Multi-material classification of dry recyclables from municipal solid waste based on thermal imaging

Area of Science:

• Environmental Science
• Robotics
• Materials Science

Background:

• Rapid urbanization and industrialization have led to a significant increase in municipal solid waste (MSW) generation globally.
• Developing countries face challenges in managing MSW due to a lack of source segregation practices.
• There is a critical need for automated systems to efficiently sort recyclables from mixed waste streams.

Purpose of the Study:

• To develop and evaluate a thermal imaging-based system for automated classification of recyclables from simulated MSW.
• To integrate this classification system with a robotic sorting mechanism for a single-step process.
• To assess the accuracy and viability of the proposed system for large-scale application in developing countries.

Main Methods:

• A simulated municipal solid waste (MSW) sample was used for experimentation.
• A thermal imaging technique was employed to capture thermal signatures of waste materials.
• The captured thermal data was processed to classify different types of recyclables.
• The classification system was integrated with a robotic arm for automated sorting.

Main Results:

• The thermal imaging system demonstrated the capability to classify useful recyclables from simulated MSW.
• The classification accuracy of the system ranged from 85% to 96%.
• The system achieved comparable accuracy to existing single-material recyclable classification techniques.
• The integration with a robotic system enabled classification and sorting in a single process step.

Conclusions:

• Thermal imaging is a viable technique for the automated classification of recyclables in MSW.
• The developed system offers a promising solution for efficient waste management in developing countries.
• The proposed technology can be a cost-effective and scalable solution for recycling plants processing MSW.