<![CDATA[The increasing demand for G0 seed potatoes in Indonesia, reaching approximately 143,740 tons in 2021 while only 8.6% of the demand could be supplied, highlights the need for more efficient production and monitoring systems at the early stage of the potato seed supply chain. Current manual sorting and counting processes are labor-intensive, time-consuming, and prone to human error, creating a need for an automated and reliable monitoring solution. This study develops and implements an Artificial Intelligence of Things (AIoT) system based on the YOLOv11n object detection algorithm for real-time detection, size classification, and counting of G0 seed potatoes. The proposed system integrates a conveyor belt, a Logitech C922 Pro USB webcam for image acquisition, and a laptop as the edge computing unit running the YOLOv11n model. Detection results are transmitted through the MQTT protocol to a Node-RED dashboard for real-time remote monitoring. Unlike conventional approaches, the system combines a lightweight YOLOv11n model with MQTT communication to support simultaneous multi-category size classification and synchronized dashboard visualization. Detected potatoes are classified into three size categories (small, medium, and large) based on calibrated bounding-box pixel areas validated with potato farmers. The model was trained and evaluated using four epoch configurations (25, 50, 75, and 100 epochs) with Precision, Recall, F1-score, mAP@0.5, and mAP@0.5–0.95 as evaluation metrics. The 100-epoch model achieved the best performance, with precision approaching 1.00, recall of approximately 0.98, mAP@0.5 of 0.986, and mAP@0.5–0.95 of 0.96. Validation confirmed that calibrated geometric measurements matched the physical potato dimensions, while dashboard data were fully consistent with edge-computing outputs. These findings demonstrate that the proposed YOLOv11n-based AIoT system provides accurate, reliable, and real-time monitoring of G0 potato production, offering a practical solution to improve operational efficiency and data accuracy in Indonesia's potato seed supply chain.]]>
