Multipurpose Agri BOT: An IoT and AI-Based Autonomous System for Smart Farming

Abstract

Agriculture remains the backbone of the Indian economy, but conventional farming practices continue to face limitations due to labor dependency, inefficiency, and lack of automation. This paper presents the design and implementation of a Multipurpose Agri Bot, an IoT-enabled and AI-integrated robotic system that automates essential farming operations such as ploughing, seed sowing, irrigation, and grass cutting. The system uses an Arduino Mega 2560 microcontroller interfaced with sensors including soil moisture, temperature, humidity, and ultrasonic sensors. A Flask-based machine learning API, powered by the Gemini AI Random Forest model, predicts the top five suitable crops based on real-time soil data. The bot is powered by a solar energy system, promoting sustainability and reducing dependence on non-renewable energy sources. Controlled through a Bluetooth-enabled mobile application built in MIT App Inventor, the bot offers remote control, data monitoring, and crop recommendation features. The results demonstrate improved accuracy, cost efficiency, and resource utilization compared to traditional methods, making it suitable for small and medium-scale farms.