Efficient Technology for Detecting Real-Time Sweet lemon leaf Diseases Using IoT: A Thing Speak-Node MCU Based Approach

Abstract

In modern precision agriculture, real-time detection of plant diseases is vital to prevent yield losses, reduce pesticide usage, and enhance crop productivity. While deep learning and edge AI have shown promising results in leaf disease classification, their deployment requires costly and computationally intensive hardware such as Raspberry Pi or Jetson Nano. These limitations make them impractical for large-scale use in economically constrained or rural areas. To bridge this gap, this paper presents an IoT-based, rule-driven diagnostic framework utilizing the NodeMCU ESP8266 microcontroller and ThingSpeak cloud platform. The proposed system leverages low-power sensors (DHT11 and soil moisture) to monitor key environmental parameters indicative of plant disease risk. Disease inference is conducted using logical threshold-based conditions executed through ThingSpeak's MATLAB analytics, enabling real-time alerts via email. The implementation has been evaluated across semi-arid farming plots for tomato and chili crops, showing an alert accuracy of 87%, uptime of 98.9%, and total cost below $15. This low-power system demonstrates high reliability and affordability for real-time field deployment in small-scale farms, thereby supporting the sustainable intensification of agriculture.