Development of a smart irrigation monitoring system employing the wireless sensor network for agricultural water management

IoT

Internet of Things

PCB

Printed Circuit Board

GPRS

General Packet Radio Service

TCP/IP

Transmission Control Protocol/Internet Protocol

UDP

User Datagram Protocol

FTP

File Transfer Protocol

CoAP

Constrained Application Protocol

GSM

Global System for Mobile Communications

FAO

Food and Agriculture Organization

TDR

Time-Domain Reflectometry

Many farmers in India irrigate their crops without adhering to irrigation scheduling criteria (George et al. 2000). This prevailing practice leads to various detrimental consequences. Enhancing agricultural yields and promoting sustainable agriculture need irrigation. However, there is a significant gap between the potential and actual use of irrigation. Most modernization efforts in India focus on implementing micro-irrigation for tube well systems, addressing the problem of water table decline experienced in many districts (Chourasia et al. 2024). Over-irrigation and under-irrigation are major challenges in agriculture. Excessive irrigation water usage contributes to waterlogging, wastage of valuable water resources, plant diseases, and soil salinity; conversely, inadequate water supply results in crop water stress. Farmers continue to employ traditional methods based on assumptions about the crop's water requirements, and farming practices have not evolved significantly over time (Gsangaya et al. 2020). To meet future food requirements while minimizing water usage, farmers must adopt precise irrigation scheduling techniques that rely on smart irrigation utilizing soil moisture sensors (Kim et al. 2008). These techniques are critical in enhancing agricultural production, optimizing irrigation efficiency, reducing water wastage, and mitigating harmful environmental impacts. Adopting advanced irrigation methods is essential for efficient water management, the precise irrigation scheduling of crops, and avoiding wasteful water use on the farm. Precision agriculture, combined with wireless sensor networks (WSNs), is essential in driving automation within the agricultural sector (Shafi et al. 2019). It involves collecting real-time data on soil, crops, and weather conditions using sensors deployed in the fields. Precision irrigation delivers water and nutrients to crops with optimal timing, placement, and quantity, ensuring efficient resource use and enhanced crop performance (González Perea et al. 2018; Costanzo et al. 2024). Using the WSN for irrigation monitoring has generated significant interest in the research community. The WSN has emerged as a promising technology with numerous advantages in irrigation management (Rasooli et al. 2020). Its real-time monitoring and data collection capacity has led to widespread recognition and exploration in this field. Using the WSN, farmers and stakeholders can efficiently monitor and evaluate various irrigation parameters, enabling informed decision-making and enhanced management of water resources in agricultural practices. Farmers often lack weather forecasts and provide irrigation without considering potential rainfall events. This lack of information can lead to excessive water use, waste energy and money, and negatively impact crop growth due to overwatering. Rainfall can occur on the same day as irrigation, before or after. Therefore, it is essential to consider these factors to save water (Goap et al. 2018). Integrating weather forecast models with smart irrigation design tools can help achieve this goal effectively. Smart irrigation can be improved by optimizing water use with the help of IoT-based technologies, which are proving to be highly beneficial in many areas of agriculture (Sharma et al. 2016). When designing a smart irrigation system, important factors, including soil moisture, precipitation, and evaporation, must be considered. Notably, soil moisture levels are strongly impacted by precipitation and evaporation, which, in turn, influence irrigation decisions. An automated irrigation scheduling system based on direct soil water measurements was proposed by O'Shaughnessy & Evett (2010), offering a more water-efficient irrigation method than manual techniques. Allen et al. (1998) introduced an evapotranspiration (ET) method as a crucial factor in determining crop water requirements. This method is influenced by various climatic factors, including solar radiation, relative humidity, temperature, wind speed, and crop-specific characteristics such as growth stage, variety, plant density, and soil conditions. In this study, we used an empirical model given by the modified Penman method in the FAO56 manual to determine the crop's water requirement.

