Estimating the lithium-ion batteries Remaining Discharge Time (RDT) are significantly affected by the condition of individual cells, yet most existing models typically focus on predicting battery-level performance. Accurate predictions of RDT at the cell level are essential for optimizing battery management systems. To predict RDT of lithium-ion cell Machine Learning (ML) is used. This study introduces a hybrid CNN-LSTM to predict the RDT of lithium-ion cell. By leveraging both convolutional and recurrent neural networks, the model effectively captures temporal and spatial relationships in cell data, leading to more precise RDT estimations. The proposed approach offers improved accuracy in predicting cell degradation and failure, enhancing battery performance and reliability. The model demonstrated a significant improvement in prediction accuracy, as reflected by a reduced Mean Absolute Error (MAE) and a Root Mean Square Error (RMSE). The achieved accuracy of above 96.5% demonstrates the model’s effectiveness in making accurate RDT predictions. \(\dots \)