Failure rate is the frequency with which any system or component fails, expressed in failures per unit of time. It thus depends on the system conditions, time interval, and total number of systems under study. [ 1 ] It can describe electronic, mechanical, or biological systems, in fields such as systems and reliability engineering, medicine and biology , or insurance and finance . It is usually denoted by the Greek letter λ {\displaystyle \lambda } ( lambda ). In real-world applications, the failure probability of a system usually differs over time; failures occur more frequently in early-life ( "burning in" ), or as a system ages ( "wearing out" ). This is known as the bathtub curve , where the middle region is called the "useful life period". Mean time between failures (MTBF) The mean time between failures (MTBF, 1 / λ {\displaystyle 1/\lambda } ) is often reported instead of the failure rate, as numbers such as "2,000 hours" are more intuitive than numbers such as "0.0005 per hour". However, this is only valid if the failure rate λ ( t ) {\displaystyle \lambda (t)} is actually constant over time, such as within the flat region of the bathtub curve. In many cases where MTBF is quoted, it refers only to this region; thus it cannot be used to give an accurate calculation of the average lifetime of a system, as it ignores the "burn-in" and "wear-out" regions. MTBF appears frequently in engineering design requirements, and governs the frequency of required system maintenance and inspections. A similar ratio used in the transport industries , especially in railways and trucking, is "mean distance between failures" (MDBF) - allowing maintenance to be scheduled based on distance travelled, rather than at regular time intervals. Mathematical definition The simplest definition of failure rate λ {\displaystyle \lambda } is simply the number of failures Δ n {\displaystyle \Delta n} per time interval Δ t {\displaystyle \Delta t} : λ = Δ n Δ t {\displaystyle \lambda ={\frac {\Delta n}{\Delta t}}} which would depend on the number of systems under study, and the conditions over the time period. Failures over time To accurately model failures over time, a cumulative failure distribution , F ( t ) {\displaystyle F(t)} must be defined, which can be any cumulative distribution function (CDF) that gradually increases from 0 {\displaystyle 0} to 1 {\displaystyle 1} . In the case of many identical systems, this may be thought of as the fraction of systems failing over time t {\displaystyle t} , after all starting operation at time t = 0 {\displaystyle t=0} ; or in the case of a single system, as the probability of the system having its failure time T {\displaystyle T} before time t {\displaystyle t} : F ( t ) = Pr ( T ≤ t ) . {\displaystyle F(t)=\Pr(T\leq t).} As CDFs are defined by integrating a probability density function , the failure probability density f ( t ) {\displaystyle f(t)} is defined such that: F ( t ) = ∫ 0 t f ( τ ) d τ {\displaystyle F(t)=\int _{0}^{t}f(\tau )\,d\tau } where τ {\displaystyle \tau } is a dummy integration variable. Here f ( t ) {\displaystyle f(t)} can be thought of as the instantaneous failure rate , i.e. the probability of failure in the time interval between t {\displaystyle t} and t + Δ t {\displaystyle t{+}\Delta t} , as Δ t {\displaystyle \Delta t} tends towards 0 {\displaystyle 0} : f ( t ) = lim Δ t → 0 + Pr ( t < T ≤ t + Δ t ) Δ t . {\displaystyle f(t)=\lim _{\Delta t\to 0^{+}}{\frac {\Pr(t<T\leq t{+}\Delta t)}{\Delta t}}.