Getting Started Guide for the Intel® Speech Enabling Developer Kit

Designed to enable commercial Smart Home device manufacturers to create new experiences with speech recognition capabilities using Amazon Alexa* Voice Service on Intel® silicon, the Intel® Speech Enabling Developer Kit can be used to create a diverse ecosystem of Smart Home devices with Alexa. Some categories of these devices include: wireless speakers, entertainment devices, and smart appliances. The kit includes a dual DSP with inference engine, an 8-mic circular array, and technology for Alexa wake word recognition, beam forming, noise reduction, and acoustic echo cancellation.

Visit the Intel® Speech Enabling Developer Kit website to learn more.

Follow along with the chapters of this getting started guide to get your developer kit up and running. We walk you through hardware assembly and set up as well as firmware and software setup. You will be able to set up a Raspberry Pi* board as a speech analytic engine host system running Raspbian* operating system (OS).

There are two phases to set up the Intel® Speech Enabling Developer Kit:

1. Hardware setup
2. Operating system setup and software downloads

If you have already completed the hardware setup instructions in the box, you can skip to Step 4: Connect the Raspberry Pi cable.

Phase 1: Hardware Setup

The hardware components in the Intel® Speech Enabling Developer Kit includes:

• Dual DSP with inference engine
• Eight digital microphone (DMIC) board
• Raspberry Pi* connector cable
• 6mm screws (x3)
• Washers (x6)
• 40mm female-female standoffs (x3)

Additional hardware needed, but not included in the kit:

• Raspberry Pi* 3 board and power supply (Try this set: Raspberry Pi 3 with 2.5A Micro USB Power Supply)
• Ethernet cable
• HDMI* cable
• HDMI compatible monitor
• USB keyboard and mouse
• External speaker with 3.5mm connector, such as the M-Audio AV32.
• Micro USB power supply (min. 5V, 1A) to power DSP board (Try this one: Micro USB Charger Power Adapter)
• MicroSD* card (16GB or larger)

You can also purchase a Raspberry Pi starter kit, like the Vilros Raspberry Pi 3 Kit, which contains a board, case, power supply and heat sink.

Contents of Kit

Note: The Raspberry Pi Connector cable may be shorter than the one pictured.

---

Assembly

Step 1: Screw and Washer Placement

Start with the DSP board with bottom side facing up. Place one washer on the first screw and insert in the screw slot. Then place another washer on the screw. Attach one of the female-female standoffs to the screw. Repeat this process with the remaining screws, washers and standoffs.

Note: Be careful not to overtighten the screws.

Step 2: Invert Setup for DMIC Board Placement

Turn the setup over so that it is standing on the 40mm female-female standoffs. Place the DMIC board into the connectors on the DSP board.

Note: Ensure that Intel's logos face the same direction on the DSP board and DMIC board and that the three large holes are aligned. All pins should line up and boards should be directly over one another. If alignment is off – even by one pin – the kit will not work.

Step 3: Raspberry Pi Connector

Invert the setup and insert the Raspberry Pi connector cable on the board.

Step 4: Connect Cable to Raspberry Pi

Step 5: Connect Power to DSP Board

Note: The LEDs will turn blue to indicate the power is connected correctly.

Step 6: Connect Speakers

Connect the speakers to the 3.5mm audio jack on the DSP board.

Completed Cable Setup Below

Now connect the USB keyboard and mouse and the HDMI monitor to the Raspberry Pi.
This image shows the correct cable setup for the needed peripherals into the Raspberry Pi ad the DSP board. (Ethernet cable connection not shown.)

Hardware setup key:

1. 5V 1A USB power to DSP board
2. 3.5mm audio jack to speaker from DSP
3. 5V 2.5A USB power for Raspberry Pi
4. HDMI output to monitor
5. USB keyboard
6. USB mouse

Phase 2: Operating System Setup and Software Downloads

Now that you have the hardware set up, you are ready to set up the Raspberry Pi as the host system for the developer kit. To do this, you’ll flash the Raspbian* OS onto the Raspberry Pi board. This is done in two steps:

• First, you will need to download the latest version of Raspbian onto your computer and then program the microSD card, using your laptop or PC.
• Second, you will move the microSD card to the Raspberry Pi and boot the OS.

Raspbian* OS Setup

To download and install the OS and additional required packages, you will need to establish an internet connection to both your laptop or PC and to the Raspberry Pi.

