AI-Powered Multipurpose Agrobot for Precision Farming:Crop Health Management, Fertilizer Recommendation and Yield Prediction

摘要

With almost 50% of its workforce employed and 20.2% of its GDP coming from agricultural products, India is the world's leading producer of these types of goods. Farmers struggle to do simple tasks like ploughing, planting, sowing, watering, and applying pesticides. It is hazardous to use pesticides on plants that are afflicted if they are not knowledgeable about how to use them. It is taking a long time to identify the diseased plant and stop its spread. Robotic precision agriculture could reduce these issues at a frenetic peak. The suggested system, In addition to a seeding mechanism, a pesticide spraying mechanism, a water spraying mechanism, and the capacity to read soil moisture and temperature data via an Android application (Blynk), the suggested prototype makes use of an ESP32 microcontroller and a number of sensors.Further makes the use of YOLO (You Only Look Once) algorithm for recognizing objects, which improves the speed and precision of disease detection on leaves by analysing leaf images in real-time at 45 images per second. The Yolo algorithm will progressively lengthen the time it takes to identify plant diseases, and the deep learning algorithm will suggest the best fertilizer based on the scanned image and the disease's severity level. As a step toward sustainability, solar panels are used to power the entire system with renewable energy. Automatic movement combined with an IR (infrared) sensor and an ultrasonic sensor for obstacle avoidance and path planning can significantly increase efficiency and reduce humanintervention. This quick algorithm (YOLO) will significantly improve the production's efficiency and scalability. and the yield and availability of human/farmer for agriculture are increased by automatic planning, the visualization of sensor data via Blynk, and the recommendation system.