} A concept closely related but different [ 2 ] to instantaneous failure rate f ( t ) {\displaystyle f(t)} is the hazard rate (or hazard function ), h ( t ) {\displaystyle h(t)} . In the many-system case, this is defined as the proportional failure rate of the systems still functioning at time t {\displaystyle t} – as opposed to f ( t ) {\displaystyle f(t)} , which is the expressed as a proportion of the initial number of systems. For convenience we first define the reliability (or survival function ) as: R ( t ) = 1 − F ( t ) = Pr ( T > t ) {\displaystyle R(t)=1-F(t)=\Pr(T>t)} then the hazard rate is simply the instantaneous failure rate, scaled by the fraction of surviving systems at time t {\displaystyle t} : h ( t ) = f ( t ) R ( t ) {\displaystyle h(t)={\frac {f(t)}{R(t)}}} In the probabilistic sense for a single system, this can be interpreted as the instantaneous failure rate under the conditional probability that the system or component has already survived to time t {\displaystyle t} : h ( t ) = lim Δ t → 0 + Pr ( t < T ≤ t + Δ t ∣ T > t ) Δ t . {\displaystyle h(t)=\lim _{\Delta t\to 0^{+}}{\frac {\Pr(t<T\leq t{+}\Delta t\mid T>t)}{\Delta t}}.} Conversion to cumulative failure rate To convert between h ( t ) {\displaystyle h(t)} and F ( t ) {\displaystyle F(t)} , we can solve the differential equation h ( t ) = f ( t ) R ( t ) = − R ′ ( t ) R ( t ) {\displaystyle h(t)={\frac {f(t)}{R(t)}}=-{\frac {R'(t)}{R(t)}}} with initial condition R ( 0 ) = 1 {\displaystyle R(0)=1} , which yields [ 2 ] F ( t ) = 1 − exp ⁡ ( − ∫ 0 t h ( τ ) d τ ) . {\displaystyle F(t)=1-\exp {\left(-\int _{0}^{t}h(\tau )d\tau \right)}.} Thus for a collection of identical systems, only one of hazard rate h ( t ) {\displaystyle h(t)} , failure probability density f ( t ) {\displaystyle f(t)} , or cumulative failure distribution F ( t ) {\displaystyle F(t)} need be defined. Confusion can occur as the notation λ ( t ) {\displaystyle \lambda (t)} for "failure rate" often refers to the function h ( t ) {\displaystyle h(t)} rather than f ( t ) . {\displaystyle f(t).} [ 3 ] Constant hazard rate model There are many possible functions that could be chosen to represent failure probability density f ( t ) {\displaystyle f(t)} or hazard rate h ( t ) {\displaystyle h(t)} , based on empirical or theoretical evidence, but the most common and easily-understandable choice is to set f ( t ) = λ e − λ t , {\displaystyle f(t)=\lambda e^{-\lambda t},} an exponential function with scaling constant λ {\displaystyle \lambda } . As seen in the figures above, this represents a gradually decreasing failure probability density. The CDF F ( t ) {\displaystyle F(t)} is then calculated as: F ( t ) = ∫ 0 t λ e − λ τ d τ = 1 − e − λ t , {\displaystyle F(t)=\int _{0}^{t}\lambda e^{-\lambda \tau }\,d\tau =1-e^{-\lambda t},} which can be seen to gradually approach 1 {\displaystyle 1} as t → ∞ , {\displaystyle t\to \infty ,} representing the fact that eventually all systems under study will fail. The hazard rate function is then: h ( t ) = f ( t ) R ( t ) = λ e − λ t e − λ t = λ . {\displaystyle h(t)={\frac {f(t)}{R(t)}}={\frac {\lambda e^{-\lambda t}}{e^{-\lambda t}}}=\lambda .} In other words, in this particular case only , the hazard rate is constant over time. This illustrates the difference in hazard rate and failure probability density - as the number of systems surviving at time t > 0 {\displaystyle t>0} gradually reduces, the total failure rate also reduces, but the hazard rate remains constant . In other words, the probabilities of each individual system failing do not change over time as the systems age - they are " memory-less ". For many systems, a constant hazard function may not be a realistic approximation; the chance of failure of an individual component may depend on its age. Therefore, other distributions are often used. For example, the deterministic distribution increases hazard rate over time (for systems where wear-out is the most important factor), while the Pareto distribution decreases it (for systems where early-life failures are more common). The commonly used Weibull distribution combines both of these effects, as do the log-normal and hypertabastic distributions. After modelling a given distribution and parameters for h ( t ) {\displaystyle