IoT-Driven Intelligent Trash Bin (IntelliBIN) Enabled with Edge Segregation for Malls

Urban waste management, especially from commercial sources like shopping malls, presents a significant challenge. A large amount of packaging waste generated by retail outlets and food courts complicates the waste management process. The current method of segregating waste at centralized facilities leads to time-consuming processes, increased transportation costs, and increased labor demands. In response, this work proposes an innovative solution: an IoT-Driven Intelligent Trash Bin with Edge Segregation for shopping malls. By deploying a deep learning model on a Raspberry Pi board, waste segregation is optimized at the edge, reducing latency and improving efficiency. The system includes an intelligent bin named IntelliBIN, featuring four compartments, and utilizes the proposed deep learning model to analyze and segregate the wastes into paper, metal, food, and others via a servo mechanism. In addition, the bin is equipped with ultrasonic sensors to monitor the filled levels of each compartments and a Wi-Fi radio device to enable real-time data transmission and monitoring. The statistical data of waste disposal is presented, and automated notifications are sent to the mall management to take appropriate action. The system assures effective waste segregation and management with the proposed model by achieving a training accuracy of 94%, and it outperforms some of the existing works such as VGG 19 (92%), ResNet (93%), and Inception Net (87%).