HSTU Student's AI Crop System: 97% Accuracy

A student from Hajee Mohammad Danesh Science and Technology University has developed an AI-based system to recommend crops. This technology analyzes soil and weather data to suggest suitable crops for farmland. Nahid Islam, an ECE student, created this system using real soil data from the university campus. It considers soil pH, nitrogen, phosphorus, potassium, moisture, temperature, and rainfall. The system can choose from 28 crop varieties. It also includes an AI chatbot that answers farmers' questions in Bangla. During testing at a BADC nursery field, the AI model achieved over 97 percent accuracy. The technology has four layers, including an IoT-based soil analysis device and an Android application for farmers. This innovation could help farmers make better decisions about what to plant.