Design and Remote Monitoring of a wireless-Controlled Smart Agricultural Greenhouse

. This article presents a wirelessly controlled smart agriculture system and offers a comprehensive literature review to explore advancements in greenhouse monitoring and control systems. The system utilizes various sensors and actuators to optimize climatic conditions and enhance the growth of Dianthus Caryophyllus (CLOVE), an aromatic plant. Developed using Internet of Things (IoT) technology, the smart greenhouse system continuously monitors crucial environmental variables, including temperature, humidity, CO2 levels, and soil moisture. The collected data is utilized by actuators to make real-time adjustments to the greenhouse environment. The prototype operates through wireless communication protocols accessible via a dedicated mobile application.


Introduction
Green Revolution has provided farmers with the opportunity to learn about and use scientific agricultural methods, reducing the need for human labor and embracing technology [1].However, farmers still facing challenges in maintaining an optimal balance of plant nutrients, appropriate temperature and humidity levels, effective soil moisture control, and disease prevention.In light of growing worries about resource depletion, climate change, and food security, greenhouse farming is poised to become a key component of sustainable agriculture in the next years due to its reliance on data that improve the accuracy of agricultural processes [1].
The Agricultural greenhouses are closed structures with covers of translucent material that provide partial or complete protection to crops [2], [3] and enable their cultivation under optimal environmental conditions.According to research, greenhouse farming offers several advantages over open-air farming, such as higher yields, reduced water spoilage, increased output, and better pest and disease management.Research conducted in Kenya [4] discovered that greenhouse farming may minimize water consumption by 50-90% when compared to traditional farming.Moreover, combining renewable energy sources with human resources and innovation can significantly decrease greenhouse gas emissions in greenhouse operations [5].Additionally, crops grown indoors are less susceptible to environmental impacts [6].
In the following sections of the paper, a literature overview on greenhouse monitoring and supervision is presented, as well as a conceptual analysis of a smart agricultural greenhouse system.Analysing the greenhouse's unique requirements, appropriate technological solutions are suggested.The design includes several parts and systems, such as sensors for tracking environmental variables, actuators for fine-tuning, and a central control system that takes instructions from the mobile app.These components work together to create a remotely controlled smart agricultural greenhouse prototype, facilitating efficient monitoring and management of the greenhouse environment.Controlling the greenhouse microclimate is a crucial step in maintaining optimal conditions during crop growth [7].The microclimate in the greenhouse is multiparametric and nonlinear [2], requiring a solid understanding of both internal factors, such as greenhouse dimensions and components, as well as external factors like weather conditions, temperature, humidity, and wind speed.
With the help of Internet of Things (IoT) technology, farmers can remotely monitor greenhouses and automatically regulate the climate data by utilizing various actuators.This enables them to receive more timely information and make necessary adjustments.IoT has been integrated to enable farmers to make real-time decisions and actions, minimizing supply losses [8].
The literature review forms the cornerstone of our research, providing readers with insights into the existing advancements in the smart greenhouse sector.The developed mobile application serves as a user-friendly interface, enabling farmers to access real-time data and efficiently control the greenhouse environment.This innovation not only enhances agricultural efficiency but also contributes to the conservation of natural resources and ensures food security.
The paper is organized as follows: the second section presents a comprehensive overview of recent advancements and research in the field of smart greenhouse systems, the design of the suggested greenhouse is then explained in the following subsections and finally, simulation and results are presented.

History of Greenhouse Farming
Agricultural greenhouses have been utilized since ancient Rome, when early greenhouses were used to cultivate exotic fruits and vegetables, as well as prevent food from freezing in the 17th century [9].Over the years, greenhouses have undergone significant evolution, expanding their use to include flowers, vegetables, and even exotic foods.In the modern era, modern agricultural greenhouses have become incredibly efficient, by employing cuttingedge technologies and climate control systems to boost crop growth and yield.The agricultural sector has evolved beyond traditional methods and become more innovative and adaptive, driven by the need to address climate change adaptations and advancements in agricultural practices [10].
The traditional greenhouse suffers from several limitations in terms of supervision and control [11].It is primarily designed for year-round crop cultivation, relying heavily on the grower's assessment of crop parameters and other growing conditions [12].Manual plant inspections are also commonly employed, which consume significant time and labor resources [13].Moreover, conventional greenhouse farming is labor-intensive and timeconsuming, increasing the risk of illness and resulting in inadequate nutrition for plants.These challenges also make it difficult to accurately monitor climatic conditions within the E3S Web of Conferences 469, 00038 (2023) ICEGC'2023 https://doi.org/10.1051/e3sconf/202346900038greenhouse [14].To overcome these limitations, researchers emphasize the importance of enhanced data sharing, reliable methods for sharing greenhouse data, data interoperability to enable communication between different systems and devices, and remote monitoring.These advancements are essential to address existing issues and pave the way for successful smart greenhouse agriculture [13].
The aim of smart farming greenhouses is to integrate analytics and sensors into farming methods [6], [14].By incorporating IoT, farmers can make real-time decisions and take actions to minimize supply losses [8] while ensuring a suitable climate for optimal plant growth and health [6].This integration facilitates the exchange of information across sensors, data streams, processes, and web-based services, ultimately leading to the development of cost-effective and reliable farming solutions that enhance productivity with minimal labor requirements [12].The data collected by sensors plays a crucial role in the development of more robust models for predicting greenhouse parameters, as well as performing tasks such as plant disease detection, weed management, and plant counting [2].

