# Tellus Consulting Bob IOT project

Setup

I had a lot of problems getting a good setup of the software on my Raspberry Pi, including NodeJS. These instructions are a conglomerations from various sources. Note that my intent is to access my Raspberry Pi remotely, so I have chosen not to have a desktop UI and assume all access is through SSH. Setup is based partially on the instructions from Codefoster's IOT Workshop instructions.

Hardware

• Raspberry Pi 3 B
• 4 GB SD Card
• Network cable
• Micro USB cable (for power)

Software

• Raspbian Jessie Lite - Version June 2017 - Kernel 4.9 - Download from Raspberrypi.org/downloads
• Etcher - Download from etcher.io

Core software installation

These instructions will install an operating systema and Node.js.

1. Download Raspbian Jessie Lite from Raspberrypi.org/downloads

2. Install Etcher from etcher.io

3. Insert the SD card in your local computer

4. Install the Raspbian Jessie Lite image you downloaded to the SD card, using Etcher:

   1. Start Etcher.
   2. Click Select image and navigate to and select the zip file downloaded from Raspberrypi.org.
   3. Click Select drive and select the SD card.
   4. Click Flash! and wait for the image to install.
   5. After the installation is complete, close Etcher.

5. Eject and re-insert the SD card.

6. Create a file named SSH (no extension) on the root of the SD Card.

7. Make sure your Raspberry Pi will successfully boot in headless mode.

   This is based on an article here: Is your Pi not booting? Look under "Here is a list of known causes why the Pi might not be booting, or might appear not to be booting:" and then "Video".

   1. From your root of your SD card, open the file config.txt.
   2. If necessary, format the text file (for me, everything appeared as one line. There should be one line for each #)
   3. Uncomment (remove the #) the following lines:

      # hdmi_safe=1
      ...
      # hdmi_force_hotplug=1

8. Eject the SD card and insert it in the Raspberry Pi.

9. Plug in power cable to the Raspberry Pi.

10. Plug in network cable to Raspberry Pi (plug in other end to your computer or router).

11. Log in to your raspberry pi

    1. Open a command window that supports SSH.

       I use Windows PowerShell, and the Git Shell's shell.ps1 configuration file, called from my profile at startup.

    2. From your command shell, type the following command and press ENTER:

       ssh pi@raspberrypi.local

    3. When prompted, enter the password raspberry

12. Verify wi-fi tools are installed.

    Based on instructions from https://www.maketecheasier.com/setup-wifi-on-raspberry-pi/

    • Run the command

      sudo apt-get install wpasupplicant wireless-tools

13. Update general wi-fi configurations. Run the command:

    sudo nano /etc/network/interfaces

14. Update the section that starts with allow-hotplug wlan0 to match the following:

    allow-hotplug wlan0
    iface wlan0 inet manual
        wpa-conf /etc/wpa_supplicant/wpa_supplicant.conf
    iface default inet dhcp

15. Press CTRL+X to exit nano. Press Y and press ENTER to SAVE the file, and press ENTER again to save it.

16. Get a list of available wireless networks. Run the command:

    sudo iwlist wlan0 scan | grep ESSID

    grep ESSID returns just the lines containing ESSID, which show the network names.

17. Update the wpa_supplicant configuration file. Run the command:

    sudo nano /etc/wpa_supplicant/wpa_supplicant.conf

18. Add the following to the file:

    network={
        ssid="<Wireless Network Name>"
        psk="<network password>"
        key_mgmt=WPA-PSK
        }

    • Replace the text <Wireless Network Name> with the name of your network.
    • Replace the text <network passowrd> with your network password.
    • You may need to use a different key_mgmt setting based on your network.

19. Exit Nano and save the file.

20. Update other Raspberry Configration options using the **Raspberry Pi Software Configuration Tool (raspi-config). Run the command:

    sudo raspi-config

21. Change the user password on your device

    1. In raspi-config select 1 Change User Password.
    2. Select Ok.
    3. In the shell, type your new password and press Enter.
    4. Confirm your password and press Enter.
    5. In raspi-config select Ok.

22. Change the host name for your device (from raspberrypi)

    1. In raspi-config select 2 Hostname.
    2. Select Ok.
    3. Type your new host name and press Enter.

23. Change the locale options

    1. In raspi-config select 4 Localisation Options.

    2. Select I1 Change Loocale.

    3. On the Configuring locales page, use the arrow keys to scroll through the list, and SPACEBAR to select your language and locale setting.

