Eco sustainable IoT based roof garden monitoring and planting recommendation system with machine learning

Sustainable urban agriculture plays a very important role in the management of food security and environmental issues of rapidly expanding cities. Inefficient irrigation, poor choice of crops and a lack of real-time monitoring are major challenges that are affecting traditional rooftop gardening. To address these issues, this work suggests a smart Internet of Things-based, eco-friendly rooftop garden sensor and machine learning-based planting suggestion system that runs on electricity. The system incorporates a group of low-cost sensors to measure soil moisture, pH, temperature, humidity, and rainfall and an automated irrigation system where the harvested rainwater is used to manage water in an efficient and sustainable manner. The architecture is centered on a Random Forest machine learning module which analyzes the real-time environmental data and decides if it is possible to plant and suggests crops that best fit the current microclimate conditions of the location of the rooftop. The system is designed to work in three synchronous levels of sensing, processing and user interface, with a responsive GrowGreen web dashboard displaying real-time monitoring, irrigation notifications and recommended crops in a ranking order. The experimental findings prove that the Random Forest model attained as high as 92% prediction accuracy and irrigation efficiency of 95%, which proves the practical feasibility of the framework proposed. This platform contributes to retro-friendly rooftop farming by helping to streamline water usage, increase the potential of crops and promoting the growth of intelligent, resilient green cities due to the innovative incorporation of IoT and machine learning technologies.