In this study, we focus on the relevance of remaining life prediction of randomly degraded equipment in the context of big data monitoring and the core issue of quantifying uncertainty in remaining life prediction. We analyze the limitations and common problems of current research. To address the limitations and common problems, a solution for predicting the remaining life of randomly degraded devices with multisource sensing monitoring in the context of big data is proposed, and the feasibility and effectiveness of the idea are verified using battery data. Finally, multiple machine learning methods, such as support vector machines, random forests, recurrent neural networks, and convolutional neural networks, are combined to predict the remaining life of batteries, and these four machine learning methods perform well in the work of battery remaining life prediction and solve the key scientific problems.