AI-DRIVEN PRECISE IRRIGATION FOR SUSTAINABLE FRUIT CULTIVATION USING IOT AND DEEP LEARNING

Abstract

Abstract— Precise irrigation is currently attracting a lot of attention as the global population continues to rise, increasing the demand for food and water. As a result, farmers will require water and arable land to meet this demand. Due to the scarcity of both resources, farmers need an alternative solution that modifies their operations. Precision irrigation is the solution for producing larger, higher-quality, and more efficient yields with limited resources. The application of Deep Learning (DL) and the Internet of Things (IoT) is essential for transforming irrigation into a more productive and ecological system. In this research, we used DL and IoT for smart irrigation to improve fruit productivity. For profitable yields, the first important step is to choose the appropriate fruit based on soil and environmental conditions. Next, the fruit should be grown in a controlled environment. For fruit prediction, the Stacked Long Short-Term Memory (Stacked-LSTM) model is used, and to maintain the controlled environment and detect abnormalities, the K-means clustering (KMC) algorithm is used. Both models are validated and deployed in the cloud. The Stacked LSTM model achieves an accuracy of 98.33%, and KMC yields a minimum relative error of 2.49%. Environmental parameters from the field are collected using sensors and sent to the cloud. The DL model in the cloud analyzes the data and provides required results, such as suggesting the appropriate fruit for cultivation. If any environmental factor increases or decreases beyond the normal range, it gives a notification. To provide all these facilities in a user-friendly way, an interactive website is developed.