For the past decade, the integration of WSN–IoT for automation in irrigation practices has achieved some success in concurrently overcoming the limitations; for example, Kumar (2017) developed a WSN architecture that includes an Arduino end node and a Raspberry Pi coordination node (CR node) for constructing an automated irrigation and security system. He used the nRF24L01 transceiver for data transfer and wireless communication between the end node and the CR node. However, this device's ability to gather and transmit data requires human intervention and depends on commands from an external server. Katyara et al. (2017) implemented a WSN as a remote terminal unit (RTU) for remote monitoring and the smart control of irrigation systems in Pakistan. Various data, such as soil moisture and temperature, were measured by these RTUs, and these data were sent to estimate and control the amount of water needed during irrigation activity. The results of tests showed positive results in reducing water used in irrigation and increasing agricultural land's productivity by almost 20–25%. Montesano et al. (2018) designed a sensor-based irrigation system for automated irrigation. The findings suggest that employing a WSN to monitor substrate water status in real time proves to be a valuable tool for precision irrigation management. This irrigation approach has the potential to be customized for various crops when combined with accurate information regarding the impact of different water availability levels. Nagarajan & Minu (2018) proposed an automated sprinkler irrigation system utilizing a WSN. Their design incorporates ZigBee technology for data transmission and GPRS technology for data storage and analysis. The system uses sensors to monitor several soil factors, including temperature, pH, and humidity. Real-time monitoring and control of soil parameters, such as temperature and water content, are made possible by transmitting the gathered data to a central controller. Additionally, the system facilitates efficient water supply control and optimization. Hamami & Nassereddine (2020) reviewed the application of WSNs in irrigation. They suggested that sensors (e.g., soil moisture, soil temperature, humidity, and pH) significantly contribute to improving irrigation systems, so they play a crucial role in promoting measurement techniques for different agricultural factors. Furthermore, the basic rules for designing an automated irrigation system should be a simple and understandable interface for effective communication with the farmer/user. However, an automated system for smartly managing irrigation using WSNs should include sensor nodes, coordinator nodes, and communication technologies (e.g., LoRa, Wi-Fi, and ZigBee). This survey concludes that a smart irrigation system can be efficient and offers significant potential to economize irrigation water if it is well and properly planned, managed, and maintained. Tenreiro et al. (2020) highlighted the framework of precision agriculture with soil hydraulic properties and hydrologic models to address spatial variability in different crops. Field-level simulations of water processes and flows in agricultural fields can be enhanced by crop modelling and hydraulic characteristics (Hussein et al. 2011; Jin et al. 2018) for a better understanding of how regional water fluctuations affect crop behaviour and overall field performance may be possible with this synergistic approach. Naji & Salman (2021) introduced a smart irrigation system that uses a WSN based on Arduino and XBee technologies. The system comprises sensor nodes to monitor agricultural conditions and automatically regulate soil moisture levels within the optimal root zone range. The results obtained from a pilot prototype indicate that a significant amount of water-saving can be obtained when using a smart irrigation system instead of the conventional approach. Matilla et al. (2022) conducted a comprehensive review of automation in irrigation systems, the evolution of communication techniques, and the introduction of budget-friendly monitoring systems for pivot irrigation. Their findings underscore a significant contribution: integrating affordable sensors with cutting-edge communication technologies such as Edge computing. This combination empowers farmers to monitor their fields efficiently without substantial investments. Ndunagu et al. (2022) developed a prototype IoT-based smart irrigation system for small vegetable farms. They demonstrate the effectiveness of a smart irrigation system using drip irrigation with corrosion-resistant soil moisture and DS18B20 soil temperature sensors. The open-source IoT cloud platform ‘Thingspeak’ was utilized for data collection, storage, analytics, and visualization. The design saves farmers time and money on irrigation, improves efficiency, and increases crop yield. Furthermore, this study suggested that solar power and backup batteries should be used to supplement the main power supply. The results indicate that the system is highly efficient and reliable in irrigation and managing water resources. It can also be adopted in rural areas to boost agricultural productivity. Costanzo et al. (2024) highlighted the concept of precision surface irrigation by focusing on three core principles: organized field geometry, precise control of hydraulic–hydrological variables, and regular performance evaluation. The work emphasizes the benefits of combining one-dimensional (WinSRFR) and two-dimensional (IrriSurf2D) modelling tools to represent the complex nature of surface irrigation dynamics by applying such concepts to a simulation environment. Additionally, the two-dimensional model offered a detailed spatial assessment of irrigation performance, surpassing the one-dimensional approach in scenarios with cross-sectional variability. Ultimately, the two-dimensional modelling approach proves to be a powerful tool for spatially characterizing surface irrigation dynamics, making it highly suitable for implementing precision surface irrigation in diverse agricultural water management.

