In reality, failure data are often collected under diffract operational conditions (covariates), leading to heterogeneity among the data. Heterogeneity can be classified as observed and unobserved heterogeneity. Un-observed heterogeneity is the effect of unknown, unrecorded, or missing covariates. In most reliability studies, the effect of unobserved covariates is neglected. This may lead to inaccurate reliability modeling, and consequently, wrong operation and maintenance decisions. There is a lack of a systematic approach to model the unobserved covariate in reliability analysis. This paper aims to present a framework for reliability analysis in the presence of unobserved and observed covariates. Here, the unobserved covariates will be analyzed using frailty models. A case study will illustrate the application of the framework.
LouitDMPascualRJardineAK. A practical procedure for the selection of time-to-failure models based on the assessment of trends in maintenance data. Reliab Eng Syst Saf2009; 94(10): 1618–1628.
2.
LiuX. Survival models on unobserved heterogeneity and their applications in analyzing large-scale survey data. J Biom Biostat2014; 5: 1000191.
3.
LawlessJF. Statistical models and methods for lifetime data. John Wiley & Sons, 2011, 662 p.
4.
FinkelsteinM. Imperfect repair and lifesaving in heterogeneous populations. Reliab Eng Syst Saf2007; 92(12): 1671–1676.
5.
ThompsonSGSharpSJ. Explaining heterogeneity in meta-analysis: a comparison of methods. Stat Med1999; 18(20): 2693–2708.
6.
HigginsJPThompsonSG. Quantifying heterogeneity in a meta-analysis. Stat Med2002; 21(11): 1539–1558.
7.
GarmabakiAAhmadiABlockJ, et al. A reliability decision framework for multiple repairable units. Reliab Eng Syst Saf2016; 150: 78–88.
8.
GuidaMGiorgioM. Reliability analysis of accelerated life-test data from a repairable system. IEEE Trans Reliab1995; 44(2): 337–346.
9.
GiorgioMGuidaMPulciniG. Repairable system analysis in presence of covariates and random effects. Reliab Eng Syst Saf2014; 131: 271–281.
10.
LawlessJF. Regression methods for Poisson process data. J Am Stat Assoc1987; 82(399): 808–815.
11.
KumarDKlefsjöB. Proportional hazards model: a review. Reliab Eng Syst Saf1994; 44(2): 177–188.
12.
GutierrezRG. Parametric frailty and shared frailty survival models. Stata J2002; 2(1): 22–44.
13.
HougaardP. Frailty models for survival data. Lifetime Data Anal1995; 1(3): 255–273.
14.
VaupelJWMantonKGStallardE. The impact of heterogeneity in individual frailty on the dynamics of mortality. Demography1979; 16(3): 439–454.
15.
BarabadiABarabadyJMarkesetT. Maintainability analysis considering time-dependent and time-independent covariates. Reliab Eng Syst Saf2011; 96(1): 210–217.
16.
AbbringJHVan Den BergGJ. The unobserved heterogeneity distribution in duration analysis. Biometrika2007; 94(1): 87–99.
17.
LeszczycPTPBassFM. Determining the effects of observed and unobserved heterogeneity on consumer brand choice. Appl Stoch Model Data Anal1998; 14(2): 95–115.
18.
BarabadiAGudmestadOTBarabadyJ. RAMS data collection under Arctic conditions. Reliab Eng Syst Saf2015;135:92–9.
19.
GaoXBarabadyJMarkesetT. An approach for prediction of petroleum production facility performance considering Arctic influence factors. Reliab Eng Syst Saf2010;95(8):837–46.
20.
GarmabakiAHSAhmadiAMahmoodYA, et al. Reliability modelling of multiple repairable units. Qual Reliab Eng Int2016; 32(7): 2329–2343.
21.
AsfawZGLindqvistBH. Unobserved heterogeneity in the power law nonhomogeneous Poisson process. Reliab Eng Syst Saf2015;134:59–65.
22.
LiMLiuJ. Bayesian hazard modeling based on lifetime data with latent heterogeneity. Reliab Eng Syst Saf2016; 145: 183–189.
23.
ClaytonDG. A model for association in bivariate life tables and its application in epidemiological studies of familial tendency in chronic disease incidence. Biometrika1978; 65(1): 141–151.
24.
AsadzadehSAghaieAShahriariH, et al. The application of proportional hazards and frailty models to multistage processes surveillance. Int J Adv Manuf Technol2014; 74(1): 461–470.
25.
SlimacekVLindqvistBH. Nonhomogeneous Poisson process with nonparametric frailty. Reliab Eng Syst Saf2016; 149: 14–23.
26.
AshaGRajaAVRavishankerN. Reliability modelling incorporating load share and frailty. Appl Stoch Models Bus Ind2018; 34(2): 206–223.
27.
XuMLiX. Negative dependence in frailty models. J Stat Plan Inference2008; 138(5): 1433–1441.
28.
MisraNGuptaNGuptaRD. Stochastic comparisons of multivariate frailty models. J Stat Plan Inference2009; 139(6): 2084–2090.
29.
SlimacekVLindqvistBH. Reliability of wind turbines modeled by a Poisson process with covariates, unobserved heterogeneity and seasonality. Wind Energy2016; 19(11): 1991–2002.
