Answers are grounded in this paper and cite their sources.
The increasing volume of waste and the limitations of manual and stationary waste management systems have led to a greater need for intelligent, autonomous waste management systems, particularly in indoor spaces such as laboratories, classrooms, and offices. Traditional recycling systems lack interactive features, autonomous capabilities, and real-time monitoring, making them inefficient for human-centered spaces. To address this need, we propose a callable autonomous robotic recycling system for dynamic spaces. The proposed system incorporates a natural human-robot interface, allowing users to invoke the system via a raised-hand gesture detected using a stereo vision approach, followed by triangulation-based localization, SLAM-based navigation and obstacle avoidance, real-time waste classification using YOLOv8, ultrasonic-based waste fill level detection, and edge-cloud-based data logging through a unified event-driven state architecture. The proposed system is a closed-loop cyber-physical system that integrates perception, localization, navigation, classification, monitoring, and docking to enable real-time waste management. Autonomous navigation was validated in real-world laboratory settings, demonstrating the proposed system’s ability to map and navigate to a user-defined target location while avoiding static obstacles. Waste classification performance was validated on the TrashNet dataset, achieving an F1 score of 0.981 at a confidence of 0.643 and an mAP of 0.989. In addition, raised-hand detection for natural user interaction was validated on a filtered subset of the HaGRID dataset, achieving a precision of 0.88, recall of 0.85, F1-score of 0.86, AP@0.50 of 0.95, and AP@0.75 of 0.89. The proposed system has demonstrated its ability and efficacy through a unified approach to interactive waste recycling using an autonomous system and real-time waste classification and monitoring.
No takes yet. Share an insight, caveat, or question.
Yaghi et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: