Abstract: The aim of the study is to design a farm environmental monitoring system based on the Internet of Things (IoT) to realize automated management of agriculture and the implementation of precision production. The system is composed of three layers: the sensing layer, the transmission layer, and the application layer. A modular structure was adopted to develop coordinator and router nodes with flexible structures and strong versatility. These nodes connect with sensors through standard analog interfaces to build a high-reliability, flexible Wireless Sensor Network (WSN). The WSN perceives environmental information for greenhouse tomato growth in real-time and transmits data to a remote server management system reliably. Field experiments show that the WSN is stable and reliable, providing a scientific basis for farm management.
Introduction
Traditional agricultural monitoring has long faced a trade-off between labor-intensive manual data collection and cumbersome wired sensor networks. Manual gathering is prone to human error and provides only periodic snapshots of data, while wired systems are expensive to install and difficult to reconfigure as crop layouts change. To address these challenges, this research proposes a precision agriculture infrastructure that is wireless, automated, and flexible. The goal is to provide farmers with high-frequency, real-time data—such as soil moisture and light intensity—to optimize yields and resource use without the infrastructure overhead of traditional cabling.
System Architecture
The researchers developed a custom hardware and software stack based on a three-layer IoT model designed to bridge the gap between the field and the end-user.
- Sensing Layer: This layer utilizes an ad hoc Wireless Sensor Network (WSN) based on the ZigBee protocol. It consists of "router nodes" that collect data and relay messages, and a "coordinator node" that acts as the network master.
- Communication Layer: A hybrid approach is employed where local ZigBee data is aggregated by the coordinator and then pushed to the broader internet via a GPRS gateway. This ensures connectivity even in rural areas lacking Wi-Fi.
- Application Layer: This consists of a remote server and monitoring terminal where users can visualize data and send control commands back down the chain to the field hardware.
Hardware and Modular Design
The heart of the system is the MC9S12XS128 microprocessor, which handles 8-channel analog signals and controls 6-channel water valves. The design prioritizes modularity, using standard analog interfaces that make nodes "plug-and-play" with various sensors, including soil temperature, humidity, and Photosynthetic Active Radiation (PAR).
By using a "master-leaf" modular structure, the system remains versatile. It can be adapted for different crops simply by swapping sensors or adjusting the software logic. Furthermore, the inclusion of a 6-channel driver circuit for solenoid valves transforms the system from a passive monitoring tool into an active participant in precision production, enabling automated drip irrigation.
Field Validation and Results
The system was validated through a field experiment at Anhui Agricultural University involving one server, one coordinator node, and ten router nodes. The system monitored soil temperature, soil moisture, air temperature, air humidity, and PAR.
Key results from the field test include:
- Reliability: The WSN remained stable under real-world agricultural conditions, successfully performing data gathering and control operations every 10 seconds.
- Data Integration: Disparate data types were successfully visualized on a single dashboard. Notably, PAR data was scaled by a factor of 20 to allow it to be graphed alongside temperature and humidity for clearer human analysis.
- Remote Command Execution: The system proved capable of routing complex command frames from a remote server through the coordinator to specific router nodes to trigger hardware actions, such as opening irrigation valves.
Discussion and Future Work
While the system demonstrates robust performance, several areas for future development remain. Power management is a critical consideration; although ZigBee is low-power, long-term viability in the field requires further study on battery life and energy harvesting. Additionally, while the 10-node mesh was successful in a greenhouse setting, the latency and performance of the ad hoc network in large-scale industrial farms with hundreds of nodes remain to be tested. Future iterations could also integrate edge computing or AI to allow nodes to make autonomous irrigation decisions based on real-time environmental data.
Conclusion
This IoT-based monitoring system provides a reliable and flexible solution for modern agriculture. By combining the low-power mesh capabilities of ZigBee with the long-range reach of GPRS, the design offers a robust "last-mile" solution for agricultural data, paving the way for more scientific and automated farm management.
Summary updated