Waste Management with IoT and Interoperability

In today’s rapidly evolving world, cities are growing faster than ever, and with this growth comes an increasing challenge of managing waste efficiently. Overflowing bins, inefficient collection routes, and environmental pollution are common problems faced by many urban areas. Traditional waste management methods struggle to keep up with the demand, leading to unnecessary costs and environmental harm. But why don’t we have a system that’s as smart as the spaces we live in?

Imagine a world where waste management adapts to our needs, seamlessly integrating with the environment to make every city cleaner, greener, and more efficient. A system where waste levels are constantly monitored, collection routes are optimized in real-time, and data from sensors and cameras inform us about waste while reducing costs and waste overflow.

Motivation

Smart waste management arises from the challenges of urbanization, where traditional systems struggle to keep up with rising population density and waste production. Pollution from inefficient waste management harms both the environment and public health. A smarter, real-time, and automated approach can optimize waste collection and contribute to healthier urban environment.

Problem Statement

This project focuses on developing a real-time smart waste management system powered by IoT technology. This revolves around inefficient waste management systems that struggle with real-time monitoring, data integration, and interoperability. Traditional waste management methods often lack real-time insights into waste levels, leading to delays in collection and unnecessary trips. However, current systems face challenges in ensuring seamless interoperability across different devices.

Our Solution

Our solution leverages IoT technology to transform waste management by providing real-time monitoring and optimization. We integrate sensors and cameras to continuously track waste levels in bins, detecting when they are full and need to be emptied. Gas sensors are also utilized to monitor odors providing additional insights into waste content. In addition, the camera tracks the number of people dumping waste and last cleaning time.

The system ensures interoperability, enabling seamless communication between devices. The platform’s data-driven approach enables timely interventions to avoid bin overflow and maintain optimal waste management operations. The sensor data is transmitted in real-time from the camera and controller to OM2M.

Additionally, we developed a dashboard to showcase the bin levels, where data from OM2M is retrieved through subscription feature and stored in a database for easy access. The dashboard provides a user-friendly interface for monitoring the status of all bins. It includes analytics on waste dumping violations at each location, displaying the bin count, the last cleaning time, and the original bin image for each site.

This provides valuable insights into waste management activities, helping identify issues and track the cleanliness and maintenance status of bins across the area. By automating waste collection and leveraging real-time data, the solution reduces operational costs, enhances sustainability, and improves the overall efficiency of waste management processes.

Programming Environment

• Camera and Sensors: Utilized Raspberry Pi for controlling hardware, implemented machine learning algorithms to process data detected by cameras, and ESP32 for gas sensor.
• Middleware: OM2M (based one oneM2M standards)
• Database: PostgreSQL
• Programming Language: Python, C, JS
• Backend: Built with Node.js to handle data processing, API integration, and dashboard functionalities.
• Frontend: Developed using ReactJs for dynamic UI and Leaflet for interactive maps.

How it Works

We have designed, developed, and deployed two devices that will sense, collect, process data from the device location. The first device is an Arduino, which comprises a ESP32 controller and a gas sensor. This is in charge of detecting emission of gases. Another one is camera which is connected to raspberry pi collects images and process the data to identify bin level and dump violations and converts bin image to base 64.

These sensors collect data and sends to OM2M (based on oneM2M standards). Then the data store in a database retrieved from OM2M through subscription. The final dashboard will connect to this and visualize the data for insights. This also consists of drilldown capabilities with bin origin images.

This modular design ensures focused data handling by the Raspberry Pi and ESP32 while leveraging the interoperability of oneM2M for centralized processing. It addresses critical waste management challenges by enabling remote monitoring, improving efficiency, and supporting sustainable practices through advanced data analytics.

Hardware Setup

The architecture represents a waste management IoT system integrating diverse components for monitoring. The Raspberry Pi focuses on handling camera data and capturing visual information related to waste, such as type identification, quantity estimation, or operational monitoring. The Camera is connected to Wi-Fi. Raspberry Pi, which is the edge device used for processing connects to the camera via RTSP and captures live feed to perform real-time bin occupancy detection and people counting. The captured video feed and applies background subtraction for people detection, tracking individuals as they cross predefined regions. For bin monitoring, the model analyzes a specific ROI in the frame to determine if the bin is full, half-full, or empty based on edge detection. The data, including people count and bin status, is sent to an OM2M platform via HTTP POST requests.

Meanwhile, the ESP32 microcontroller independently manages data collection from environmental sensors, specifically the CO2 and Gas sensors (methane detection). These sensors monitor the emission of gases, which is critical for assessing decomposition and ecological impact. The Adafruit Display likely offers localized information for immediate on-site monitoring. All collected data—whether from the camera, CO2 sensor, or methane sensor—flows into the oneM2M platform, which standardizes communication and enables seamless integration across devices.

Software Setup

We have used OM2M the open-source Interoperability Layer. The OM2M acts as central framework/ Middleware, facilitating data transfer to a database for storage and analysis.

oneM2M API Usage and Overview

The OM2M resource tree is as follows

Access Control Policy (ACP)

ACP defines the rules and policies that govern access to resources within a OM2M. It specifies the permissions and restrictions associated with different types of devices and entities. In the architecture we have used acp_admin as default ACP for posting data from the sensor end and retrieving data as well.

Subscription

The external script will subscribe to container discovered previously according to all the sensors specified in infrastructure. Each sensor will post on each corresponding node in onem2m. After having subscribed, the script will listen for each subscription sector as parameters specified.

After receiving a request, the script will store in the database the new event, depending on what kind of data is transmitted. On the other hand, the script keeps a relation between the database and the http client. The client will ask for data to be displayed and returns the data.

Dashboard

A dashboard is connected to the database for bin data and OM2M for gas parameter details, providing real-time insights and analytics for waste management stakeholders.

The dashboard initially displays nodes at each location. When a user selects a specific location, it then presents detailed insights for that location.

The dashboard also suggests optimized collection routes based on bin status, recommending the shortest path to bins that are full.

Features

• User friendly dashboard
• Drill Down capabilities
• Real-time Bin Monitoring
• Data Integration
• Location-based Insights
• Origin Bin Images