A Cyber-Physical Digital Twin Architecture Integrating IoT Sensor Networks for Real-Time Multi-Parameter Greenhouse Monitoring

Abstract

Efficient real-time monitoring and control of greenhouse environments remain critical challenges in precision agriculture due to the dynamic interplay of multi-parameter agroenvironmental variables and the lack of integrated cyber-physical decision frameworks. This work develops a cyber-physical Digital Twin (DT)-driven architecture that tightly couples distributed IoT sensing, cloud-based data management, and real-time actuation for intelligent greenhouse management. The proposed system leverages a distributed IoT sensor network based on ESP8266 NodeMCU modules for multi-modal data acquisition (temperature, humidity, light intensity, and soil moisture), interconnected via the MQTT lightweight protocol to a Raspberry Pi edge gateway and HiveMQ cloud infrastructure. A Digital Twin layer synchronizes physical and virtual states, enabling continuous data streaming, real-time visualization, and closedloop control through a web-based dashboard. The system implements threshold-driven control logic to regulate key actuators, including irrigation, heating, and ventilation, ensuring adaptive environmental stabilization. Experimental validation on a polyethylene tunnel greenhouse equipped with six sensor nodes demonstrates reliable real-time operation and effective environmental regulation, maintaining temperature ($20-24^{\circ} \mathrm{C}$) and soil moisture ($60-68 \%$) within optimal agronomic ranges. The architecture exhibits low-latency data transmission and stable control responsiveness under dynamic conditions. The main contribution lies in the end-to-end integration of a practical DTenabled cyber-physical system, bridging sensing, communication, data processing, and actuation within a unified and deployable framework. This approach advances scalable, cost-effective digital farming solutions and establishes a foundation for future integration of predictive analytics and intelligent optimization in smart greenhouse ecosystems.