<![CDATA[Inorganic waste management in waste banks faces challenges in sorting and quality evaluation processes that still rely on manual methods with high levels of subjectivity. Bank Sampah 34 Ngasemrejo experiences problems with community uncertainty regarding waste eligibility standards, causing high material rejection rates and suboptimal community behavior in waste deposit. Therefore, this research aims to develop an automatic inorganic waste eligibility detection system before depositing to waste banks. This research develops an inorganic waste eligibility detection system based on computer vision using the You Only Look Once version 11 (YOLOv11) algorithm to classify plastic bottles, duplex, and newspaper waste based on eligible and ineligible physical conditions for deposit. The research dataset consists of 2,800 images divided into 70% training data, 20% validation data, and 10% testing data. Data preprocessing was performed using the Roboflow platform including annotation, augmentation, and resize to 640x640 pixels. The YOLOv11n model was trained for 50 epochs with optimized hyperparameters. Evaluation results show excellent performance with mAP50 of 99.4%, mAP50-95 of 95.8%, precision rate of 98.3%, and recall of 98.6%. Testing on testing data shows that the system can accurately classify the eligibility of inorganic waste according to waste bank standards. This system is expected to help residents sort waste independently, improve waste bank operational efficiency, and support higher quality and sustainable recycling processes.]]>
