A Scalable IoT-driven Smart Agriculture System: Ontology-based Inference and Automation for Hydroponic Farming

• Published in: Sensors and Materials Vol. 37; no. 3; p. 883
• Main Authors: Lin, Yu-Ju, Tu, Yu-Ming
• Format: Journal Article
• Language: English
• Published: Tokyo 株式会社ミュー 12.03.2025; MYU K.K; MYU Scientific Publishing Division
• Subjects: Agriculture; Automation; Big Data; Cost effectiveness; Crop growth; Data analysis; Data collection; Deep learning; Machine learning; Ontology; Population growth; Real time
• ISSN: 0914-4935; 2435-0869; 2435-0869
• DOI: 10.18494/sam5410

Global warming and increasing disasters have worsened conditions for crop growth, intensifying the global food crisis alongside population growth. IoT technology is critical in smart agriculture, enabling the real-time monitoring and optimization of crop environments through big data analysis and machine learning. However, deep learning models struggle to adapt to diverse conditions owing to reliance on specific training scenarios. In this study, we propose an ontology-based smart agriculture system that emphasizes flexibility and scalability. Unlike deep learning models, ontology models can adapt to different crops or environmental changes by simply adding or modifying relevant classes, eliminating the need for extensive retraining. The system integrates IoT circuits for real-time data collection and ontology reasoning using Owlready2. It automates decision-making and device control, demonstrated in a hydroponic environment where it successfully responded to changes and executed appropriate actions. This approach combines enhanced adaptability, operational efficiency, and cost-effectiveness, lowering the barriers for farmers to adopt smart agriculture and enabling seamless management across diverse scenarios.