AUTORECYCLER: Prototype based on artificial vision to automate the material classification process (Plastic, Glass,...

Area of Science:

• Environmental Science
• Computer Science
• Engineering

Background:

• Growing global pollution necessitates improved waste management.
• Inadequate classification of solid waste presents a significant environmental challenge worldwide.

Purpose of the Study:

• To develop an automated system for recognizing and separating recyclable materials.
• To leverage artificial vision and machine learning for efficient waste classification.

Main Methods:

• An automated sorting system utilizing a webcam and Nvidia® Jetson Nano™.
• Implementation of a convolutional neural network (CNN) trained for waste material recognition.
• Real-time classification of Plastic, Glass, Cardboard, and Metal.

Main Results:

• The system achieved 95% accuracy in separating plastic.
• Achieved 96% accuracy for glass and metal separation.
• Demonstrated 94% accuracy in cardboard classification.

Conclusions:

• The developed system effectively contributes to recycling initiatives.
• Automated waste sorting has a positive impact on reducing global environmental pollution.