SMART AGRICULTURE ROBOT

Abstract

Agriculture continues to support a large part of the population, yet many field operations are still carried out by hand. Ploughing small plots, sowing seeds at regular spacing, watering, spraying pesticides, and checking crop health demand considerable labour and time, and farmers increasingly find it difficult to hire workers during peak seasons. Manual spraying also exposes workers to harmful chemicals. This paper presents a smart agriculture robot that performs several routine field tasks, monitors soil and crop conditions with IoT sensors, and reports to the farmer through a cloud dashboard and mobile application. The robot is built on a four-wheel chassis driven by high-torque DC motors and controlled by an ESP32 microcontroller. It carries a seed dispensing mechanism driven by a servo motor, a small tiller blade for loosening soil, a water and pesticide spraying unit with a pump and nozzle, and a soil probe fitted with moisture and temperature sensors. Additional sensors measure air temperature, humidity, and light intensity. A Raspberry Pi with a camera captures images of plants for disease and weed detection, and a solar panel helps charge the battery during operation. The robot can be driven manually from a mobile application or can move along crop rows automatically using ultrasonic sensors and a simple row-following method. As it moves, it drops seeds at set intervals, inserts the soil probe at sampling points, and records readings together with the position along the row. The camera images are processed by a lightweight convolutional neural network that classifies leaves as healthy or affected by common diseases and distinguishes weeds from crop plants, so that spraying can be applied only where it is required. All sensor readings, images, and task logs are sent to a cloud platform using the MQTT protocol over Wi-Fi or a GSM data connection. The dashboard shows soil moisture maps, crop health summaries, and the progress of each task, and the system sends alerts to the farmer's phone when soil moisture is low, disease is detected, or the battery is running out. Irrigation can be started from the application for dry patches identified by the robot.