AgriTwin-GH: Agricultural Digital Twin for Sm...

AgriTwin-GH: Agricultural Digital Twin for Smart Greenhouse Horticulture of Tomato Cultivation

Author Name : Amala Margret. A, Dr. V. Govindasamy, Arjun Christopher, Vantapati Raja Rajeswari, Bhuvanalakshmi. J. P

DOI: https://doi.org/10.56025/IJARESM.140426327

ABSTRACT Growing tomatoes in a controlled greenhouse sounds straightforward, but in practice it means managing a constantly shifting web of temperature, humidity, CO₂, irrigation, and disease pressure — often all at once. Most greenhouses still handle this reactively, which means problems have already taken hold by the time anyone responds. This paper presents AgriTwin-GH, an AI-driven digital twin framework that flips that dynamic from reactive to predictive. The system fuses time-series environmental data with real-time plant image analysis to enable continuous health monitoring and early risk detection, driving a fully automated control strategy. For crop intelligence, EfficientNet-based deep learning models classify tomato growth stages and detect diseases, reaching 98.61 % and 98.23 % accuracy respectively.