annex A and B) demonstrating model application and numerical algorithms.
Applications
IEC 61164 is practical for professionals involved in reliability planning, testing and assurance:
- Reliability engineers applying statistical models to quantify design improvements and to track growth during validation testing.
- Test managers designing Type I/Type II reliability growth tests and interpreting results.
- Design teams and program managers planning reliability improvement activities and setting reliability goals.
- Quality and systems engineers estimating MTBF, producing confidence intervals and performing model fit checks for field or test data.
- Procurement and regulatory stakeholders evaluating supplier reliability growth claims.
Practical uses include planning design‑phase improvements, monitoring test‑phase reliability growth, projecting remaining test time to achieve goals, and documenting statistical evidence of reliability improvement.
Related Standards
- IEC 61014:2003 - Programmes for reliability growth (to be used in conjunction with IEC 61164).
- Other referenced IEC testing standards (IEC 60605 series, IEC 60300‑3‑5) for complementary test principles and statistical procedures.
Keywords: IEC 61164, reliability growth, statistical test, estimation methods, power law model, MTBF, confidence intervals, goodness‑of‑fit, reliability improvement programme, IEC 61014.