Serviceability is one of the most important factors in machineries associated with aerospace, nuclear plant, power plant, turbine generator system, liquid engines, spacecraft etc. The lack of service ability can raise potential threat to safety, production and economy. To improve the service ability proper maintenance technique is necessary. The traditional data-driven methods of predicting remaining useful life (RUL) have proven to be inefficient. Recently prognostic and health management (PHM) has gained popularity in predicting the service ability or remaining useful life (RUL) of the machineries. Due to the special features such as automatic feature extraction capability, massive data handling ability, and lack of human intervention, deep learning has emerged as a tool for PHM.