The machines used by humans are becoming more and more complex and intelligent, and the environments in which machines are used are more diverse. Reliability prediction has become a very challenging research topic, attracting more and more interested researchers. Many products are repairable systems, such as vehicles, boats, and aircraft. Deep learning is the latest artificial intelligence technology. Long short-term memory (LSTM) and CNNLSTM models among neural network models can be used to predict temporal data. Based above, this study proposes the LSTM and CNN-LSTM models for reliability prediction of repairable products. Both academia and industry can benefit from this research.