h(t)} , the failure probability density f ( t ) {\displaystyle f(t)} and cumulative failure distribution F ( t ) {\displaystyle F(t)} can be predicted using the given equations. Measuring failure rate Failure rate data can be obtained in several ways. The most common means are: All components of a design, The functionality of each component, The failure modes of each component, The effect of each component failure mode on the product functionality, The ability of any automatic diagnostics to detect the failure, The design strength (de-rating, safety factors) and The operational profile (environmental stress factors). Given a component database calibrated with field failure data that is reasonably accurate, [ 4 ] the method can predict product level failure rate and failure mode data for a given application. The predictions have been shown to be more accurate [ 5 ] than field warranty return analysis or even typical field failure analysis given that these methods depend on reports that typically do not have sufficient detail information in failure records. [ 6 ] Decreasing failure rates A decreasing failure rate describes cases where early-life failures are common [ 7 ] and corresponds to the situation where h ( t ) {\displaystyle h(t)} is a decreasing function . This can describe, for example, the period of infant mortality in humans, or the early failure of transistors due to manufacturing defects. Decreasing failure rates have been found in the lifetimes of spacecraft - Baker and Baker commenting that "those spacecraft that last, last on and on." [ 8 ] [ 9 ] The hazard rate of aircraft air conditioning systems was found to have an exponentially decreasing distribution. [ 10 ] Renewal processes In special processes called renewal processes , where the time to recover from failure can be neglected, the likelihood of failure remains constant with respect to time. For a renewal process with DFR renewal function, inter-renewal times are concave. [ clarification needed ] [ 11 ] [ 12 ] Brown conjectured the converse, that DFR is also necessary for the inter-renewal times to be concave, [ 13 ] however it has been shown that this conjecture holds neither in the discrete case [ 12 ] nor in the continuous case. [ 14 ] Coefficient of variation When the failure rate is decreasing the coefficient of variation is ⩾ 1, and when the failure rate is increasing the coefficient of variation is ⩽ 1. [ clarification needed ] [ 15 ] Note that this result only holds when the failure rate is defined for all t ⩾ 0 [ 16 ] and that the converse result (coefficient of variation determining nature of failure rate) does not hold. Failure rates can be expressed using any measure of time, but hours is the most common unit in practice. Other units, such as miles, revolutions, etc., can also be used in place of "time" units. Failure rates are often expressed in engineering notation as Failures Per Million Hours ( FPMH ), or 10 −6 , especially for individual components, since their failure rates are often very low. The Failures In Time ( FIT ) rate of a device is the number of failures that can be expected in one billion (10 9 ) device-hours of operation [ 17 ] (e.g. 1,000 devices for 1,000,000 hours, or 1,000,000 devices for 1,000 hours each, or some other combination). This term is used particularly by the semiconductor industry. Combinations of failure types If a complex system consists of many parts, and the failure of any single part means the failure of the entire system, and the chance of failure for each part is conditionally independent of the failure of any other part, then the total failure rate is simply the sum of the individual failure rates of its parts however, this assumes that the failure rate λ ( t ) {\displaystyle \lambda (t)} is constant, and that the units are consistent (e.g. failures per million hours), and not expressed as a ratio