During the California Wildfires of 2018 and 2019 San Francisco air quality turned dangerously unhealthy. This made me interested in having an indoor air quality monitoring station. At the end of 2019 I moved to Shenzhen, China where I again experienced multiple 'bad air' days. Faced with free time during the COVID-19 pandemic I decided to take action and build a wifi enabled air quality monitoring station.
As of October 25, 2020 this project has reached a public state and is visible at https://dudeyo.com/particle.
The remainder of this readme is an overview of the project architecture and a timeline of the development history. Questions, comments, and feedback are welcome at jasonford1 at gmail.
Hardware instructions can be found at https://docs.particle.io/quickstart/aqmk-project/
Wifi and GPIO are provided by the Particle Argon. Additional sensors measure temp, humidity, air pressure, harmful gas levels, and pm2.5 particulate levels.
The sensors are attached via sticky tack to the wall behind my desk. Sensor measurements are recorded every 30 seconds and sent over wifi to console.particle.io.
When console.particle.io receives sensor data an event is triggered that sends a webhook (POST request) to dudeyo.com/particle. The dudeyo.com Linux server receives the POST request, processes the JSON data package, and then uploads the data to an Azure SQL instance for long term storage.
When an end user visits dudeyo.com/particle the Linux server responds with a static HTML page. The backend server then requests current conditions and the last hour of data from Azure SQL. Once Azure responds the data is sent to the end user for local rendering using D3.js.
March 5, 2020
- Particle Argon, sensors, and display are assembled into a cohesive working device
- Code from https://github.com/particle-iot/air-quality-kit is customized and deployed to the Particle Argon
- Device measures and displays air quality from my office desk
October 23, 2020
- Particle IoT features implemented such that sensors communicate via local wifi with console.particle.io
- Particle.io webhook implemented
- Linux server deployed as backend for the project
- Nginx deployed on linux server as reverse proxy
- NodeJS and PM2 deployed as the production and development environments to maintain (yet to be written) Particle.js app
- Data model developed and deployed to Azure SQL instance
October 25, 2020
- Particle.js app (this repository) is written and deployed to Linux server
- Particle.js receives console.particle.io webhook POST request
- Particle.js processes JSON sensor data package
- Particle.js sends sensor data to Azure SQL database via Tedious
- App requirements: Express, Tedious
December 18, 2020
- MSSQL implemented for Azure SQL connection pooling
- Particle.js processes requests for last hour, 24 hours, 7 days, 30 days, or all data for all sensors
- End user can choose timeperiod of last hour, 24 hours, 7 days, 30 days, or all data for all sensors
- App requirements: Express, MSSQL(added), D3(added), Tedious(removed)
December 21, 2020
- Particle.js processes requests for current conditions
- Current conditions rendered in graph title for improved at-a-glance viewing
December 23, 2020
- Particle.js renders graphs dynamically to window size on page load
- Bootstrap implemented
- Many small improvements in spacing and typography
- Graphs render within their own card
- Added 'seconds since last sensor reading'
- App requirements: Express, MSSQL, D3, Bootstrap(added)
Notifications
- Ability to set thresholds for notification
- SMS or push notification sent when a threshold is met
Frontend improvements
- Custom date ranges for graphs
- Timeperiod selection by graph instead of globally
- Graphs resize on window resize events
- Graphs show value on hover at mouse location
- Graphs show progress indicator while data is being fetched
Backend improvements
- Implement data compression strategies for better load times:
- Reduce data resolution by finding time block averages (eg. a single data point is 5 minutes or 30 minutes of averaged 30 second data)
- Select graph timeperiods individually such that graphs can request their own data sets independently
The code deployed to Particle Argon is a variation of the code found at https://github.com/particle-iot/air-quality-kit. The modifications made were primarily in changing the brightness of onboard LEDs and screen such that they dimmed or turned off, especially at night.