<![CDATA[Milkfish farmers in Pangkahwetan, Ujungpangkah, Gresik, commonly rely on visual observations and reactive responses to assess pond-water conditions, underscoring the need for accessible real-time monitoring. This community service programme introduced and demonstrated the Smart-Bandeng System, an Internet of Things (IoT)-based platform that integrates temperature, pH, and turbidity sensors with a web-based monitoring interface and a K-Nearest Neighbors (KNN) classification component. The programme involved the Pangkahwetan Pond Farmers Association, comprising 30 farmers who manage approximately three hectares of ponds. Activities included socialisation, operator training, device installation, field demonstration, mentoring, and follow-up planning. During a documented two-hour field test, sensor readings were transmitted at five-second intervals. The available dataset recorded temperatures of approximately 28.00-28.19 °C and turbidity values of 29-30 NTU. pH was initially 7.03, rose abruptly to 12.68-13.08, and subsequently returned to 7.00. Because this rapid fluctuation was not verified using a calibrated reference instrument, it is interpreted cautiously as a possible sensor, calibration, electrical, or local sampling artefact rather than a confirmed pond-wide change. Programme documentation did not report sensor models, calibration accuracy, participant pretest-posttest outcomes, harvest effects, or validated KNN performance metrics. The primary documented achievement was therefore the successful deployment and real-time demonstration of a community-oriented monitoring prototype, rather than evidence of reduced harvest failure or improved production. Future implementation should prioritise sensor validation, verification of water-quality thresholds, model-performance assessment, operator competency evaluation, and longitudinal monitoring of technology adoption, pond management, and production outcomes before wider operational deployment]]>
Copyrights © 2026
