Important: This Guidance requires the use of Amazon Forecast, which is no longer available to new customers. Existing customers of Amazon Forecast can continue using and deploying this Guidance as normal.
This Guidance demonstrates how to use historical battery health data with artificial intelligence and machine learning (AI/ML) algorithms to improve the accuracy of battery State of Health (SoH) and Remaining Useful Life (RUL) estimations. Currently, these estimations largely rely on a static formula-based approach, which can provide near-term battery health information. Using this Guidance, automotive original equipment manufacturers (OEMs) can predict battery SoH and RUL into the future with easy-to-train AI/ML models built using historical data stored in the Cloud.
Predictions of battery health will help OEMs and EV owners proactively plan for battery replacement, and most importantly, can be used to move battery into a new life and promote the overall circular economy of a battery. OEMs can retrain these models at regular intervals using incoming battery health status data, and monitor the battery fleet health using out-of-the box dashboards. Along with information such as driving trends, charge and discharge behaviors, OEMs can also use this Guidance to provide EV owners with recommendations on how to slow SoH decline and extend the battery lifespan.
Architecture Diagram
Step 1
Battery health data is ingested through the Connected Mobility Platform.
Step 2
Streaming data ingested through the Connected Mobility Platform is stored in Amazon Timestream for near real-time monitoring. Amazon Simple Storage Service (Amazon S3) is used to store historical battery data.
Amazon DynamoDB stores the state of each phase of the machine learning (ML) pipeline, including pre-processing, model generation and prediction, post-processing, and drift monitoring.
Step 3
When new data becomes available, AWS Lambda is initiated to start a pre-processing job in AWS Glue using the uploaded processing plugin script to transform battery health data into the desired format.
Step 4
Processed datasets get imported into Amazon Forecast, which uses AutoML to automatically choose the most optimal data processing and ML algorithm to build. It then trains and makes predictions on battery state of health (SoH) and remaining useful life (RUL).
Step 5
Predictions made using Forecast are exported into Amazon S3 after getting processed using AWS Glue. This data is used to serve data through a web interface for Original Equipment Manufacturer (OEM) and the vehicle dashboard for connected vehicles.
Step 6
Exported forecasts in Amazon S3 are also used to monitor drift. You can set a drift threshold, for example 10%, above which the forecast pipeline gets triggered to retrain the model.
Step 7
You can view battery health status and predictions through the Connected Mobility Platform.
Step 1
Battery health data is ingested through the Connected Mobility Platform.
Step 2
Streaming data ingested through the Connected Mobility Platform is stored in Amazon Timestream for near real-time monitoring. Amazon Simple Storage Service (Amazon S3) is used to store historical battery data.
Amazon DynamoDB stores the state of each phase of the machine learning (ML) pipeline, including pre-processing, model generation and prediction, post-processing, and drift monitoring.
Step 3
When new data becomes available, AWS Lambda is initiated to start a pre-processing job in AWS Glue using the uploaded processing plugin script to transform battery health data into the desired format.
Step 4
Processed datasets get imported into Amazon Forecast, which uses AutoML to automatically choose the most optimal data processing and ML algorithm to build. It then trains and makes predictions on battery state of health (SoH) and remaining useful life (RUL).
Step 5
Predictions made using Forecast are exported into Amazon S3 after getting processed using AWS Glue. This data is used to serve data through a web interface for Original Equipment Manufacturer (OEM) and the vehicle dashboard for connected vehicles.
Step 6
Exported forecasts in Amazon S3 are also used to monitor drift. You can set a drift threshold, for example 10%, above which the forecast pipeline gets triggered to retrain the model.
Step 7
You can view battery health status and predictions through the Connected Mobility Platform.
Well-Architected Pillars
The AWS Well-Architected Framework helps you understand the pros and cons of the decisions you make when building systems in the cloud. The six pillars of the Framework allow you to learn architectural best practices for designing and operating reliable, secure, efficient, cost-effective, and sustainable systems. Using the AWS Well-Architected Tool, available at no charge in the AWS Management Console, you can review your workloads against these best practices by answering a set of questions for each pillar.
The architecture diagram above is an example of a Solution created with Well-Architected best practices in mind. To be fully Well-Architected, you should follow as many Well-Architected best practices as possible.
-
Operational ExcellenceRead the Operational Excellence whitepaper
Amazon CloudWatch provides centralized logging with metrics and alarms to raise alerts for operational anomalies. CloudWatch is also used to monitor model drift. You may want to establish a drift threshold, above which the ml pipeline will be triggered to retrain the model. This helps continuously improve battery health predictions.
-
SecurityRead the Security whitepaper
Forecast uses TLS with AWS certificates to encrypt any data sent to other AWS services. Forecast endpoints support only secure connections over HTTPS.
Data at-rest is encrypted using server-side encryption with Amazon S3 managed keys (SSE-S3).
-
ReliabilityRead the Reliability whitepaper
This Guidance follows an event-driven architecture with loosely coupled services, making it easy to isolate behaviors and therefore increase resilience and agility. It uses managed serverless services, such as Amazon EventBridge and Lambda, to communicate between loosely coupled services.
-
Performance EfficiencyRead the Performance Efficiency whitepaper
Services selected for this Guidance are all purpose-built. For example, Forecast is a purpose-built time series forecasting service based on machine learning. EventBridge is a purpose-built serverless service to connect loosely coupled components using events.
-
Cost OptimizationRead the Cost Optimization whitepaper
This Guidance follows an event-driven architecture with AWS services that are fully managed, such as Lambda, AWS Glue, and Amazon S3. These services autoscale according to the workload demand. As a result, you only pay for what you use.
-
SustainabilityRead the Sustainability whitepaper
The Amazon S3 Intelligent-Tiering storage class is designed to automatically move data to the most sustainable access tier in Amazon S3.
Implementation Resources
The sample code is a starting point. It is industry validated, prescriptive but not definitive, and a peek under the hood to help you begin.
Disclaimer
The sample code; software libraries; command line tools; proofs of concept; templates; or other related technology (including any of the foregoing that are provided by our personnel) is provided to you as AWS Content under the AWS Customer Agreement, or the relevant written agreement between you and AWS (whichever applies). You should not use this AWS Content in your production accounts, or on production or other critical data. You are responsible for testing, securing, and optimizing the AWS Content, such as sample code, as appropriate for production grade use based on your specific quality control practices and standards. Deploying AWS Content may incur AWS charges for creating or using AWS chargeable resources, such as running Amazon EC2 instances or using Amazon S3 storage.
References to third-party services or organizations in this Guidance do not imply an endorsement, sponsorship, or affiliation between Amazon or AWS and the third party. Guidance from AWS is a technical starting point, and you can customize your integration with third-party services when you deploy the architecture.