       I chose en_US.UTF-8

    4. Press ENTER and on the next page, select the language you just chose and press ENTER.

    5. Update the Timezone and Wi-fi Country as needed from the same Localisation Options menu.

24. Enable I2C and Camera interfaces on Raspberry Pi

    This article provides a decent article that explains the interfaces:Raspberry Pi SPI and I2C Tutorial. If not enabled, you will receive an error later in NodeJS.

    1. In raspi-config select 5 Interfacing Options.
    2. Select P1 Camera, select Yes and select OK.
    3. In raspi-config select 5 Interfacing Options again.
    4. Select P5 I2C, select Yes and select OK.
    5. When finished, in raspi-config select Finish and press ENTER.
    6. Select Yes to reboot your Raspberry Pi

25. Log in to your Raspberry Pi, again, using the new host name (pi@hostname.local)

26. Install Node.js on the device

    1. Check Nodejs.org to verify what the current or latest version is.

       At the time of this instruction, the Long Term Support (LTS) version was 6.11.0 and current 8.1.3. I used LTS.

    2. In bash, run the following commands:

       cUrl -sL http://deb.nodesource.com/Setup_6.x | sudo -E bash
       sudo  apt-get -y install nodejs

Build your circuit

The Bob-IOT project uses a very similar setup to that used by @CodeFoster in his project. @CodeFoster's project is based on a kit provided in the workshop. Mine is based on this commercially available electronics kit from elegoo.com: Elegoo Upgraded Electronics Fun Kit While the components come from a different source, the setup of the circuit board is the same. The basic instructions are:

1. Install the LED into the board, providing plenty of room to both left and right sides, with longest leg to the left.
2. Install a 10K ohm resistor, with one end in a column matching the longest leg of the LED and the other end connected in a column to the left of the led.
3. Use a jumper wire to connect a column matching the shorter leg of the led to a colum to the right.
4. Attach a 330 ohm resist to a column matching the right end of the jumper wire, and attach it to a column to its right.
5. Attach a button to the board, with one side connected to the column matching the right end of the 330 ohm resistor.
6. Use a female-to-male jumper wire, conect GPIO Pin 26 to a column matching the left end of the 10K ohm resistor.
7. Use a female-to-male jumper wire, connect GPIO 20 to a column matching the left side of the button (and right end of the 330 ohm resistor).
8. Use a female-to-male jumper wire, connect a 3.3v (PIN 1) pin to a column matching the right side of the button.
9. Use a female-to-mail jumper wire, connect a ground (PIN 6) to the column that matches the left side of the 330 ohm resistor (and the right end of the jumper that connects to the right side of the LED).

Completed breadboard and Raspberry Pi

Close-up of button and 330 ohm resistor

Close-up of LED and 10K ohm resistor

You can use the topic, GPIO: MODELS A+, B+, RASPBERRY PI 2 B AND RASPBERRY PI 3 B as a reference to the GPIO pins. Note that the power and ground pins are not numbered. I've used the physical pin numbering (see the above topic for explanation) to specificy them.

Set up your development environment

These instructions will setup your development environment on your host computer and in Microsoft Azure. If you have your local development environment set up already, you can skip some of this.

1. Set up Visual Studio Code. Why VS Code? Because its free and easy.

   1. Navigate to https://code.visualstudio.com/
   2. Click Download for Windows (or whatever OS you use)
   3. Run the installer and follow the prompts as appropriate.

2. Sign up for an Azure account. You can do a free trial or simply sign up for pay-as-you-go.

   There are several services you can use for free, as long as your usage is low.

   • Navigate to Azure.com and click Free Account -or- Navigate to portal.azure.com

3. Create an Azure IoT Hub

   1. In the Azure portal (portal.azure.com), click + New.
   2. Click Internet of Things and then click IoT Hub.
   3. Enter a Name for your IoT Hub.
   4. Select a Pricing and scale tier.

      The F1 tier is free.

   5. Select the number of units.

      1 is sufficient.

   6. Under Resource group select Create new and enter a name for the resource group.
   7. Select the Location that is closest to you.
   8. Click Create.

4. Get the connection string for your IoT Hub.

   1. In the Azure portal, open your IoT Hub.

      The setup of your dashboard will affect how you find your IoT Hub. If all else fails, click All resources in the left-hand menu, and select your IoT Hub.

   2. In the IoT Hub blade, select Shared access policies.
   3. Select iothubowner.
   4. Copy the Connection string-primary key.

5. Install IoT Hub Explorer

   • On your local computer, in the command shell, run the following command:

     npm install -g iothub-explorer

6. Add your device to IoT Hub

   1. In the command shell, run the following command:

      iothub-explorer login '<connection string>'

      Replacing <connection string> with the connection string you copied from your IoT Hub.