or as probability densities. This is useful to estimate the failure rate of a system when individual components or subsystems have already been tested. [ 18 ] [ 19 ] When adding "redundant" components to eliminate a single point of failure , the quantity of interest is not the sum of individual failure rates but rather the "mission failure" rate, or the "mean time between critical failures" (MTBCF). [ 20 ] Combining failure or hazard rates that are time-dependent is more complicated. For example, mixtures of Decreasing Failure Rate (DFR) variables are also DFR. [ 11 ] Mixtures of exponentially distributed failure rates are hyperexponentially distributed . Suppose it is desired to estimate the failure rate of a certain component. Ten identical components are each tested until they either fail or reach 1,000 hours, at which time the test is terminated. A total of 7,502 component-hours of testing is performed, and 6 failures are recorded. The estimated failure rate is: which could also be expressed as a MTBF of 1,250 hours, or approximately 800 failures for every million hours of operation. Mathematics portal Annualized failure rate Failure modes, effects, and diagnostic analysis Force of mortality Frequency of exceedance Reliability engineering Reliability theory Reliability theory of aging and longevity Survival analysis Weibull distribution ↑ MacDiarmid, Preston; Morris, Seymour; et al. (n.d.). Reliability Toolkit (Commercial Practices ed.). Rome, New York: Reliability Analysis Center and Rome Laboratory. pp. 35– 39. 1 2 Todinov, MT (2007). "Chapter 2.2 HAZARD RATE AND TIME TO FAILURE DISTRIBUTION". Risk-Based Reliability Analysis and Generic Principles for Risk Reduction . ↑ Wang, Shaoping (2016). "Chapter 3.3.1.3: Failure Rate λ(t)". Comprehensive Reliability Design of Aircraft Hydraulic System . ↑ Electrical & Mechanical Component Reliability Handbook . exida. 2006. ↑ Goble, William M.; Iwan van Beurden (2014). Combining field failure data with new instrument design margins to predict failure rates for SIS Verification . Proceedings of the 2014 International Symposium - BEYOND REGULATORY COMPLIANCE, MAKING SAFETY SECOND NATURE, Hilton College Station-Conference Center, College Station, Texas. ↑ W. M. Goble, "Field Failure Data – the Good, the Bad and the Ugly," exida, Sellersville, PA ↑ Finkelstein, Maxim (2008). "Introduction". Failure Rate Modelling for Reliability and Risk . Springer Series in Reliability Engineering. pp. 1– 84. doi : 10.1007/978-1-84800-986-8_1 . ISBN 978-1-84800-985-1 . ↑ Baker, J. C.; Baker, G. A. S. . (1980). "Impact of the space environment on spacecraft lifetimes". Journal of Spacecraft and Rockets . 17 (5): 479. Bibcode : 1980JSpRo..17..479B . doi : 10.2514/3.28040 . ↑ Saleh, Joseph Homer; Castet, Jean-François (2011). "On Time, Reliability, and Spacecraft". Spacecraft Reliability and Multi-State Failures . p. 1. doi : 10.1002/9781119994077.ch1 . ISBN 9781119994077 . ↑ Proschan, F. (1963). "Theoretical Explanation of Observed Decreasing Failure Rate". Technometrics . 5 (3): 375– 383. doi : 10.1080/00401706.1963.10490105 . JSTOR 1266340 . 1 2 Brown, M. (1980). "Bounds, Inequalities, and Monotonicity Properties for Some Specialized Renewal Processes" . The Annals of Probability . 8 (2): 227– 240. doi : 10.1214/aop/1176994773 . JSTOR 2243267 . 1 2 Shanthikumar, J. G. (1988). "DFR Property of First-Passage Times and its Preservation Under Geometric Compounding" . The Annals of Probability . 16 (1): 397– 406. doi : 10.1214/aop/1176991910 . JSTOR 2243910 . ↑ Brown, M. (1981). "Further Monotonicity Properties for Specialized Renewal Processes" . The Annals of Probability . 9 (5): 891– 895. doi : 10.1214/aop/1176994317 . JSTOR 2243747 . ↑ Yu, Y. (2011). "Concave renewal functions do not imply DFR interrenewal times". Journal of Applied Probability . 