An online application of edge-cloud computing for lithium-ion battery with SOC estimation

The adoption of cloud and edge computing for several application motivates the monitoring and controlling a Lithium-ion batteries (LIB) online and remotely. In these scenarios, this work presents a case of study by the application of a hybrid concept of edge-cloud computing. The edge collects the data from the cells and sent directly to the cloud where several Machine Learning algorithms can be built, trained and tested. Then, the trained models are downloaded to the edge and do inferences in real time. The proposed idea presented still an SOC estimator based on Long Short-Term Memory (LSTM) with average error below 5%.