Smart Garbage Sorting System: Integrating Robotic Arm and Machine Learning for Segregation

Abstract

Efficient waste management is crucial for sustainable development. Emerging technologies like robotics and machine learning offer promising solutions. Automated segregation reduces exposure to hazardous waste materials for workers, promoting safer working conditions and minimizing health risks associated with manual sorting. This paper explores the integration of robotic arm technology with machine learning algorithms for automated dry waste segregation. By leveraging computer vision and deep learning techniques, the robotic arm is trained to identify and sort different types of garbage materials namely paper, plastic, and metal effectively. Key components include image recognition models to classify garbage in realtime and adaptive control algorithms to guide the robotic arm's movements during sorting. During training the three types of garbage are classified with an average accuracy around 99% using GoogleNet. Through this interdisciplinary approach, the garbage is accurately identified and segregated to respective bins.