Environmental Monitoring with LTPS, Azure, and PowerBI

About the Application

This tutorial shows how to collect the environmental data — humidity, ambient pressure, temperature, and ambient lighting level — from multiple LTPS devices using Azure IoT Hub, process this data with Azure Stream Analytics, and visualize it in real time with Power BI.

What you need:

Hardware

• 1 x Linux TPS3 (LTPS3) board, optionally in a TPB3 enclosure
• 1 x Tibbit #10 (power supply)
• 1 x Tibbit #18 (power jack)
• 3 x Tibbit #00_3 (two direct I/O lines, +5V power, ground)
• 1 x Tibbit #30 (ambient humidity/temperature meter)
• 1 x Tibbit #35 (barometric pressure sensor)
• 1 x Tibbit #28 (ambient light sensor)

Onboard Software

• Node.js V6.x.x (pre-installed during production)

External Services

• Microsoft Azure subscription
• Microsoft Power BI account

Prerequisites

• Define the required configuration in the LTPS Web Interface
• Login to the LTPP3 board from the SSH client
• Install NPM and other tools as needed

Install the App

First, do this:

# Create a folder for the application
cd /opt/node-apps/
mkdir environment
cd environment

# Install the required modules
npm install @tibbo-tps/tibbit-28
npm install @tibbo-tps/tibbit-30
npm install @tibbo-tps/tibbit-35

Create azure-app-0.js file with your application:

var tibbit28 = require("@tibbo-tps/tibbit-28").init("S11");
var tibbit35 = require("@tibbo-tps/tibbit-35").init("S13");
var tibbit30 = require("@tibbo-tps/tibbit-30").init("S15");

setInterval(function(){
    var illuminationData = tibbit28.getData();
    var humidityData = tibbit30.getData();
    var pressureData = tibbit35.getData();
    var dateTime = new Date();

    console.log("Date/Time: "+dateTime);
    console.log("Illumination: "+illuminationData.illumination);
    console.log("Humidity: "+humidityData.humidity);
    console.log("Temperature: "+humidityData.temperature);
    console.log("Pressure: "+pressureData.pressure);
},1000);

Upload it to the /environment folder and run:

$ node azure-app-0.js

Date/Time: Tue Jul 19 2016 13:50:54 GMT+0000 (UTC)
Illumination: 115
Humidity: 43.755340576171875
Temperature: 28.05670928955078
Pressure: 738.7398681640625

Set Up Microsoft Azure IoT Hub

Azure IoT Hub is a service that allows bi-directional communications between your devices and a "solution back end." To continue with the tutorial, create an IoT hub.

• Create an account on the Azure Portal
• Sign in to the Azure Portal
• Choose New > Internet of Things > Azure IoT Hub
• Configure your IoT Hub and click Create.

• After the IoT hub is ready, select All Resources > and click the Keys icon.
• In the Shared access policies blade select the iothubowner item and copy the value of the Connection string field.

Register Device Identity

There are three ways to register device identities: with a script using Azure API, with GUI App (Windows only), and with a multiplatform CLI tool. The third way seems to be the most convenient one.

Note: To use iothub-explorer you will need Node.js V4.x or higher installed on your PC

• Run the following on your PC:

npm install -g iothub-explorer@latest

• Now this:

$ iothub-explorer login <iothubowner-connection-string>

Session started, expires on Thu Jul 21 2016 13:22:10 GMT+0400

where <iothubowner-connection-string> is an iothubowner connection string from the previous step.

• Register a new device and get its connection string:

Replace <device-name> with the actual name of your LPTS, for example, "tps-centreville"

$ iothub-explorer create <device-name> --connection-string

Created device tps-centreville

...

connectionString: HostName=iot-tps.azure-devices.net;DeviceId=tps-centreville;SharedAccessKey=fSCVQIY..TOprSsDE=

• Save the value of the connectionString — you will need it later.

