In this study the Maximum Likelihood Estimator is utilized to identify the characteristics of failure of class-H insulation by considering accelerated life test data under censored situations from Nelson. The hazard rate function is considered in terms of the reliability, h(R), so-called AE model. The AE model is used to model the failures which are expressed as the serial connection between three modes, namely the turn, phase, and ground. This is the so-called competing failure. The main concern in the present investigation relates to the characteristic of changes in cumulative damage with temperature. The characteristic of the damage process basically change, with less capability of cumulation. The failure tends to be unpredictable in a constant hazard rate situation in much higher temperature environments. The parameters of the model are related to the temperature and follow the Arrhenius law. The numerical results indicate that the AE model is well fitted to the data and gives more information to identify the failure modes with fewer parameters. This is better than the using Weibull distribution with both parameters varied with temperature. According to the predicted lifetime, the turn needs to be rearranged primarily, followed by the phase. The ground mode only has influence on the failure at much higher temperatures.