Shed light on the elusive mysteries of your pets’ secret home life with the Internet of Things (“IoT”) Bark Back project! This system monitors and uploads the volume of noise in your home to the Cloud so you can check in on your beloved pet(s) while you are away. The best part: if it gets too loud, i.e. Fido is barking or making some other raucous, the program will play back pre-recorded tracks to help distract and quiet down the pup.
This project uses a Raspberry Pi computer to read the SparkFun MEMS microphone breakout board and trigger an audio player. Data is uploaded to the CloudMQTT service using the MQTT communication protocol.
This tutorial will show you how to do the following:
- Connect and read in the SparkFun MEMS Microphone using the Raspberry Pi 3;
- Upload the volume data to the CloudMQTT service; and
- Use a volume threshold to trigger your “bark back” at your pooch if they get too noisy.
To build this project, you’ll need a fully configured, WiFi-connected Raspberry Pi 3 computer with Raspbian OS. It’s also helpful to know some Python programming as well as the following things: (1) how to use and control the Raspberry Pi GPIO pins; (2) MQTT communication; and (3) analog sensors. If any of this is unfamiliar, or if you’re just curious (be curious!), check out the tutorials below!
Raspberry Pi 3
- Raspberry Pi 3 Starter Kit Hookup Guide
- Raspberry Pi GPIO
- SPI Communication with the Raspberry Pi
MQTT Communication Protocol
MQTT (Message Query Telemetry Transport) is a popular IoT communication protocol. We’ll use the Paho Client Python library and an MQTT service called CloudMQTT. Here’s more about MQTT and how to use it:
- Exploring Communication Protocols for IoT
- Getting Started with CloudMQTT
- Overview of Eclipse Paho MQTT Python client library
MEMS Microphone Breakout Board
The MEMS microphone is an analog microphone, so we’ll need an Analog-to-Digital converter (“ADC”) to read in the analog signal with the Raspberry Pi digital GPIO pins.
- Getting started with the SparkFun MEMS Microphone Breakout Board
- MEMS Microphone Datasheet
- MCP3002 ADC Datasheet
We’ll also need the following peripherals: Raspberry Pi 3 Case; SD Card (minimum 8 GB); Raspberry Pi 3 GPIO cable; MicroUSB power cable; HDMI cable and HDMI-compatible monitor; USB keyboard; USB mouse; speakers with 1/8″ headphone port.
Configure the Raspberry Pi
Step 1: Check & Install Updates
Checking for and installing updates is always a good way to start. Run the following commands in the terminal window:
sudo apt-get update
sudo apt-get upgrade
Step 2: Set up SPI Interface for MEMS Microphone + MCP3002
To use the SPI (Serial Port Interface) to read in the MEMS Microphone via the MCP3002, we’ll need the Python Dev Package:
sudo apt-get install python-dev
We’ll also need the SPI Interface (may want to create a subfolder to save this in):
git clone git://github.com/doceme/py-spidev
sudo python setup.py install
Here’s the SPI-Dev Documentation if you run into any issues.
Step 3: Playing Sounds with OMXPlayer
The OMXPlayer is an audio and video player pre-loaded on Raspbian OS. It works with most sound file types, including: .wav, .mp3, and .m4a. This is what we’ll use to play back sounds when Fido gets too loud. The Python library to control the OMXPlayer is included in Raspbian (woo!).
To test the OMXPlayer from the terminal, type the following:
If that doesn’t work, try forcing it over the local audio-out device:
omxplayer -o local /home/.../SongFilePath/SongFileName.mp3
Step 4: Configure CloudMQTT Server
Now we set up an MQTT server! To do this using CloudMQTT, do the following:
- Set up a CloudMQTT account (the “Cute Cat” plan is free).
- Create a new MyCloud instance.
- In the Console, create a new ACL rule.
- You can monitor published messages in the “Websocket” UI.
