WasteSortBot: An Innovative Solution for Multi-Categorical Waste Separator Bin Using Vision-Based Control With Robot Operating System and Cloud Computing

The increase in environmental pollution is largely attributed to inadequate waste management. A problematic situation occurs when categorical waste is mixed during the initial stages of waste collection. Automated waste management associated with advanced technology assists in sorting out categorical wastes. The technical challenge arises when multiple categorical wastes are sorted simultaneously. Categorical waste sorting can reduce environmental pollution. To address these issues, in this research work, the problem of multiple categorical waste sorting techniques was proposed. The proposed IoT-integrated automated bin (WasteSortBot) can sort out three types of categorical waste with its intelligent, tailored mechanism and algorithm. After experimenting with YOLO models, YOLO version 8 provided the best results for training and testing. For real-time waste detection on an IoT device, the YOLOv8 model was deployed in the cloud for computation. Intelligent features of the system: The waste collector disk, an automated arm integrated with an air pressure mechanism, helps the system accurately sort the category of waste with real-time detection enhancement and low energy cost. The specific Raspberry Pi’s bidirectional communication for model execution on centralized cloud storage, and the real-time monitoring of the fill label on the hosted dashboard of the bin is performed through an internet connection. The proposed WasteSortBot demonstrates accurate waste separation and offers a scalable and effective solution for smart cities with the aim of minimizing manual intervention, optimizing waste collection logistics, reducing environmental pollution, and promoting environmental sustainability.