Abstract
In this study, an automated greenhouse prototype has been developed for smart farming applications. With the integration of sensors and actuators in greenhouses, smart systems can now monitor the environment and control plant growth to ensure healthy growth. With Artificial Intelligence, not only the environment, but also plant growth monitoring and disease control can be implemented. The greenhouse environment and plant monitoring system provides a plug-and-play, easy-to-use system that is founded on the principles of efficiency, durability, and low energy consumption. The system consists of four nodes, and all communication among the nodes is wireless. At startup, the sensor-collecting node acquires data from the temperature, humidity, and ambient light sensors. Data from the soil sensor, which is attached to the sensor processing node, is also collected. This information is sent using Bluetooth Low Energy to the sensor processing node. The third node, equipped with a camera, captures images and uses artificial intelligence to detect plants and assess their health. The sensor processing node and the image detection node communicate via MQTT. In the image detection node, a convolutional neural network processes captured images.