30.
FinkelsteinM. Shocks in homogeneous and heterogeneous populations. Reliab Eng Syst Saf2007; 92(5): 569–574.
31.
BarkerKBaroudH. Proportional hazards models of infrastructure system recovery. Reliab Eng Syst Saf2014; 124: 201–206.
32.
LiuBLiangZParlikadAK, et al. Condition-based maintenance for systems with aging and cumulative damage based on proportional hazards model. Reliab Eng Syst Saf2017; 168: 200–209.
33.
SlimacekVLindqvistBH. Nonhomogeneous Poisson process with nonparametric frailty and covariates. Reliab Eng Syst Saf2017; 167: 75–83.
34.
ChaJHFinkelsteinM. Some notes on unobserved parameters (frailties) in reliability modeling. Reliab Eng Syst Saf2014; 123: 99–103.
35.
BarabadiAGarmabakiAYuanF, et al. Maintainability analysis of equipment using point process models. In: 2015 IEEE international conference on industrial engineering and engineering management (IEEM), Singapore, 6–9 Decemeber 2015, pp.797–801. New York: IEEE.
36.
MaZKringsAW. Multivariate survival analysis (I): shared frailty approaches to reliability and dependence modeling. In: 2008 IEEE Aerospace Conference, Big Sky, MT, USA, 1–8 March 2008, pp.1–21. New York: IEEE.
37.
BarabadiABarabadyJMarkesetT. Application of reliability models with covariates in spare part prediction and optimization–a case study. Reliab Eng Syst Saf2014; 123: 1–7.
38.
LagakosSW. The graphical evaluation of explanatory variables in proportional hazard regression models. Biometrika1981; 68(1): 93–98.
39.
ArjasE. A graphical method for assessing goodness of fit in Cox’s proportional hazards model. J Am Stat Assoc1988; 83(401): 204–212.
40.
GrayRJ. Some diagnostic methods for cox regression models through hazard smoothing. Biometrics1990; 46(1): 93–102.
41.
CoxDR. A note on the graphical analysis of survival data. Biometrika. 1979; 66(1): 188–190.
42.
SchoenfeldD. Partial residuals for the proportional hazards regression model. Biometrika. 1982; 69(1): 239–241.
43.
PettittANDaudIB. Investigating time dependence in Cox’s proportional Hazards model. J R Stat Soc Ser C Appl Stat1990; 39(3): 313–329.
44.
ParkSHendryDJ. Reassessing Schoenfeld residual tests of proportional hazards in political science event history analyses. Am J Pol Sci2015; 59(4): 1072–1087.
45.
KalbfleischJDPrenticeRL. The statistical analysis of failure time data. Hoboken, NJ: John Wiley & Sons, Inc., 2011.
46.
SchoenfeldD. Chi-squared goodness-of-fit tests for the proportional hazards regression model. Biometrika1980; 67(1): 145–153.
47.
O’QuigleyJPessioneF. Score tests for homogeneity of regression effect in the proportional hazards model. Biometrics1989; 45(1): 135–144.
48.
McKeagueIWUtikalKJ. Goodness-of-fit tests for additive hazards and proportional hazards models. Scand J Stat1991; 18(3): 177–195.
49.
GillRSchumacherM. A simple test of the proportional hazards assumption. Biometrika1987; 74(2): 289–300.
50.
HorowitzJLNeumannGR. A generalized moments specification test of the proportional hazards model. J Am Stat Assoc1992; 87(417): 234–240.
51.
KimberAC. A Weibull-based score test for heterogeneity. Lifetime Data Anal. 1996; 2(1): 63–71.
52.
GrayRJ. Tests for variation over groups in survival data. J Am Stat Assoc2012; 90: 198–203.
53.
CommengesDAndersenPK. Score test of homogeneity for survival data. Lifetime Data Anal1995; 1(2): 145–156.
54.
MittlböckMHeinzlH. A simulation study comparing properties of heterogeneity measures in meta-analyses. Stat Med2006; 25(24): 4321–4333.
55.
IoannidisJPA. Interpretation of tests of heterogeneity and bias in meta-analysis. J Eval Clin Pract2008; 14(5): 951–957.
56.
Huedo-MedinaTBSánchez-MecaJMarín-MartínezF, et al. Assessing heterogeneity in meta-analysis: Q statistic or I2 index?Psychol Methods2006; 11(2): 193.
57.
BrewerMJButlerACooksleySL. The relative performance of AIC, AICC and BIC in the presence of unobserved heterogeneity. Methods Ecol Evol2016; 7(6): 679–692.
58.
YurekliKKuruncA. Performances of stochastic approaches in generating low streamflow data for drought analysis. J Spat Hydrol2005; 5(1): 20–32.
59.
KvaløyJTLindqvistBH. TTT-based tests for trend in repairable systems data. Reliab Eng Syst Saf1998; 60(1): 13–28.
60.
LeeETWangJ. Statistical methods for survival data analysis. John Wiley & Sons, 2003, 528 p.
61.
CoppiniDBowtellPWengC, et al. Showing neuropathy is related to increased mortality in diabetic patients—a survival analysis using an accelerated failure time model. J Clin Epidemiol2000; 53(5): 519–523.