Greenhouse Monitoring and Control System
In the field of greenhouse monitoring and supervision, several studies have examined the effectiveness of smart greenhouses compared to traditional ones.
For instance, [11] compared the agronomic and quality metrics of tomato plants grown in a conventional greenhouse to those grown in a smart greenhouse to verify the efficiency and efficacy of the smart greenhouse system in supplying fresh and high-quality tomato products.The smart greenhouse was built using a Raspberry Pi microcontroller to monitor the impact of temperature and humidity changes on tomato growth.The study revealed that the smart greenhouse system yielded 10-15% better results compared to the traditional greenhouse.Similarly, [15] established that an Internet of Things (IoT)-based system might be a versatile and appropriate option for monitoring and regulating the greenhouse environment.They presented a sensor-based system for monitoring temperature, humidity, and soil moisture levels.
[8] introduced a technology platform based on IoT and convolutional neural networks to analyze physical variables and detect plant diseases for precision management in agriculture.The platform comprises processing and storage, gateways, sensors, and camera nodes.Sensor nodes wirelessly exchange data from sensors and collect data from sensor nodes, while camera nodes deliver pictures to the gateway node, which stores a CNN model for detecting plant diseases.The platform is connected to the internet by the gateway node, which also sends data to the cloud server.Farmers can access and manage the platform remotely using a web and mobile application.To ensure communication with sensors, cameras, and the cloud service, ZigBee, Wi-Fi, and a cellular interface were employed.The CNN model analyzes color photos of plant leaves to identify diseases and presents the image with a label describing the plant's health.The CNN model demonstrated its effectiveness with an average accuracy of 95%.
[1] underline the potential of IoT and fog computing in improving astute agriculture practices.They explore the emerging computing technology of fog computing, which supports cloud computing and enhances the quality of service of IoT devices.This technology offers more possibilities to increase automation in greenhouse farming through three systems: the Data Acquisition System (DAS) for collecting sensor data, the Decision Support System (DSS) as the operating system that supervises and oversees all activities, and the Central Actuator Manager (CAM) for executing tasks using actuators.The authors also suggest the utilisation of machine learning for environmentally friendly management practices.
In their review, [10] discussed the application of smart technology, the internet of Things (IoT), and data recording in agricultural production systems .They explored innovative E3S Web of Conferences 469, 00038 (2023) ICEGC'2023 https://doi.org/10.1051/e3sconf/202346900038techniques to combat climate change and maintain sustainable crops, such as regression model [16], neural network [17], [18], fuzzy logic [19], [20], and linear regression [21] for soil moisture, seeding, yield predictions, and irrigation management.Image processing techniques and genetic algorithms were also discussed for smart disease management, achieving precision accuracy up to 97.2% [22].[13] emphasize that existing smart greenhouse farming encounters a number of issues, such as data sharing, data interoperability, data processing, and remote monitoring.To address these challenges, the authors propose utilizing blockchain technology as a secure and distributed platform to create a smart contract system.This system aims to maximize production management, plan an energy-efficient greenhouse, and improve overall crop production.The results demonstrate that the proposed strategy saves 19% more energy than the prediction-based approach and 41% more energy than the baseline scheme.
[23] proposed a four-layer architectural system for monitoring greenhouse microclimate and disease risk indicators.The research aimed to develop and deploy a scalable, costeffective IoT-based monitoring system for protected cultivations.They highlighted Sigfox as a major LPWAN technology for IoT deployments, citing its affordability, ease of use, and independence.The Sigfox network architecture was utilized to connect IoT devices and transmit data over the Sigfox protocol to the internet.The system was developed using machine learning techniques for environmental prediction, and its performance was experimentally validated by monitoring temperature and humidity in a tomato crop greenhouse in Mexico for over six months.The system achieved an availability percentage of 92%.The study conducted by [24] focused on predicting and managing temperature and humidity in greenhouses.In this study, the authors developed two ANFIS models to forecast the behavior of internal temperature and internal relative humidity based on historical data.The findings emphasize ANFIS as a valuable tool for improving climate control and optimizing greenhouse conditions.To summarize, these studies show that smart greenhouse technology, internet of things integration, and data-driven techniques can improve agricultural practices and output (table1).