48 (2): 583– 588. arXiv : 1009.2463 . doi : 10.1239/jap/1308662647 . S2CID 26570923 . ↑ Wierman, A. ; Bansal, N.; Harchol-Balter, M. (2004). "A note on comparing response times in the M/GI/1/FB and M/GI/1/PS queues" (PDF) . Operations Research Letters . 32 : 73– 76. doi : 10.1016/S0167-6377(03)00061-0 . ↑ Gautam, Natarajan (2012). Analysis of Queues: Methods and Applications . CRC Press. p. 703. ISBN 978-1439806586 . ↑ Xin Li; Michael C. Huang; Kai Shen; Lingkun Chu. "A Realistic Evaluation of Memory Hardware Errors and Software System Susceptibility" . 2010. p. 6. ↑ "Reliability Basics" . 2010. ↑ Vita Faraci. "Calculating Failure Rates of Series/Parallel Networks" Archived 2016-03-03 at the Wayback Machine . 2006. ↑ "Mission Reliability and Logistics Reliability: A Design Paradox" . Further reading Goble, William M. (2018), Safety Instrumented System Design: Techniques and Design Verification , Research Triangle Park, NC: International Society of Automation Blanchard, Benjamin S. (1992). Logistics Engineering and Management (Fourth ed.). Englewood Cliffs, New Jersey: Prentice-Hall. pp. 26– 32. ISBN 0135241170 . Ebeling, Charles E. (1997). An Introduction to Reliability and Maintainability Engineering . Boston: McGraw-Hill. pp. 23– 32. ISBN 0070188521 . Federal Standard 1037C Kapur, K. C.; Lamberson, L. R. (1977). Reliability in Engineering Design . New York: John Wiley & Sons. pp. 8– 30. ISBN 0471511919 . Knowles, D. I. (1995). "Should We Move Away From 'Acceptable Failure Rate'?". Communications in Reliability Maintainability and Supportability . 2 (1). International RMS Committee, USA: 23. Modarres, M. ; Kaminskiy, M.; Krivtsov, V. (2010). Reliability Engineering and Risk Analysis: A Practical Guide (2nd ed.). CRC Press. ISBN 9780849392474 . Mondro, Mitchell J. (June 2002). "Approximation of Mean Time Between Failure When a System has Periodic Maintenance" (PDF) . IEEE Transactions on Reliability . 51 (2): 166– 167. doi : 10.1109/TR.2002.1011521 . Archived from the original (PDF) on 2008-12-17 . Retrieved 2008-07-14 . Rausand, M.; Hoyland, A. (2004). System Reliability Theory; Models, Statistical methods, and Applications . New York: John Wiley & Sons. ISBN 047147133X . Turner, T.; Hockley, C.; Burdaky, R. (1997). The Customer Needs A Maintenance-Free Operating Period . Leatherhead, Surrey, UK: ERA Technology Ltd . {{ cite book }} : | work= ignored ( help ) U.S. Department of Defense, (1991) Military Handbook, “Reliability Prediction of Electronic Equipment, MIL-HDBK-217F, 2 Fault Tolerant Computing in Industrial Automation Archived 2014-03-26 at the Wayback Machine by Hubert Kirrmann, ABB Research Center, Switzerland Mean Arithmetic Arithmetic-Geometric Contraharmonic Cubic Generalized/power Geometric Harmonic Heronian Heinz Lehmer Arithmetic-Geometric Generalized/power Average absolute deviation Coefficient of variation Interquartile range Standard deviation Central limit theorem Moments Kurtosis L-moments Skewness Index of dispersion Contingency table Frequency distribution Partial correlation Pearson product-moment correlation Rank correlation Kendall's τ Spearman's ρ Stem-and-leaf display Scatter Plot Matrix Data transformation Log transformation Power transform Box–Cox transformation Yeo–Johnson transformation Box–Cox transformation Yeo–Johnson transformation Variance-stabilizing transformation Anscombe transform Fisher transformation Feature scaling Standardization (z-score) Min–max normalization Unit vector normalization Dimensionality reduction Principal component analysis Factor analysis Seasonal adjustment

