<![CDATA[Manual water-quality monitoring and schedule-based feeding may delay responses to unsuitable pond conditions and lead to inappropriate feed dispensing in catfish cultivation. This study developed an Internet of Things (IoT)-based monitoring and automatic feeding system that evaluates water temperature and pH using Mamdani fuzzy inference. The prototype integrates an ESP32 microcontroller, temperature and pH sensors, an ultrasonic sensor for feed-level monitoring, and a servo motor for feed dispensing. Temperature and pH measurements are classified into five linguistic categories and evaluated through a 25-rule base to produce a binary Feed ON or Feed OFF decision. Sensor validation yielded accuracies of 95.52% for temperature measurement, 95.87% for pH measurement, and 95.97% for ultrasonic distance measurement. Validation using five representative input combinations showed that all program outputs matched the expected rule-base decisions, resulting in 100% decision conformity for the tested cases. During seven days of field monitoring, all 21 evaluations conducted at 09:00, 15:00, and 21:00 produced Feed ON because the paired temperature and pH measurements satisfied the minimum rule-base requirements. Web and mobile dashboards displayed sensor measurements, feed availability, device status, operating mode, and feeding activity in real time. The main contribution is a compact and interpretable mechanism that performs condition-based feeding decisions locally on the ESP32 while supporting remote supervision through IoT interfaces. The results indicate adequate sensor performance and consistent fuzzy decision implementation under the tested conditions.]]>
Copyrights © 2026
