Technology Architecture for Building Smart Meters using AWS IoT Services

In this architecture, the first thing is the connectivity of the smart meters with the cloud. Here we are using AWS IoT core as the IoT hub, which is connecting to each device. AWS IoT core is sending results to the AWS Kinesis Firehose. After this, the architecture is divided into four parts: Alerting and Notification, Real-time Results, Analytical dashboard, Predictive results on the dashboard.

In alerting part, Kinesis Analytics is consuming stream from Kinesis Firehose and doing the transformation on the streamed data. After these transformations, AWS Lambda Function is used for further data validation and filtering to identify an anomaly. Based on these anomalies, the message has been sent to AWS SNS queue. AWS SQS can also be used at the place of AWS SNS.

While integrating real-time data with the dashboard, we are using the same transformed result which is being used by AWS Lambda. This real-time transformed stream is sent to AWS Kinesis Stream. After that, it will be sent to Web Socket Server, which can be further integrated with the dashboard for the real-time visualization.

For the Analytical Dashboard, real-time firehose stream is writing data to AWS S3 bucket with is raw data zone for the IoT events. In the next step, an AWS Glue ETL job is defined to collect the data from the data zone and putting it into the trusted data zone after doing all the transformation. On the top of AWS S3, we are using AWS Athena for interactive SQL queries on AWS S3 bucket. This a complete serverless service by AWS which can be used to do analytical queries on the AWS S3. This can be integrated directly with the dashboard.

Now the last and most important part is running the Machine learning model on the incoming results and getting predictive results. At this stage, AWS Sagemaker is using trusted data lake for the model training. And for the model inference, AWS Fargate is being used, which is calculating the prediction results and storing it into AWS DynamoDB. AWS DynamoDB is integrated with the dashboard to show the predictive results.

Benefits of Enabling AWS IoT based Smart Meters

  • Estimation of energy bills
  • More control over the energy usage