Numerous studies have explored the design of smart irrigation systems, as highlighted in the literature. However, most of these studies utilized industrial-grade soil moisture sensors for data collection and employed open-source IoT cloud platforms such as Thingspeak and Blynk for data visualization. While several prototype innovative irrigation models have been developed, most were tested only in small-scale farms or laboratories. Few studies have considered environmental factors (e.g., rainfall forecasting, ET, and soil temperature) in their irrigation models. Additionally, most irrigation systems lack the design of energy-efficient power supplies. This negligence frequently leads to the waste of energy and freshwater and negative impacts on crop growth, including overwatering when irrigation is followed shortly by rain. The development of automated irrigation systems that depend on the continuous monitoring of soil moisture patterns and upcoming precipitation information is essential for the efficient and optimal use of freshwater in irrigation. Therefore, developing a smart irrigation model that can be directly applied to large-scale agricultural lands is crucial, providing practical benefits to farmers.

We proposed a novel sensor node designed with an Android application (SIMS) and used PCB-based in-house designed FR4 capacitive soil moisture sensors, which are tested and calibrated with the standard gravimetric method. The results of these calibrations revealed that 85% of the sensors accurately detected patterns in soil moisture fluctuations during the cropping period (Kushwaha et al. 2024). This study presents a design of WSN–edge–cloud architecture for real-time irrigation monitoring under field conditions. The system is integrated with the upcoming rainfall event model, which receives daily data from the Internet through an automatic weather station. Additionally, the system provides irrigation based on the net irrigation requirements (NIRs) while accounting for anticipated future rainfall events. This design uses an energy-efficient node (i.e., 0.011 mA at 4.2 V) to minimize the loss of power compared to current sensors for soil moisture and make it easier for farmers to use in the irrigation field. To summarize, the design aims of this study are as follows:

• A design of a WSN–Edge–Cloud-based architecture, featuring a specific sensor node (iNODE) to support smart irrigation, along with a DS18B20 soil temperature sensor and PCB-based fringe capacitive soil moisture sensors, is proposed. These sensors are strategically placed within the wheat crop field to observe real-time moisture levels and monitor soil temperature to improve irrigation management.

• A smart irrigation monitoring system (SIMS) is implemented, and its possibilities are illustrated in a real-world scenario, i.e., irrigation monitoring and an Android application.

• Experimental research is performed to determine the advantages of the configuration mode, utilizing real-time soil moisture data.

• The rainfall event model is integrated with a designed smart irrigation system for optimal automated irrigation scheduling.

• It is an automated irrigation system based on an Android application (SIMS App).

The hardware component of this layer is mainly focused on iNODE, which is the core element of this module. The continuous collaboration and interoperability of iNODE's hardware and software components, which operate perfectly harmoniously, are essential to its efficient operation. The necessary hardware components for the device's operation include a module for communication, power circuits, general-purpose input–output (GPIO) interface channels, and microcontroller units (MCUs). On the other hand, the software is necessary for managing various tasks, such as data aggregation, data integrity checks to ensure against corruption, and energy optimization to ensure the iNODE uses its power effectively. Together, the hardware and software elements build a reliable and effective system that enables iNODE to carry out its intended functions successfully.

The power consumption of sensor nodes depends upon the specific sensors, including the number and type of sensors. As a result, the power consumption of each iNODE varies when it is in active mode and when it enters sleep mode. In the scope of this research, two iNODEs were utilized, each interfacing with a total of four sensors (comprising one for soil temperature and three for soil moisture). Table 1 shows the power consumption in various components interfaced with a module.