Finally, install the MQTT Paho Client Python library:
pip install paho-mqtt
Build it! Hardware
Here’s a pinout for the Raspberry Pi 3:
1. Insert MCP3002 pins into breadboard. Here’s a pinout for this board:
The MCP3002 uses 4 SPI pins for communication: Serial Clock (“SCL”), Master Input Slave Output (“MISO”), Master Output Slave Input (“MOSI”), and Chip Select (“CS”). These pins correspond to Raspberry Pi GPIO pin 11 (SCLK), GPIO pin 9 (MISO), GPIO Pin 10 (MOSI), and GPIO Pin 8 (CE0).
Make the following connections with MCP3002 pins:
- Connect Pin 1 to Raspberry Pi GPIO Pin 8 (CE0)
- Connect Pin 2 to the analog output of the MEMS Microphone breakout board
- Connect Pin 4 to GND
- Connect Pin 5 to Raspberry Pi GPIO Pin 10 (MOSI)
- Connect Pin 6 to Raspberry Pi GPIO pin 9 (MISO)
- Connect Pin 7 to Raspberry Pi GPIO Pin 11 (SCLK)
- Connect Pin 8 to Raspberry Pi 3.3V out
2. Solder wires to the MEMS Microphone breakout board. Connect to MCP3002 and Raspberry Pi.
- Connect Vcc to Raspberry Pi 3.3V.
- Connect GND to Raspberry Pi GND
- Connect AUD to MCP3002 Pin 2
3. Plug in all the cables for the Raspberry Pi and turn everything on.
Build it! Software
Our goal with the Bark Back is twofold: trigger a playback sound when the dog barks, and send the data to a server where we can check it.
Here’s the open-source Python program for this project. Feel free to (and please do) adjust and modify the code
To get the program up and running, you need to fill in two things:
– songList: Write in the file path and file name for each of the songs you want to play.
– creds: Input your CloudMQTT information in this dictionary.
Step 1: Read in the SparkFun MEMS Microphone breakout board.
Read in the ADC value (between 0 and 1023) from the MEMS Microphone breakout board (via the MCP3002) using the SPI library and calculate the signal peak-to-peak amplitude.
Map the signal peak-to-peak amplitude to a Volume Unit. The current code maps the ADC range between 0 and 700 (based on quick experimentation) to a Volume Unit between 0 and 10. To adjust the sensitivity of the microphone, adjust the ADC input range.
For a thorough overview of the MEMS mic, check out this tutorial.
Step 2: Trigger audio player.
First we’ll need songs to play! You can quickly record sounds in GarageBand (or on your smartphone) and send ’em to the Raspberry Pi. In Python, use the subprocess library to call the omxplayer.
In the code, input the file path of the songs you want to play back in the *songList* variable (line 26). The current volume threshold is set to 7 in the main function.
Step 3: Send data to CloudMQTT Server
Use the Paho Client Python library to communicate with the CloudMQTT servers. To broadly summarize: Set up a Client server; define communication protocols; connect with our credentials (aka creds); and subscribe and publish our data. Most of this is done in the main function (lines 129 – 149, and lines 169 – 174).
To check on received data, go to the “Websocket UI” tab in the CloudMQTT console.
Test & Install & Enjoy!
Run the BarkBack.py program in Terminal or in the Python IDE (you can also use SSH to run the program after you’ve already left).
Check that you are getting volume levels in your Websocket UI tab.
Test the system by triggering the mic (clap, yell, bark, etc.) to be sure that the speakers play through all of the sounds.
Once everything is up and running, it’s recommended to solder the components to a PCB (Printed Circuit Board) if you intend to install the system for more than just a few days.
1. Add in an RPi camera module to include video.
2. We’re already connected to the IoT, so why not throw in some more sensors! For example, you could monitor hazardous gases (perhaps methane?).
4. Prefer a reward over barking back? Create an automatic treat dispenser to reward your pet for being quiet for a few hours. Here’s a candy-dispensing tutorial for inspiration and to help get you started.
5. Share your findings and adaptations in the comments below!