Smart Waste Compactor for Plastic Bottles and Cans

Abstract

This research aims to develop a smart compactor for plastic bottles and aluminum cans, focusing on improving waste management efficiency and promoting systematic recycling. The prototype utilizes image recognition technology through an AI-enabled camera integrated with the CorgiDude board, a high-performance microcontroller capable of running mathematical models, computational algorithms, and machine learning models such as image classification. Additionally, object detection sensors are employed to enhance the accuracy of distinguishing between plastic bottles and aluminum cans. Data from various sensors are collected and processed using a well-trained model before entering the compression phase. Field testing of the prototype in environments such as shopping centers and schools demonstrated high classification performance, with an accuracy rate of 96.88%. These results align with the research objectives and highlight the machine's potential for practical deployment in real-world applications.