Sound Pollution Monitoring IoT Platform

This is an IoT platform for real-time noise monitoring and analysis. The platform includes a set of sensors that can be deployed in different locations to collect sound data, which is then processed, analyzed and made available to provide insight into noise pollution levels.

This project was originally realized as a course project in my academic career. I decided to independently extend the solution to allow real-time measurement and use of retrieved data, by creating the entire backend structure.

Features

• Real-time monitoring of sound pollution levels
• Data visualization and analysis of sound data
• Data storicization by means of a MongoDB server
• Scalable data acquisition
• Prototypal IoT devices
• Prototypal Andoird application
• Scalable NodeJS Servers
• Socket based communications for efficient data monitoring

Project Structure

The root folder contains code divided into subfolders for each of the main components of the solution:

• database_mongodb contains some python scripts for testing the most important queries;
• meter_arduino contains code to handle the Arduino micro-controller board (IoT sound meter)
• mqtt_broker contains the script to easily start the MQTT architecture broker on a local machine
• mqtt_publisher_android is a basic Android project to register and send sound level data to the MQTT platform
• mqtt_publisher_raspberry contains python code for the RaspberryPI platform and is required to connect the meter_arduino subproject to the MQTT platform
• mqtt_subscriber is a python project of a MQTT subscriber service which retrieves the data generated and sends it to the MongoDB server
• server_frontend is a basic NodeJS frontend server provider, the frontend is developed in React
• server_monitor is another NodeJS server used to remotely monitor data changes with high efficiency and scalability

The following image displays the main elements of the Sound Pollution platform architecture.

Devices and meters

Two solutions are provided to enable sounds level measurement over time. First solution for noise measurement with Raspberry PI platform and Arduino:

• Analog sensor KY-038 [5] performs the measurement;
• Arduino manages the sensor and sends the data to Raspberry PI via serial channel;
• RaspberryPI receives the data at Arduino and communicates it to an MQTT platform

Second solution with Android smartphone:

• The device's microphone can be used to measure the noise level;
• The Foreground Service framework allows long-duration and background tasks to run;
• The application constantly retrieves data from the microphone and communicates it to an MQTT platform;

The following photo shows an implementation of the demo IoT device using Arduino and Raspberry PI.

MQTT data transfer architecture

Simple overview of the MQTT-based data transfer between devices and IoT platform.

Remote access and dashboard

Monitor server handles all monitoring requests that come in from users

• A device is monitored only once
• Efficiency in relation to connected users

Frontend server provides a web app to users

• OS independence
• A two-way channel between browser and monitor server
• Client requests monitoring
• Server communicates updates