Many people have suffered the consequences of not having purified air when spending long periods of time at home: allergies, health issues related to breathing, headaches and anxiety among others. We decided to focus on learning how the air quality is monitored in the ISS in order to apply similar techniques to home air monitoring.

We used basic DIY hardware based on the NodeMCU platform to develop a simple prototype and showcase the idea. The platform would get measurements from a series of sensors and send them over the network to a dedicated server through HTTP requests. A simple webpage allows the user visualizing real-time air quality data, historical average data, and getting customized advices produced by an artificial intelligence. 

The hardware 

The hardware code is built on top of open source Arduino libraries. The system acquires the data from the different sensors, serializes it into a JSON string and sends it to the server by using a POST request.

The platform used for the prototype was the NodeMCU development board, together with the DHT11 sensor for temperature and humidity measurement and the Keyes water sensor to get the water vapour concentration.

In the future, a board with higher capabilities and reliability and the rest of needed sensors would have to be used.

The server side layer 

It is a Web Api built on .Net Core technology and hosted by Azure app services. This project handles http requests, processes business logics and storages data into SQL Server data base. 

Future versions will also have an Artificial Intelligence (AI) module. Through supervised learning techniques, the AI will be able to:

• Request for information to the user to understand the action causing sudden changes or common patterns. For instance, at the same time of day, there is a sudden increase in the oxygen concentration. The AI wouls ask the users about the possible reason behind this fact.
• Analyze global data and user-supplied information to create cause-and-effect knowledge.  Namely, based on user feedbach, the AI could conclude that opening a window in the morning might cause an increase in the oxygen concentration.
• Give advices to keep optimal air quality values. For example, the AI could suggest opening a window to get a better oxygen level to users without a significant increase of oxygen concentration in the morning.

The visualization layer 

Angular 9 Web Application hosted on the free resouce Porkbun offered by Space Apps. That application will be able to:

• Show realtime data graphics
• Visualize historical data
• Input information about sudden changes detected by AI
• Give advices, based on machine learning knowledge, to help users keep Air quality as optimal as possible

It would be nice to have native mobile applications to make use of PUSH Notification, Geolocalization, and other useful smartphone native features. In addition, this would lead to a more accessible and usable user application.

Our real-time monitoring prototype can be visualized here.

And our short video explaining our project is here.

• ESA Air Revitalization System inspired us selecting the particles to be measured by the system
• ECLSS 
• Quality Air Index Calculation reference from MeteoGalicia (Spanish)
• Arduino Reference
• Open Source Libraries for hardware programming: ArduinoJson, ESP8266HTTPClient, ESP8266WiFi.
• FlexClip free online video editor
• Microsoft Azure App Services documentation
• Microsoft dotnet Core documentation
• Stack overflow open community for sharing software knowledge
• Angular documentation 
• ChartJS documentation third party library used by angular
• Swagger API specification and documentation