Stationarity transformation Sample size determination Statistical power Sampling Cluster Stratified Factorial experiment Random assignment Randomized controlled trial Randomized experiment Scientific control Adaptive clinical trial Stochastic approximation Up-and-down designs Cross-sectional study Natural experiment Quasi-experiment Probability distribution Sampling distribution Order statistic Order statistic Empirical distribution Density estimation Density estimation Statistical model Model specification L p space Model specification Parameter location scale shape Parametric family Likelihood (monotone) Location–scale family Exponential family Likelihood (monotone) Location–scale family Exponential family Statistical functional Bootstrap U V Optimal decision loss function Statistical distance divergence Robustness ( Sensitivity analysis ) Estimating equations Maximum likelihood Method of moments M-estimator Minimum distance Maximum likelihood Method of moments Minimum distance Unbiased estimators Mean-unbiased minimum-variance Rao–Blackwellization Lehmann–Scheffé theorem Median unbiased Mean-unbiased minimum-variance Rao–Blackwellization Lehmann–Scheffé theorem Rao–Blackwellization Lehmann–Scheffé theorem Median unbiased Confidence interval Likelihood interval Prediction interval Tolerance interval Resampling Bootstrap Jackknife Power Uniformly most powerful test Uniformly most powerful test Permutation test Randomization test Randomization test Multiple comparisons Likelihood-ratio G -test Score/Lagrange multiplier Wald Z -test (normal) Z -test (normal) Student's t -test Kolmogorov–Smirnov Anderson–Darling Normality (Shapiro–Wilk) Model selection Cross validation AIC BIC Cross validation Sign Sample median Signed rank (Wilcoxon) Hodges–Lehmann estimator Hodges–Lehmann estimator Rank sum (Mann–Whitney) Nonparametric anova 1-way (Kruskal–Wallis) 2-way (Friedman) Ordered alternative (Jonckheere–Terpstra) 1-way (Kruskal–Wallis) 2-way (Friedman) Ordered alternative (Jonckheere–Terpstra) Van der Waerden test Bayesian probability prior posterior Credible interval Bayesian estimator Maximum posterior estimator Maximum posterior estimator Regression analysis Pearson product-moment Partial correlation Confounding variable Coefficient of determination Errors and residuals Regression validation Mixed effects models Simultaneous equations models Multivariate adaptive regression splines (MARS) Template:Least squares and regression analysis Simple linear regression Ordinary least squares General linear model Bayesian regression Nonlinear regression Homoscedasticity and Heteroscedasticity Exponential families Logistic (Bernoulli) / Binomial / Poisson regressions Analysis of variance (ANOVA, anova) Analysis of covariance Multivariate ANOVA Degrees of freedom Contingency table Graphical model Log-linear model Cochran–Mantel–Haenszel statistics Principal components Canonical correlation Discriminant analysis Cluster analysis Structural equation model Factor analysis Factor analysis Multivariate distributions Elliptical distributions Normal Elliptical distributions Normal Seasonal adjustment Exponential smoothing Structural break Granger causality Q-statistic (Ljung–Box) Breusch–Godfrey Autocorrelation (ACF) partial (PACF) Cross-correlation (XCF) ARIMA model (Box–Jenkins) Autoregressive conditional heteroskedasticity (ARCH) Vector autoregression (VAR) ( Autoregressive model (AR) ) Spectral density estimation Fourier analysis Least-squares spectral analysis Whittle likelihood Kaplan–Meier estimator (product limit) Proportional hazards models Accelerated failure time (AFT) model First hitting time Nelson–Aalen estimator Clinical trials / studies Medical statistics Methods engineering Probabilistic design Process / quality control System identification Actuarial science Crime statistics National accounts Official statistics Population statistics Environmental statistics Geographic information system Mathematics portal Actuarial science Engineering failures Reliability engineering Survival analysis Statistical ratios Articles with short description Short description matches Wikidata Wikipedia articles needing clarification from December 2024 Webarchive template wayback links CS1 errors: periodical ignored