MLSFDD: Machine-Learning-Based Smart Fire Detection Device for Precision Agriculture

Abstract

The agriculture sector contributes significantly to the overall development of the Indian economy. This sector can be revamped by modern technological interventions like the Internet of Things (IoT) and Machine Learning (ML) along with traditional processes. To improve sustainable growth in the agriculture field, monitoring of parameters like temperature, light, smoke, and flame is given top priority in crop yields. In this work, a smart IoT-based device (Machine Learning-based Smart Fire Detection Device (MLSFDD)) is designed for smart agriculture. The proposed MLSFDD has gathered data from agricultural crop fields through sensors and sensed data are analyzed by employing state-of-the-art ML algorithms like Random Forest (RF), Support Vector Regression (SVR), K-Nearest Neighbours (KNN), and Decision Tree (DT) to detect the fire status by sending the notification through Android phones during strange hours. This model has been realized and examined using raw data received from different sensors. The accuracy, Root Mean Square Error (RMSE), Coefficient of Determination (R <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> ), and Ratio of Prediction to Deviation (RPD) of the proposed model have been calculated via extensive simulation of four ML algorithms. The accuracy of the prediction model of 94% for DT, 93% for RF, 90% for SVR, and 92% for KNN have been achieved. This suggested that mapping the field area’s agricultural fire content can be accomplished using the DT ML model. The study’s findings provide a valuable resource for accurate fire prediction in precision agriculture.

Topics & Concepts