r future, so a controlled maintenance event can be scheduled. Next, real-time failure warning notifications are sent to Amazon SNS. Predictions are displayed on the Amazon QuickSight dashboard. Once completed, predictions are stored in Amazon Aurora for recordkeeping and future improvements. AWS Lambdauses the previouslytrained predictions model on the sensor logs to make predictions. Then, the sensor logs are sent to AWSLambdafor analysis (running the model against the data). Telemetry messages are sent to AWS Lambda from IOT Core for analysis by the model. Connected vehicle sendssensor logs to AWSIoT Core.Logs can also be sent using HTTP interface. Amazon SageMaker model is deployed after the model is trained to predict battery failures. A
