IoT Decibel Meter and Dashboard

This is the companion repo for my talk "Bringing ideas to life with DIY IoT: Visualizing Noise Pollution".

Instructions

Setting up the hardware

1. First, you will need a raspberry pi, I'm using a 3B+, but any model that supports HATs should do.

To set up the pi, it is recommended to use their imaging software here

2. Get an Enviro module for the PI. I got the base model without air quality, but either will work. You can purchase them online here or from Microcenter
3. Attach the enviro to the Pi, simply push it down on the pins, no soldering necessary. You can find full instructions for this here
4. Before finding a spot to place your pi, be sure to connect it to your wifi network and take note of its IP address. You will also need to enable SSH if you have not done so already.
5. Find a good spot inside or outside where you want to capture data and place the pi. Make sure you're able to run power to it, it's within range of wifi, and won't get wet.

Server

Next on the agenda is getting a home server ready to go. Nothing fancy is needed here, just a computer that has some storage on it that can store your data and run the dashboard. Once you've got the server, you will need to install Docker on it. The rest of this tutorial will assume you are using Linux or a Linux-like system, if you're using Windows you will need to adjust accordingly.

InfluxDB

1. Pull this repo down on your server, you will then need to open a terminal and navigate to the influx folder as you will be running commands and using some of the files.
2. Next run what is in setup.sh. Influx depends on mytelegraf.conf being in the right place. You can take a look at the file but shouldn't need to change anything in it.

If you get a permissions error on setup.sh, trying running chmod 777 setup.sh

2. Create a influx2.env file. You will then need to populate it with your own values, most of these values can be whatever you want to set them to.

DOCKER_INFLUXDB_INIT_MODE=setup
DOCKER_INFLUXDB_INIT_USERNAME=yourusername
DOCKER_INFLUXDB_INIT_PASSWORD=yourpassword
DOCKER_INFLUXDB_INIT_ORG=yourorgname
DOCKER_INFLUXDB_INIT_BUCKET=environmental
DOCKER_INFLUXDB_INIT_ADMIN_TOKEN=yoursecrettoken

When influx starts up, it will depend on that file being created and in the right place, the root of the influx directory.

3. Now it's time to spin everything up. Take a look at the docker-compose.yml file and make sure everything looks good. Then run docker compose up -d

4. After your containers come up, you should then be able to login to Influx at localhost:8006. The login credentials should be the same values you set in the env file earlier. If something goes wrong, check the logs of your main influx container.

docker ps will list all the containers, then [docker logs](https://> docs.docker.com/reference/cli/docker/container/logs/)

5. Generate an API token and take note of it, you will need it later

6. Login to influx and make sure the bucket you specified in the env was created. Then make sure you can hit influx from a machine other than the server itself. The address should be something like ${server_ip}:8086.

Recording the sound data

Now that you have InfluxDb up and running we can now set up the meter and start sending the data to influx.

1. Login to the pi over ssh and clone this repo to it.
2. Copy the src/.asoundrc file in the repo over to the home directory of the pi
3. Using vi or vim, create src/.env and fill out the following variables

API_TOKEN=yourtokenfrominflux
SERVER_ADDRESS='http://serverip:8086'
ORG='yourorg'
BUCKET='yourbucket'

4. Install the dependencies needed for the python program to run by running pip3 install -r src/requirements.txt

5. Test out the Python script by running python3 src/DecibelMeter.py

6. If all goes well, stop the script. If there are any missing dependencies install them.

7. In src/decibel-meter.service, replace the path for ExecStart to the path where your python script is located.

8. Copy the service file to your system cp src/decibel-meter.service /etc/systemd/system/decibel-meter.service

9. Reload the daemon so it knows about the new service sudo systemctl daemon-reload

10. Enable the new service sudo systemctl enable decibel-meter.service

11. Use the following commands to start, stop, or restart the service

sudo systemctl start decibel-meter
sudo systemctl stop decibel-meter
sudo systemctl restart decibel-meter

You can also give the Docker approach a try. I have started Dockerizing the script but as of now it is not finished. If the service is started, it will continuously run the Python script, even after a reboot.

For some reason, after a reboot, something in the system deletes the .asoundrc file, and the Python script is unable to run successfully without it. Upon reboot, you will need to stop the service, wait for the file to get deleted, restore it, then start the service again. I'm sure there is a solution for this, but I've not put time into solving it at this time.

Most of the code in DecibelMeter.py is not my own. It is generously made available by roscoe81 and has been adapted for the meter. You can find the full list of acknowledgements at the bottom of the file.

Grafana

If you made it this far, congrats! The worst of it is over. Setting up Grafana is fairly simple and straightforward and very similar to how InfluxDB was setup. Grafana is the tool we will be using to actually read the data and build a meaningful dashboard.

1. The easiest way to go about this may be to follow the official docs for running via docker compose, however I have supplied an example grafana/docker-compose.yml here as well. You will need to replace the env values in the file for your own.

On my setup I am hosting grafana on my own domain. I hope to add full instructions for how to do this at some point, but for now that is beyond the scope of this guide. If you have a domain and know how to port-forward, you should be able to set the proper env vars in the compose file and run from there.

2. One Grafana spins up, check to see if you can login at port 3000.
3. Once logged in, go to datasources and setup InfluxDB as a datasource, pointing to it your hosted Influx instance.
4. After getting the datasource configured create a new dashboard and point it to the datasource you just configured. You are now free to add whatever visualizations you want. An example dashboard config has been supplied for you at grafana/dashboard.json. You can import this to a new dashboard but you will probably have to change the datasource as yours will be different.

End

If you made it to the end, nice work! I hope this was a fun and inspiring project for you to work on. If you have found a problem with the code or instructions, please open an issue or a PR to this repo. If you would like to contribute data to the project and hook into my existing dashboard open an issue so we can get in touch.