IoT and Data Analytics for Greenhouse Optimization: A Critical Review of Methods, Limitations, and Future Directions

Abstract

The integration of Internet of Things (IoT) technologies with advanced data analytics is transforming greenhouse agriculture by enabling greater precision, automation, and resource efficiency. This review critically examines recent developments in IoT-enabled greenhouse optimization, focusing on sensor technologies, communication infrastructures, and machine learning models, digital twins, and autonomous control systems. Drawing on peer-reviewed studies, recent reports, and emerging research trends, the review moves beyond descriptive analytics to evaluate practical deployment challenges and real-world applicability. Although predictive analytics, deep learning, and reinforcement learning have demonstrated strong performance in controlled environments, their adoption in commercial greenhouses remains limited by data scarcity, poor model generalization, simulation-to-reality gaps, interoperability issues, and the absence of standardized benchmarking frameworks. The review further highlights challenges related to cybersecurity, explainable artificial intelligence, federated learning, and economic scalability. While digital twin technologies show potential for optimization and decision support, their widespread implementation is constrained by calibration and synchronization complexities. By identifying these technological and research gaps, this review proposes a roadmap for developing robust, scalable, and autonomous greenhouse systems. The findings emphasize the need to shift from algorithm-centric research toward deployment-oriented solutions that support sustainable and intelligent protected agriculture.