<![CDATA[Water efficiency in agriculture plays a crucial role in addressing climate change and freshwater scarcity challenges. Conventional irrigation systems often result in excessive water usage due to non-adaptive watering schedules. This study proposes an adaptive automatic irrigation system based on the Internet of Things (IoT) integrated with the Reinforcement Learning (RL) Q-Learning algorithm to optimize water usage. The system utilizes an ESP32 microcontroller connected to DHT22 and YL-69 sensors for real-time monitoring of temperature, humidity, and soil moisture, with data transmitted to the Firebase Realtime Database. The system was experimentally tested for 30 days under three soil moisture conditions with repeated measurements, resulting in an average sensor accuracy of 97.9% and a 70% reduction in daily water consumption compared to manual irrigation. The implementation of RL enables the system to autonomously adjust irrigation decisions based on environmental dynamics while providing remote monitoring via a web dashboard. The results demonstrate that the proposed IoT-RL solution offers an effective, intelligent, and sustainable approach for improving agricultural water resource management.]]>
Copyrights © 2026