System design
The main aim of the project is to remotely control and regulate greenhouse climate parameters using a wireless network (Wifi), record and display the parameters remotely, and configure regulation set points using a mobile application adapted to the type of plants grown.

Selection of plant
Carnation (Dianthus caryophyllus), also known as grenadine or wallflower rose [25], is the plant used.It's a versatile, low-maintenance plant that's easy to grow, thrives in warm, sunny environments, and is drought-resistant.

System Overview and Block Diagram
The block diagram for the proposed controlled smart greenhouse is shown in Figure 1.The system is based on Arduino Mega and Node MCU (ESP8266).The architecture integrates various sensors for monitoring environmental variables, including temperature, humidity, soil moisture, CO2, and light intensity.And employs several actuators to adjust environmental parameters in response to sensor data.The Arduino Mega serves as the core control device.It is linked to the different sensors.The Node MCU ESP8266 module serves as the link between the Internet and the smart greenhouse, facilitating communication and data transfer to the Arduino IoT platform's cloud-based architecture, which provides real-time data visualization through dashboards and graphs accessibles via a dedicated mobile application.

Sensors and Actuators
The Dianthus thrives in full sunlight and prefers a warm, humid atmosphere with temperatures ranging between 20°C and 24°C [26].To achieve these conditions, a DHT22 sensor monitors air temperature and humidity, while an LDR (Light Detector Resistor) sensor ensures precise lighting adjustments.Additionally, we incorporate an MQ135 sensor to measure CO2 levels and a YL69 soil hygrometer for soil moisture detection.
Through careful programming, the system intelligently responds to environmental fluctuations.If the temperature falls below 20°C, the heating system activates, and if it exceeds 24°C, the extractor is initiated to maintain the ideal range.When the humidity drops below 60%, the fog system is activated automatically or manually to raise humidity levels.If soil moisture drops below 80%, a pump efficiently irrigates the plants through garden hoses, and if CO2 levels rise above 800 ppm, the motor opens the greenhouse roof, maximizing air circulation.

Communication
In terms of connectivity, Arduino IoT Cloud supports the MQTT protocol, which offers reliable communication between IoT devices.It operates on the publish-subscribe principle for machine-to-machine (M2M) communications.
The Arduino Mega represents the microcontroller used to control and monitor the smart greenhouse through various sensors.The ESP8266 (NodeMCU) acts as a wireless communication link to enable the Arduino Mega to connect to the MQTT Broker and the Mobile Application.MQTT Broker serves as a centralized cloud server that enables data exchange between Arduino Mega, ESP8266 (NodeMCU), and the Arduino IoT Cloud.The cloud platform stores data received from the MQTT Broker.It enables the user to monitor and control the agricultural greenhouse remotely via a dedicated web application (see figure 2).The mobile app provides advanced visualization features, allowing users to display trends and variations in key parameters such as temperature, humidity, brightness, CO2, etc (see figure 5).Graphs offer a quick and easy visual understanding of the greenhouse's environmental conditions evolution.

Fig. 5. Real-time Climate Data Supervision
The user can program setpoints based on desired environmental parameters, such as temperature range, optimum humidity, and ideal CO2 level.The greenhouse system will then use these setpoints to automatically adjust actuators according to sensor measurements.For example, if the temperature exceeds a predefined threshold, the heating or cooling system will be activated automatically to restore the desired conditions.In addition, manual control functionality allows growers to quickly adjust parameters to the specific needs of each type of plant, particularly in response to sudden changes or special situations (see figure 6).
E3S Web of Conferences 469, 00038 (2023) ICEGC'2023 https://doi.org/10.1051/e3sconf/202346900038Fig. 6.Arduino IOT Cloud remote Our system's ability to optimize critical climatic factors, these adjustments, in turn, had a considerable impact on the growth of Dianthus Caryophyllus and, by extension, the greenhouse's total production.Real-time adjustments minimize water and energy waste.However, our system does not account for the intricate and non-linear changes that can occur in a greenhouse environment based on sensor data.For precise adaptation, it is imperative to integrate more sophisticated models, like machine learning algorithms, to model extreme weather fluctuations and interactions between different climatic parameters.

Conclusion
By combining various sensors and actuators, the wirelessly controlled smart agricultural greenhouse presented in this article demonstrates the transformative potential of IoT technology and data-driven techniques.Remote control of crucial climatic variables within the greenhouse not only optimizes crop growth but also contributes substantially to sustainable agricultural practices.The comprehensive literature review forms the backbone of this study, providing a robust foundation for understanding existing advancements in smart greenhouses.Looking forward, future work in this field should explore more advanced data analysis and machine learning algorithms to optimize greenhouse management and predict its evolution.

Table 1 .
Modern techniques for smart greenhouse.