Clean Air

Clean Air Infrastructure

IoT & 3D intelligent systems project 2020-2021

Clean Air is a automatic system able to keep the internal air quality at best possible level, measuring the internal CO2 and the external PM10 and PM25. It opens the windows when the volume of CO2 is high and the PMs values are low enough.

Structure

Description Codebase
Database SQLite
Web App Backend Python/Flask
Web App Frontend Jinja2/Javascript
Android App Java
Predictions Prophet
Window Controller Arduino/C++
Communications HTTP/MQTT

Web App

The Web App is the centre of this system. The backend part manages the database and all the requests, while in the frontend is possible to manage ones dispositives and remotely activate them.

Android App

The Smartphone Client let us add new devices to our profile via QRCode and manage them.

Predictions

Our system is also able to predict PM10 and PM25 pollution values in order to keep the window close even before that the air quality starts worsening. To train the model we use Prophet with open source datasets of italian cities. In production the datasets keep growing thanks to the sensors and the model keeps getting retrined and fine tuned.

Aniello Panariello
Aniello Panariello
PhD Student in Artificial Intelligence

My research interests include Computer Vision and Self-supervised Learning.