Climate-Resilient Smart Agriculture Using IoT and Deep Learning

Abstract

Climate change is having an increasing impact on agriculture, leading to erratic weather patterns and reduced crop yield. Traditional forms of farming do not always solve these problems. In an attempt to aid intelligent farming decisions, this paper proposes a smart agriculture system based on deep learning and the Internet of Things (IoT), designed to be climate-resilient. Real-time monitoring of variables such as temperature, humidity, and soil moisture is made possible by IoT sensors. Deep learning models, such as CNNs for crop health assessment and LSTMs for prediction, are used to analyse the collected data. The system provides suggestions for crop management and irrigation after analyzing them. An accuracy of 92.48% was achieved in classifying the crop using the CNN model. The LSTM model has achieved an RMSE of 0.0071, MAE of 0.0048 and R² of 0.865, indicating good predictive performance.