<![CDATA[Aquaponic systems integrate fish and plant cultivation in a single production cycle, but their success depends heavily on water quality particularly ammonia levels, which can be harmful to living organisms if left unmonitored. This study develops an IoT-based monitoring system using an ESP32 microcontroller equipped with eight sensors (pH, water temperature, air temperature, humidity, DO, EC, TDS, and ammonia) integrated through a sensor fusion approach. Sensor data were processed using mean imputation and Z-score normalization, then analyzed with Pearson Correlation for feature selection and K-Means clustering for anomaly detection in an aquaponic system cultivating Channa striata and Amaranthus sp. Results show that ammonia correlates most strongly with pH (r = 0.50), while correlations with other parameters were relatively low. K-Means successfully distinguished normal from anomalous conditions automatically, and biological testing confirmed that optimal growth occurred at ammonia levels below 1 mg/L. Compared to single-parameter monitoring systems, this multivariate approach provides a more comprehensive picture of environmental conditions and offers a foundation for developing smart, efficient, and sustainable aquaponic systems.]]>
Copyrights © 2026
