Heru Pranoto, Henokh Sinaga, Ulfa Hasnita

SISTEM MONITORING REAL-TIME PADA DIAGNOSA PENYAKIT KAKAO BERBASIS IOT

Introduction

Sistem monitoring real-time pada diagnosa penyakit kakao berbasis iot. Sistem IoT & pakar untuk diagnosis penyakit kakao real-time. Memantau suhu, kelembaban & tanah, mendeteksi Antraknosa, VSD, Kutu Daun dengan akurasi tinggi untuk pertanian presisi.

Abstract

Cocoa is one of Indonesia's leading commodities with high economic value but is vulnerable to environmental changes and disease infections. This study designs a real-time monitoring and disease diagnosis system for cocoa plants based on the Internet of Things (IoT) and expert systems. The system utilizes an ESP32 microcontroller connected with DHT11, soil moisture sensors (surface and root), and HC-SR04 ultrasonic sensors. Sensor data are transmitted via HTTP POST to a server, displayed on an I2C LCD, and combined with user-selected symptoms through a web interface (HTML/PHP). An expert system processes the data to identify diseases such as Anthracnose, Vascular Streak Dieback (VSD), and Leaf Beetle infestations with up to 95% confidence. Automatic actuators (pump, buzzer, LED) respond to soil moisture conditions. Testing on 20 samples shows stable operation and effective integration between hardware and software, enabling precision farming support.

Review

This paper presents an innovative real-time monitoring and disease diagnosis system for cocoa plants, a critical commodity in Indonesia. Addressing the significant challenge of cocoa vulnerability to environmental changes and disease infections, this research proposes an Internet of Things (IoT) and expert system-based solution to enhance plant health management. The study's core contribution lies in its integrated approach, combining diverse sensor data, web-based user input, and an expert system to provide timely disease identification and automated responses, thereby supporting precision farming practices for a more resilient cocoa production. Methodologically, the system is well-structured, leveraging an ESP32 microcontroller interfaced with a suite of sensors including DHT11 for environmental data, surface and root soil moisture sensors, and HC-SR04 ultrasonic sensors. Sensor data are efficiently transmitted via HTTP POST to a server, displayed locally on an I2C LCD, and accessible through a comprehensive HTML/PHP web interface where users can input additional symptoms. The strength of this system is further exemplified by its expert system component, which effectively diagnoses specific diseases such as Anthracnose, Vascular Streak Dieback (VSD), and Leaf Beetle infestations with a commendable confidence level of up to 95%. Moreover, the integration of automatic actuators (pump, buzzer, LED) for real-time responses to soil moisture conditions demonstrates a practical and proactive approach to plant care, with testing on 20 samples confirming stable operation and effective hardware-software integration. While the current system demonstrates strong potential and effective foundational work, future enhancements could significantly amplify its impact. Expanding the validation process to a larger and more diverse set of cocoa plantations over a prolonged period would provide a more robust assessment of its real-world applicability and reliability. Additionally, integrating advanced machine learning models or image processing techniques, perhaps for early visual symptom detection rather than solely relying on environmental parameters and user input, could complement the expert system, potentially offering even earlier and more autonomous disease identification. Addressing these areas could solidify the system's utility, transforming it from a promising prototype into a comprehensive, scalable solution vital for sustaining and growing Indonesia's important cocoa industry.

Full Text

Comments

You need to be logged in to post a comment.