It is critical that these devices can run continuously without requiring much maintenance when using WSN nodes for soil moisture monitoring (Zhang 2004). It is essential to optimize the power consumption of sensor nodes since wireless monitoring devices have limited battery life (Roy et al. 2020). When the low-power external interrupt detection feature is on, the iNODE's power consumption in sleep mode is 0.011 mA at 4.2 V, as shown in Table 1. Here, we present the expected operational lifetime and energy usage of the iNODE at idle mode. Consider the case where an iNODE is set to detect the data for 5, uses 1 s in idle mode to report data, and switches to sleep mode for 20 min. Considering one soil moisture sensor, one soil temperature sensor, and one AA-size Li-ion battery interfaced with a node at idle conditions, the power consumption and life of the node are computed as follows:

In Equation (1), E_iNODE denotes the hourly energy consumption in J. If the number of times iNODE switches to active mode in an hour is n, then t_active = active mode time in seconds and V = nominal voltage of Li-ion battery.

I_sleep = Current consumption in sleep mode.

This design offers user-friendly management options for the WSN, allowing configuration adjustments to optimize its performance. Subsequently, the data are recorded into a file stored on the SD card, located within an edge server, and transmitted to a cloud server via the Internet. The edge node operates a small server and keeps a database of the wheat crop's required soil moisture levels. Additionally, it can control the opening and closing of the water supply pump. The edge server activates the irrigation water pump after verifying the received data parameters with the database when it receives inputs from the sensor nodes. The edge node turns on the water supply until the amount of soil moisture detected by every sensor reaches preset threshold values. A MQTT v5 broker establishes a connection between the edge devices and the shared repository through a distributed local network. The module uses the MQTT as a gateway between edge devices since the edge devices run on batteries, which lowers power usage. Furthermore, every edge device has a ZigBee module to link the sensor nodes. The edge network's common repository contains the incremental data log file of soil moisture-related data relayed from end nodes. In addition to serving as an edge server, the most substantial edge device manages the shared storage.

The aim of scientific and precise soil moisture-based irrigation scheduling approaches is to increase irrigation efficiency, crop growth, and yields by giving the crops with the proper amount of water at the right time (Gu et al. 2021). To determine soil moisture depletion and, subsequently, the need for irrigation, sensor values must be used with the soil moisture thresholds of FC and PWP (Sui 2017). This approach conducted a performance analysis to monitor optimal water according to the NIR. When the edge server receives data from the nodes, it verifies the received data parameters with the database and accordingly actuates the notification for irrigation purposes. The sensors' observed digital values range from 630, representing dry soil (with a minimum threshold of 19.5% moisture), down to 284, which denotes wet soil (within the maximum threshold of 40% moisture). These pre-threshold values are decided according to the FC of the soil. If the moisture content reaches its FC (SMC > 19.5%), the microcontroller gets activated, sending a signal to the central server to stop the irrigation. However, the data received from three capacitive soil moisture sensors placed at different depths showed that soil moisture is the highest at 45 cm depth. Interestingly, readings from the sensor at 30 cm constantly match the intended soil moisture levels. Despite these variations, our irrigation strategy for the field is designed to optimize water use efficiency, ensuring that the soil's moisture characteristics manage crop irrigation for optimal results. An Android app (SIMS) was also developed for the farmer, showing the status of their agricultural land.

