Reverse Vending: A Comprehensive Review of Innovative Approach in Embedded System, Sensing Accuracy, and Sustainability Implications

Reverse vending machines (RVMs) offer a promising technological intervention to improve recycling rates and reduce plastic pollution by automating material collection and providing user incentives. This review synthesizes literature (2015–2025) retrieved via Google Scholar and selected through a PRISMA-guided screening to examine RVM design, embedded-system architectures, sensor accuracy, and sustainability impacts. Findings show that contemporary RVMs commonly employ infrared, load-cell, LDR, ultrasonic and similar sensors—often managed by microcontrollers such as Arduino and higher-level processors like Raspberry Pi—to validate and sort plastic bottles with reported detection accuracies frequently exceeding 95%. Despite high baseline performance, studies identify persistent challenges: environmental interference, limited material discrimination, vulnerability to fraudulent inputs, and long-term durability concerns. Advances in sensor fusion, computer-vision classification, depth sensing, and high-precision weight measurement are highlighted as effective strategies to reduce false accepts/rejects and increase robustness. Sustainability and social-impact evidence indicates that RVMs integrated with deposit-return schemes and incentive mechanisms markedly raise return rates and community participation, while hybrid designs (e.g., dual-mode token/coin operation), renewable energy integration, and user-identification controls can enhance accessibility and prevent misuse. The review concludes that combining advanced sensing, layered embedded architectures, and well-designed incentive and governance measures is essential for scaling RVMs as reliable, equitable components of sustainable waste-management systems.