A repairable system operates under a maintenance strategy that calls for complete preventive repair actions at prescheduled times and minimal repair actions whenever a failure occurs. Under minimal repair, failures are modeled according to a nonhomogeneous Poisson process. When the intensity function is assumed to grow proportional to a power of time, this process is called the power law process. Assuming that the system will be in operation for an infinite time, we find the expected cost per unit of time for each preventive maintenance policy and hence obtain the optimal strategy as a function of the intensity function of the process. Large-sample procedures to estimate the optimal maintenance check points for the power law process are also discussed. The results are applied in a real data set concerning failures histories of power transformers.
Os dados provenientes de estudos longitudinais se caracterizam pela seqüência de duas ou mais observações em cada indivíduo. Nos estudos de coorte, esses dados geralmente apresentam estrutura desbalanceada. Uma casuística que envolve a avaliação longitudinal de crescimento de lactentes nascidos de mães infectadas pelo HIV foi acompanhada no ambulatório de AIDS pediátrica do Hospital das Clínicas da Universidade Federal de Minas Gerais, Minas Gerais, Brasil. O objetivo deste estudo é demonstrar a aplicação do modelo linear misto na análise de dados longitudinais desbalanceados provenientes dessa coorte. Os resultados mostram que, aos seis meses de idade, os meninos eram, em média, 1,8cm maiores que as meninas e as crianças sororrevertoras eram, em média, 2,9cm maiores que as infectadas. Aos 12 meses, a diferença na altura entre meninos e meninas passou a ser, em média, de 2,4cm enquanto a diferença entre infectados e sororrevertores passou a ser, em média, de 3,5cm. Além de descrever o comportamento longitudinal do crescimento, o modelo também permite estimar a velocidade de crescimento das crianças por sexo e grupo.
Traditionally, reliability assessment of devices has been based on (accelerated) life tests. However, for highly reliable products, little information about reliability is provided by life tests in which few or no failures are typically observed. Since most failures arise from a degradation mechanism at work for which there are characteristics that degrade over time, one alternative is monitor the device for a period of time and assess its reliability from the changes in performance (degradation) observed during that period. The goal of this article is to illustrate how degradation data can be modeled and analyzed by using "classical" and Bayesian approaches. Four methods of data analysis based on classical inference are presented. Next we show how Bayesian methods can also be used to provide a natural approach to analyzing degradation data. The approaches are applied to a real data set regarding train wheels degradation.
To describe and evaluate notified scorpion envenomation incidents in children and adolescents in the Brazilian state of Minas Gerais between 2001 and 2005, and to calculate the incidence and mortality and identify factors related to fatal outcomes. Analysis of 14,406 mandatory notifications of scorpion envenomation involving patients aged 0 to 19 years in Minas Gerais. After performing a univariate analysis of factors related to mortality (p < 0.25), multivariate analysis was carried out using logistic regression. The number of notifications of accidents involving scorpions increased over the period. The age group that suffered most accidents was 5 to 9 years. Local pain was reported in 95% of cases. The majority of cases were classified as mild (60%) and antivenom was given to 42% of this population. The mean mortality rate was 0.7%. The state capital, Belo Horizonte, was the municipality with the greatest occurrence of incidents. Each hour's delay before receiving first medical care was associated with a 9% increase in the odds ratio for the likelihood of death, which was 46 times greater in the severe group than in the moderate and mild groups. A decrease of 1 year in patient age was associated with a 13% increase in the odds ratio for the likelihood of death. There was an increase in the number of notifications over the period studied. The following factors were identified as linked with a higher odds ratio for the likelihood of death: delayed first medical care, being classified as a severe case, and younger age of victim.
1 January 1996 in IEEE Transactions on Reliability