IoT and Machine Learning-Driven Greenhouse Framework for Smart and Sustainable Agriculture

Agriculture plays a major role in food production, making efficient environmental monitoring and resource management essential for sustainable farming. This project presents an IoT and Machine Learning-Driven Greenhouse Framework for Smart and Sustainable Agriculture. The proposed system uses an Arduino Mega microcontroller integrated with soil moisture, pH, temperature, and humidity sensors to continuously monitor greenhouse environmental conditions in real time. A DHT11 sensor is used for temperature and humidity monitoring, while the soil moisture and pH sensors analyze soil conditions for better crop management. The collected sensor data is processed using a Random Forest machine learning algorithm developed in Python to predict suitable agricultural conditions and support smart farming decisions. A NodeMCU module enables IoT-based cloud uploading for remote monitoring and data analysis. An LCD display shows real-time environmental conditions, while a buzzer provides alerts during abnormal situations. If unfavourable soil conditions are detected, the relay-controlled DC water pump automatically turns ON to normalize soil moisture levels and maintain suitable conditions for plant growth. The proposed system improves water management, supports automated greenhouse monitoring, reduces manual agricultural efforts, and enhances sustainable farming through IoT and machine learning technologies.