• Make sure your laptop or PC has an internet connection.
• Connect one end of your Ethernet cable to Raspberry Pi’s Ethernet port and the other end of the Ethernet cable to an Ethernet switch/hub or a router that provides a direct internet connection (i.e. no proxy servers to get access to internet).
  1. On your laptop or PC, install the base Raspberry Pi OS image: http://downloads.raspberrypi.org/raspbian/images/raspbian-2017-09-08/2017-09-07-raspbian-stretch.zip. It is important to follow this link to the Raspbian OS because the developer kit has been validated on this version of the OS. For more information on installing the OS images, please visit: https://www.raspberrypi.org/documentation/installation/installing-images/README.md
  2. Load the OS image from the laptop or PC to the micro SD card using a card reader and dedicated software, like Win32DiskImager for Windows*, dd or Etcher for Linux* or Mac. You will need to use a micro SD to SD card adapter or a micro SD to USB adapter depending on what ports are available on the PC/Mac.
  3. Once the image is written to the micro SD card, remove the card from the laptop or PC and Insert the card into the Raspberry Pi. The microSD card slot is located on the bottom of the Raspberry Pi 3 board.
  4. Plug the USB cable connected to a 2.5A wall adapter into the micro USB port on Raspberry Pi and power on. Do not plug the USB cable powering the Raspberry Pi into your laptop or PC.

Change Keyboard Layout

The default set up for the keyboard is UK English in the Raspbian image. If you are using a keyboard with a different layout, like US English, these steps need to be followed to update the keyboard configuration.

1. Select Raspberry Menu in the top left of the desktop.
2. Select Preferences.
3. Select Raspberry Pi Configuration.
4. Select the Localization tab.
5. Select Set Keyboard.
6. A new window will pop-up prompting you for a keyboard layout selection. Change the settings to United States, English (US), or whichever keyboard layout is your preference, and click OK.
7. Click OK on the Raspberry Pi Configuration window.

Security

For security purposes, it is strongly advised that you change the default password for default user pi and set the password for the root user. You can learn more about updating the default password here.

Note: The remaining tasks will be completed on the Raspberry Pi browser. If you can see the Raspberry Pi desktop, you should use it for all remaining steps.

Get an Amazon Developer Account

Create an Amazon developer account here: https://developer.amazon.com/. Click Sign In and select Create your Amazon Developer account.

Create a device profile following these instructions: https://github.com/alexa/alexa-avs-sample-app/wiki/Create-Security-Profile

Note: You might notice slight differences between the user interface and the instructions posted on GitHub. Please continue to follow the instructions as stated on GitHub.

Install and Configure AVS Device SDK

There are two options to do this:

1. Download pre-built kernel, driver, SDK and dependent libraries and install them. This option (known as "use prebuilt") takes about 30 minutes to complete.
2. Download the sources and dependent libraries, build from the kernel, driver and SDK from the sources. This option (known as "from scratch") can take 3 to 4 hours to complete.

TIP: You may find it easier to open the Getting Started Guide on the Raspberry Pi so that you can copy and paste long commands: https://software.intel.com/en-us/articles/getting-started-guide-for-the-intel-speech-enabling-developer-kit.

1. Open a terminal window on the Raspberry Pi.
2. Execute the followings command to download the installation script:

   cd ~
   wget https://raw.githubusercontent.com/intel-iot-devkit/avs-device-sdk-intel-speech-enabling-kit/master/install_avs_sdk.sh

3. Execute one of the followings commands depending on the option you choose for the installation
   Use prebuilt components (30 mins):

   sudo bash ./install_avs_sdk.sh --use-prebuilt

   Build from scratch (3 to 4 hours):

   sudo bash ./install_avs_sdk.sh --from-scratch

4. The script will prompt you for the device credentials. Enter the device credentials you created in your Amazon developer portal. Press the Enter on the keyboard after every entry.
   TIP: Login to the Amazon developer portal using your developer account in a browser on the Raspberry Pi desktop and have the page for device/product you created. You can then copy/paste the credentials from the browser window to the terminal script.
5. Once the credentials are entered, the script will continue with the installation option you chose. Upon completion of installation, the script will launch the Raspberry Pi’s browser with the URL (https://localhost:3000).
6. Press Continue or login to your Amazon Developer account as prompted by the website. Wait for the script to complete.
7. Reboot the Raspberry Pi.

   sudo reboot

8. If you selected the "from scratch" option, you may optionally run unit tests to make sure all the components work correctly. If you did not select the "from scratch" option, you should skip this step.
   In terminal window, run the following steps to execute the series of unit tests on your prototype:

   cd /home/pi/sdk-folder/sdk–build
   sudo make all tests

   You should see and hear your prototype run through a series of audio tests as well as functional tests. These tests should take around 4 minutes and result in 571 tests complete with 100% success. As a developer, any time you modify your client’s on-device SDK software, you should run these Unit Tests to ensure nothing was unintentionally broken.
9. Now you’re ready to launch your client! Start the AVS Sample App with the following commands:

   cd sdk-folder/sdk-build/SampleApp/src/
   TZ=UTC ./SampleApp ../../Integration/AlexaClientSDKConfig.json

10. Now you can interact with the client using voice commands. Try giving it some commands, such as:
    • "Alexa, what time is it?"
    • "Alexa, what is 1+1?"
    • "Alexa, what’s the weather in New York City today?"
    • "Alexa, do you know rap?"

You should now have a working Alexa prototype!

To exit the Sample App, simply press CTRL-C, or enter "q" into the terminal window and press Enter.

If the device is not responding, or if you have any technical questions or issues, please contact Intel Customer Support.

Useful Links

• Visit the Intel Download Center for prebuilt component updates.