   2. Run the following command:

      iothub-explorer create '<device id>'

      Replace <device id> with a unique ID for your device

7. Install Node.js on your local machine.

   1. Navigate to nodejs.org.
   2. Click the link for the version of Node.js you want to install.
   3. Run the installer and follow the prompts as appropriate.

8. Create and initialize your local development folder.

   1. On your local machine, in the command shell, navigate to the location where you want to create your development folder.

   2. In the command shell, run the following command:

      new-item device -Type directory

      Keep in mind, I'm using Windows PowerShell.

   3. Navigate to the device folder.

   4. Initialize the folder by running the following command:

      npm init -y

9. Install the code dependency libraries. They are:

   • johnny-five - Libraries that allow you to talk to the Raspberry Pi easily
   • azure-iot-device - Libraries for Azure IoT SDK
   • azure-iot-device-amqp - AMQP transport for Azure IoT SDK

   Run the following command: posh npm install johnny-five azure-iot-device azure-iot-device-amqp

   There will likely be some errors from the johnny-five install.

Create and deploy the project

1. Create your default js file index.js.

   1. Open Visual Studio Code.

   2. Click File and then click Open Folder

   3. Navigate to the device development folder and click Select folder.

   4. Click New file and enter the file name index.js.

   5. Enter the following code into the file:

      // index.js
      //'use strict';
      
      //Load dependencies
      var five = require('johnny-five');
      var raspi = require ('raspi-io');
      var device = require('azure-iot-device');
      var deviceAmqp = require('azure-iot-device-amqp');
      
      //INIT
      let connectionString = '<connection string>';
      let hubClient = deviceAmqp.clientFromConnectionString(connectionString);
      
      //MAIN CODE
      //establishing connection to gpio
      log('establishing connection to gpio...');
          let board = new five.Board({ io: new raspi() });
      board.on('ready', () => {
          let led = new five.Led('GPIO26');
          let button = new five.Button('GPIO20');
          led.stop().off();
      
          //open connection to iot hub
          log('connecting to iot hub...');
          hubClient.open(err => {
              if (err)
                  log(err.message)
              else {
                  log('READY');
                  led.stop().off();
      
                  button.on('press', () => {
                      led.blink(500);
                      log('sending message to iot hub...');
                      let message = new device.Message(JSON.stringify({ deviceId: '<device id>', tags: ['foo', 'baz', 'bar'] }));
                              hubClient.sendEvent(message, (err, res) => {
                                  if (err) log(err.message);
                                  else {
                                              log(`Sent message to your IoT Hub`);
                                              log('READY');
                                          }
                                          led.stop().off();
                                      });
                                  });
                          }
                      });
                  });
          
      
      function log(msg) {
          console.log(msg);
      }

      Replace <connection string> with your IoT Hub connection string Replace <device id> with your device ID.

2. Update the file package.json to include additional dependencies The package.json file contains the list of dependencies that will need installed.

   1. In Visual Studio Code, open package.json.
   2. Add the following entries at the end of the dependencies section, before the closing bracket:

      "raspi-io": "^8.0.1",
      "camerapi": "^0.1.0"

      Be sure to add a comma after the previous entry.

   Note: The camerapi library is used for a Raspberry Pi camera, as the name suggests. We won't be using it in this project, but we'll load it anyway. Also, update the version to the latest. The easiest way to do this is remove the version information provided and let VS Code show you the current version.

3. Save the files in the project.

4. Deploy the project files to your device.

   1. Create the project folder on your device. Run the command:

      ssh <username>@<devicename>.local 'mkdir device'

   2. Enter your password.

      This command runs mkdir device on the device, without logging in and then logging out.

   3. Copy the project files to the device. Run the command:

      scp index.js package.json username>@<devicename>.local:~/device

5. Install the dependencies on the device.

   1. From the command shell, log in to your device.
   2. Navigate to the device folder on your device.
   3. Run the following command:

      npm install

6. Run the application using the following command from within the device folder:

   sudo nodejs .

7. You should see a message in your shell window something like this:

   1499137143153		RaspberryPi-IO
   1499138143653		RaspberryPi-IO
   1499137143675	Initialized
   >> connecting to iot hub...
   READY

8. Press the button on the breadboard to and you should see the following message in the shell:

   sending message to iot hub...
   Sent message to your IoT Hub
   READY

   In addition to the message, the led on your breadboard should blink 3 times.

9. Type the following command to exit NodeJS:

   .exit

You can also check your IoT Hub in Azure to see how many messages you have received. You'll need to do some work in Azure to actually do something with those messages.