IoT-enable your LTPS

• Login to the LTPP3 board from the SSH client
• Change the working folder to /environment and do the following:

npm install azure-iot-device
npm install azure-iot-device-amqp

• Create azure-app-1.js file with the following app:

var clientFromConnectionString = require('azure-iot-device-amqp').clientFromConnectionString;
var Message = require('azure-iot-device').Message;

var connectionString = '<THE DEVICE CONNECTION STRING FROM THE PREVIOUS STEP>';
var client = clientFromConnectionString(connectionString);

var tibbit28 = require("@tibbo-tps/tibbit-28").init("S11");
var tibbit35 = require("@tibbo-tps/tibbit-35").init("S13");
var tibbit30 = require("@tibbo-tps/tibbit-30").init("S15");

client.open(function(err){
    if(err){
        console.log('Could not connect: ' + err);
    }else{
        console.log('Client connected');

        setInterval(function(){
            var illuminationData = tibbit28.getData();
            var humidityData = tibbit30.getData();
            var pressureData = tibbit35.getData();

            var time = new Date().toISOString();

            var data = JSON.stringify({
                deviceId: 'tps-centreville',
                humidity: humidityData.humidity,
                temperature: humidityData.temperature,
                pressure: pressureData.pressure,
                illumination: illuminationData.illumination,
                time: time
            });

            var message = new Message(data);

            client.sendEvent(message, function (err) {
                if(err){
                    console.log(err.toString());
                }else{
                    console.log("Message sent: " + message.getData());
                }
            });
        },60000)
    }
});

• Upload the application to /opt/node-apps/environment folder onto the LTPS and run it:

$ node azure-app-1.js

Client connected
Message sent: {"deviceId":"tps-centreville","humidity":37.1016960144043,"temperature":31.370407104492188,"pressure":742.8632202148438,"illumination":136,"time":"2016-07-21T10:19:07.490Z"}
Message sent: {"deviceId":"tps-centreville","humidity":37.1016960144043,"temperature":31.370407104492188,"pressure":743.2034301757812,"illumination":137,"time":"2016-07-21T10:20:10.582Z"}
Message sent: {"deviceId":"tps-centreville","humidity":37.1016960144043,"temperature":31.380477905273438,"pressure":743.2034301757812,"illumination":138,"time":"2016-07-21T10:21:12.003Z"}

The most impressive and useful debugging feature offered by iothub-explorer is event monitoring.
With iot-explorer you can monitor events sent by your devices to the cloud and from the cloud to the devices.

• Execute the following:

This command requires you to provide the iothubowner connection string even if you're already logged in.

$ iothub-explorer <iothubowner-connection-string> monitor-events tps-centreville

Monitoring events from device tps-centreville

Event received:
{ deviceId: 'tps-centreville',
  humidity: 37.1016960144043,
  temperature: 31.380477905273438,
  pressure: 743.2034301757812,
  illumination: 138,
  time: '2016-07-21T10:21:12.003Z' }

Configure Stream Analytics

To complete this part of the tutorial, you'll need an active Microsoft Power BI subscription.

Before the information can be delivered to Power BI, it must be processed by the Azure Stream Analytics job.

• Choose New > Internet of Things > Stream Analytics Job
• Configure the Job and click Create
• Wait a couple of minutes until the Job is created, then select All resources > name of your Stream Analytics Job
• Click Inputs > Add

Configure the input:

• Input alias: data-from-tps
• Source Type: Data stream
• Source: IoT Hub
• Subscription: Use IoT hub from current subscription
• Endpoint: Messaging
• Shared access policy name: iothubowner
• Consumer group: powerbi
• Click Create

Configure the output:

• Output alias: data-to-powerbi
• Click Outputs > Add
• Choose Power Bi as sink
• Click Authorize and enter your Power BI credentials in the window
• Enter, respectively, Dataset Name and Table Name, e.g. telemetry
• Click Create

Now it's time to enter the query:

• Click Query
• Enter the following:

SELECT
    AVG(humidity) AS humidity,
    AVG(temperature) AS temperature,
    AVG(pressure) AS pressure,
    AVG(illumination) AS illumination,
    System.Timestamp AS time,
    IoTHub.ConnectionDeviceId AS deviceId
INTO
    [data-to-powerbi]
FROM
    [data-from-tps] TIMESTAMP by time
GROUP BY
    TumblingWindow(Second, 60), IoTHub.ConnectionDeviceId

• Click Save

Stream Analytics Query Language is a subset of SQL. The complete language documentation can be found here. There is also a very useful set of examples.

The query above splits the timeline into 60-second slices and returns average values of humidity, temperature, pressure, and light level for each slice and each deviceId.

• Start your job

Visualize Your Data

• Open your Power BI workspace.
• After the Stream Analytics job starts, a new dataset will appear in the Navigation Bar.
• Create a report. Use time as the axis, deviceId as the legend, and temperature/humidity/illumination/pressure as the value.

Add devices

• If you have another LTPS device, register it as described in Register Device Identity.
• Upload and run the code from IoT-enable your LTPS, use the modified connectionString value.
• After that, the second curve appears in the report.