Build Your Own Personal Voice Assistant Part 1

In the last few years, personal voice assistants like the Amazon Alexa and Google Home have become commonplace in our humble home. Many would say that the personal voice assistant is the first essential when it comes to a smart home. In this tutorial, by Yap Zi Qi, we teach you how to make your own with a Raspberry Pi and Google Assistant!

Hardware and software you need and where to find them:

Retailers are provided for reference only. GovTech TechNews is not affiliated with any of the retailers listed in this tutorial.

What you need

1. Computer with Windows/Mac OS to flash the Raspbian OS onto the SD card

2. SD card slot/reader

3. USB mouse & keyboard

4. HDMI-compatible Monitor & HDMI cable

Step 1: Flash the SD card

1. Download the Raspbian Image.Raspbian is the OS for the Raspberry Pi.

2. Download and install Etcher. Etcher cross-platform tool for flashing images to SD cards & USB drives.

3. Insert micro SD card to your computer

4. Open Etcher and select the downloaded Raspbian Image (either .img or .zip file)

5. Select the micro SD card.

6. Click ‘Flash!’ to begin writing Raspbian to the micro SD card

Step 2: Boot Raspbian for the first time

1. Login with the following credentials

    User: pi
    Password: raspberry 
   

   (You can change the password by running Terminal and typing sudo raspi-config and selecting the Change Password option)

   Testing the microphone and speaker

2. Plug in the microphone and speaker

3. Open Terminal and run the following commands

    arecord -l
    aplay -l
   

   arecord -l displays the list of input hardware devices

   aplay -l displays the list of output hardware devices

4. Note down the card and device number of the microphone and speaker

   

   In the above screenshot, the recording device is USB PnP Sound Device and the playback is bcm2835 ALSA. So, the card no. and device no. for recording device is 1 and 0 respectively, and for playback it is 0 and 0.

5. Create a new file by typing the following command in Terminal

    sudo nano .asoundrc
   

6. Copy the following code into the file, and adjust the card and device number according to what card and device numbers you have recorded down in step 3.

    pcm.!default {
        type asym
        capture.pcm "mic"
        playback.pcm "speaker"
    }
    pcm.mic {
        type plug
        slave {
        pcm "hw:1,0"	# "hw:card,device (for recording)"
        }
    }
    pcm.speaker {
        type plug
        slave {
        pcm "hw:0,0"	# "hw:card,device (for playback)"
        }
   

   

7. To test the audio setup, run the following code in Terminal

    arecord --duration=5 test.wav
   

   The raspberry pi will record audio through the microphone for 5 seconds. Then, run:

    aplay test.wav
   

   to listen to the audio recorded.

   

   And there you have it, your microphone and speaker are working and all that’s left is to set up the Google Assistant and link up the WiFi plug and lightbulb.

   Be sure to look out for part 2 coming next week where you learn to light up your room with a simple voice command!