Creating an AWS-Based End-to-End Cloud Solution using the STM32U5 IoT Sensor Node

Introduction

In this tutorial we will:

• create a rule in AWS IoT Core to send sensors' data from MQTT topic to Amazon Timestream
• observe the data in Amazon Timestream database (which is AWS specialized serverless database for time-series data)
• log in to Grafana workspace and configure the dashboard
• visualize the data from STM32U5 IoT Sensor Node environmental sensors (temperature, humidity, and barometric pressure) in Grafana

Every participant will have individual AWS account to experiment with. Accounts are provisioned using Workshop Studio and will be terminated after the workshop. A number of components were pre-provisioned in the accounts already using AWS CloudFormation, namely:

• Amazon Timestream database and table
• AWS Fargate running Docker container with Grafana application
• Amazon Cloudfront CDN distribution with TLS certificate to protect Grafana internet facing dashboard.
• Number of AWS Identity and Access Management policies and roles, governing how the services can interact with each other.

If you want to recreate workshop in your own AWS account, please download sensor_data_ingestion.yaml and iot_policy.yaml CloudFormation templetes to create those stacks.

Log in to Workshop Studio AWS account

Let's log in to the workshop account using One Time Password (OTP) method from the workshop landing page

1. Select Email One-time Password

2. Enter your email and press Send passcode

3. Check your email and enter the 9-digit value in the prompt and press Sign in

4. Tick the checkbox accepting the Terms and conditions and press Join Event

5. At the left bottom panel "AWS Account access" select Open AWS Console (us-east-1)

MQTT topics structure and payload format

In the STM32U5 AWS QuickConnect workshop, you should already know the deviceID. If not open the serial console to STM32U5 and reset the board pressing the black RST button. Look for the string similar to:

<INF>      790 [MQTTAgent ] Client Certificate: CN=stm32u5-a61b1620323443M0, SN:0x5A7C3C321EDBC7F9

The CN value is the first level MQTT topic that your device is publishing on. Environmental sensors are published to:

stm32u5-XXXXXXXXXXXXXXXX/env_sensor_data

In my case it is stm32u5-8b74032038373303/env_sensor_data

The payload for environmental sensors is a structured JSON:

{
  "temp_0_c": 34.11742,
  "rh_pct": 30.602741,
  "temp_1_c": 34.150002,
  "baro_mbar": 1000.073242
}

It is important as it allows rule engine to do transformation and enrichment of data on the fly.

Let's verify that our STM32U5 IoT Discovery Node sends messages to AWS IoT Core.

1. In the AWS Console navigate to Services $\to$ Internet of Things $\to$ IoT Core.

2. Navigate to MQTT Test Client. Subscribe to all MQTT topics by entering # symbol in the Subscribe field and click Subscribe

3. Observe messages. Click on pause, copy and save your MQTT topic. Remember, it has format stm32u5-XXXXXXXXXXXXXXXX/env_sensor_data. We will need that value later.

Now that we know our topic structure and payload, lets publish that data to Amazon Timestream database.

AWS IoT Core Rules

AWS IoT rules give your devices an ability to send the data from MQTT topics to a variety of AWS services, such as S3, DynamoDB, Lambda, OpenSearch, Timestream, etc. Rules also can perform data filtering and transformation. To learm more refer to official documentation.

1. In the AWS Console navigate to Services $\to$ Internet of Things $\to$ IoT Core. On that screen from the left-hand panel chose Message Routing $\to$ Rules

2. Create the rule with the name MQTTtoTimestream and description Send data from STM32U5 environmental sensors to Amazon Timestream timeseries database. Click Next

3. Enter the SQL statement to get temperature and humidity reading from environmental sensors payload and click Next

SELECT temp_0_c AS temperature, rh_pct AS humidity, baro_mbar AS pressure FROM 'stm32u5-XXXXXXXXXXXXXXXX/env_sensor_data'

4. Attach the action. Choose Timestream table as the Action, select SensorData database and stm32u5 as a table. Configure the dimension with the dimension name name and dimension value stm32u5. For IAM role, search for sensor-data-ingestion-TimestreamInsertRole-XXXXXXXXXXXX role.

5. Inspect the input and click Create

Amazon Timestream

Let's verify that the rule is sending the data to the database

1. In the AWS Console navigate to Services $\to$ Database $\to$ Amazon Timestream. Or Search for "Timestream"

2. Expand the left-side navigation panel by clicking on $\equiv$ symbol

3. Navigate to Management Tools $\to$ Query Editor. Find stm32u5 table and click on the triple vertical dots on the right-hand side. Click on Preview data

4. Run the query

5. Inspect the result

Amazon Grafana

1. In AWS Console navigate to Services $\to$ Management&Governance $\to$ CloudFormation
2. Click on sensor-data-ingestion stack and then inspect Outputs tab. There should be Grafana URL filed. Click on the URL that should look similar to http://dxxxxxxxxxxxx.cloudfront.net/
3. In the Grafana login screen use workshop as username and stm32u5 as a password

4. Dashboard should start displaying the values from Amazon Timestream stm32u5 table. If values are not displayed, please click Edit on correspondent widget and verify that the query and the measurement fields are correct. The dashboard should look like this.

5. Please note that temperature on the sensor can be significantly higher than ambient due to the proximity to WiFi module that gets warm quickly
6. Breathe at the board and observe the humidity graph changes

This concludes the workshop