Smart Irrigation System Prototype for Chili Plants with Voice Control Using Wit.ai Based on NodeMCU 8266

Abstract

Abstract—Traditional irrigation practices for chili cultivation often lead to inefficient water usage and inconsistent scheduling, adversely affecting crop yields and sustainability. To address these limitations, this study proposes a smart irrigation system prototype utilizing the NodeMCU ESP8266 microcontroller integrated with both environmental sensing and dual-mode voice control. The system operates in two modes: (1) automatic, based on real-time sensor inputs from soil moisture, rainfall, and water level detectors; and (2) manual, through voice commands processed via the Wit.ai API or offline triggers using the KY-037 high-sensitivity sound sensor. The ESP8266 serves as the core controller, executing irrigation logic and relay-based pump activation programmed through the Arduino IDE. Experimental testing demonstrated a 95% accuracy rate in voice command recognition and consistent sensor performance aligned with predefined irrigation thresholds. This dual-control approach ensures operational flexibility under varying connectivity conditions, making it well-suited for small to medium-scale agriculture, particularly in rural environments with intermittent internet access. The system represents an original contribution to the integration of IoT and natural language processing in precision agriculture, with potential for scalable implementation.