Abstract
This study introduces an AI-driven approach to sustainable agriculture by integrating real-time image analysis with adaptive lighting in hydroponic vertical indoor farming (VIF) systems. Focusing on Lactuca sativa ‘Olmetie’, YOLOv11 and EfficientNetV2S models monitored leaf coloration to assess plant health and adjust light intensity accordingly. Trials compared AI-assisted variable lighting with a conventional 16-hour constant setup, measuring shoot length, leaf width, leaf height, fresh weight, dry weight, root length, and leaf count. Under cooler conditions, the AI-assisted system improved growth uniformity and resource efficiency, achieving a fresh weight of 119.62 g, root length of 18.313 cm, and an average of 22 leaves. Findings highlight the potential of AI-integrated lighting to boost productivity and advance sustainable urban farming.
