Efektivitas Teknologi Pemilahan Limbah Medis: Meta-Analisis Terhadap Sistem Sensor, IoT, Robotika, Dan Klasifikasi Citra Digital

The mixing of medical and non-medical waste in healthcare facilities increases the risk of workplace accidents, such as the transmission of nosocomial infections, and raises hospitals’ operating costs. The use of smart digital technology in smart waste bins offers a solution for automatic waste sorting without direct contact. This study employed a meta-analysis approach in accordance with the PRISMA guidelines. A literature search was conducted in the PubMed, ScienceDirect, and Google Scholar databases for publications from 2021 to 2026. Of the 164 studies identified, 10 articles that met the inclusion criteria were extracted and critically analyzed using a random-effects model and narrative synthesis. The results indicate that segregation technologies utilizing AI-based image processing and physical sensors achieved waste classification accuracy rates of 90% to 100%. The Inception-V3 architecture delivered the highest medical image accuracy at 93%, while the Inductive Proximity sensor achieved 100% accuracy with an extremely fast response time. The implementation of edge computing based on local microcontrollers has proven capable of eliminating dependence on the internet, saving power consumption, and reducing direct physical contact by users by up to 85%. Additionally, the integration of volume monitoring and IoT telemetry achieves waste volume measurement accuracy of up to 98%, with network node efficiency ranging from 93.75% to 96.66%. The implementation of Smart Waste Bin technology based on the integration of AI, edge computing, and electronic sensors has proven to be effective, hygienic, and economical in improving compliance with medical waste sorting.