FPGA based Smart and Sustainable Agriculture

• Main Authors: Peraka, Shyam, Ali, SK Irfan, Mogili, Durga Vasu, Palivela, Ashok Kumar, Reddy, Sudheer, Bavisetti, Jyotsna, Y, Dhanush Reddy
• Format: Conference Proceeding
• Language: English
• Subjects: Azure Cloud; Cloud computing; Convolution Neural Networks; Crop Recommendation; Crops; Cyclones; DE10-Nano Cyclone V; Disease Detection; FPGA; Irrigation; Machine Learning Algorithms; Prototypes; Real-time systems; Sensors; Smart Agriculture; Smart Irrigation; Soil; Soil Testing; Sustainable Agriculture; Voice commands; Web Application; Weed Detection; XG Boost; Yolo V5
• ISSN: 2380-6923
• DOI: 10.1109/VLSID57277.2023.00063

The agriculture sector is one of the key areas of concern interlinked with most of the Sustainable Development Goals (SDGs), as stated by most organizations. However, technological advancements and agricultural intervention, particularly in developing and underdeveloped countries, are still in their early stages. IoT, cloud computing, data analysis, and other technologies are being widely used in most areas, where they are bringing significant changes in outcomes compared to previous conventional practices. It is also observed that there have been some initiations in bringing such technologies into agriculture to assist farmers, but they are not comprehensive and real-time working prototypes. Hence, we have come up with a novel, comprehensive solution to address different issues. The focus of the proposed paper is to develop and design a real-time prototype model implementing the following: 1. Control of the irrigation system based on crop stage requirements and other factors such as soil nature, humidity levels, rainfall levels, and so on. 2. Detection of weeds and disease using image processing and machine learning algorithms 3. Identification and recommendation of suitable crops based upon the nature of the soil, fertility levels, climatic conditions, water availability, etc. 4. Voice-based responsive system-mobile application or web-based application. The above tasks were successfully implemented and tested on the prototype developed with the help of the DE10 Nano Cyclone FPGA board. The sensors used in the project include the NPK sensor, moisture sensor, pH sensor, electric conductivity sensor, and water level sensor. Azure cloud services are used for the deployment of the website. The farmer is able to get the voice-based response through a mobile application developed for this purpose.