Table 2 shows the cost of a developed automated irrigation monitoring system with five modules: a controller unit and sensor nodes, a solar charging module, an IoT gateway, a soil moisture sensor, and a soil temperature module. An automatic irrigation monitoring system was estimated to cost $239, including annual maintenance costs. The proposed WSN irrigation monitoring system was relatively low cost compared with other existing systems. For example, Jamroen et al. (2020) designed a prototype automated irrigation model with a cost of about $288.98. Khriji et al. (2014) developed the soil moisture nodes, container nodes, base station, and weather nodes for precise irrigation based on a WSN, and the total cost was observed to be $388.95. Nawandar & Satpute (2019) developed an IoT-based, low-cost, and intelligent module for smart irrigation utilizing three modules: a unified sensor pole, a low-cost and intelligent IoT-based module, an irrigation unit, and a sensor information unit; the overall cost of the system was $800–$1,000. Furthermore, Sahu & Behera (2015) designed a low-cost smart irrigation control system for Indian farmers. The overall cost of hardware and software components was estimated at approximately $237.90 (without annual maintenance cost).

This article introduces the smart agriculture model designing and deployment of a complete wireless sensor networking system for real-time monitoring of soil moisture and temperature parameters. Additionally, the performance of the entire designed system (WSN–Edge–Cloud) through a real-field experiment has been evaluated. The experiment's findings validate the viability of the design by showing that the iNODE can move between many modes to reduce power consumption. Furthermore, integrating weather forecast data and ET with soil monitoring can provide valuable insights for optimizing irrigation practices in agriculture, ensuring efficient water use, and maintaining optimal soil conditions for crop growth. The specific conclusions drawn from this study are as follows:

• The context-adaptive iNODE automatically modifies its data acquisition frequency in irrigation fields following the soil water content. It collects all the necessary data while simultaneously maintaining optimal power utilization.

• Along with its design, iNODE uses less energy when sleeping, i.e., 0.011 mA at a voltage of 4.2 V in standby mode.

• These nodes ensure independence in a power-constrained environment using energy collection and a backup strategy. Additionally, this makes charging batteries at night in severe weather easier. Due to these features, iNODE is a self-contained, durable gadget that can survive harsh external conditions and guarantee long-term, accurate irrigation monitoring.

• The designed WSN system plays a pivotal role in addressing water conservation challenges, outperforming the previously used automatic timer-based irrigation system in terms of efficiency. Moreover, the developed system proves to be a cost-effective solution in terms of time and financial resources, and it boasts an impressive 98.56% accuracy in delivering real-time data. The experimental results demonstrate that the SIMS Android application enhances efficient water management for diverse crop types and improves network performance and system functionalities compared to existing systems.

• The findings also indicate that the current design outperforms previously developed sensors, such as the PR2 probe and TDR, when deployed at the same depth. Additionally, capacitive-based soil moisture sensors were utilized to regulate soil moisture levels effectively and have the potential to automate irrigation systems.

Although the designed system has certain limitations, this study offers insightful information about an economical smart irrigation monitoring system (SIMS) for agricultural water management. The findings do not apply to extreme weather conditions, e.g., prolonged droughts or heavy rainfall, which could compromise the dependability of the system's coordinator nodes, sensor nodes, and circuit designs. Furthermore, the system's performance accuracy could vary under different environmental conditions, particularly due to sensor sensitivity and calibration issues. Variations in soil type, temperature, and humidity might also influence sensor readings and system responses, potentially impacting the precision of irrigation scheduling. Future research should focus on refining the system for extreme weather conditions and improving sensor robustness to ensure broader applicability and reliability in diverse environments. Additionally, machine learning algorithms could enhance an IoT-based SIMS for soil moisture prediction and be tested under extreme weather conditions. It should be evaluated across crops such as rice and sugarcane to broaden its applicability.

The authors sincerely thank the Department of Water Resources Development and Management (WRD&M) and the Electronics & Communication Engineering (ECE), IIT Roorkee, for generously providing the necessary facilities for conducting this study. The authors sincerely thank the National Water Mission (NWM), Department of Water Resources, RD&GR, Ministry of Jal Sakti, Government of India for their financial support of this study.

Y.K.K. and A.J. were involved in the literature survey, concepts, formal analysis, investigation, writing original drafts, writing reviews, and editing. R.K.P. and A.P. did conceptualization, supervision, and writing review and editing.

The research was supported by the National Water Mission (NWM) Scheme, Department of Water Resources, RD&GR, the Ministry of Jal Sakti, Government of India in this study.

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.