Smart Technology-based Microclimate Control

System for Greenhouse

Seddiki Noureddine Mokeddem Kamal Abdelmadjid

dept. of math and computer sciences, lab. L.T.I.T dept. of math and computer sciences, lab. L.T.I.T
university of tahri Mohamed university of tahri mohamed
Bechar, Algeria Bechar, Algeria
[email protected] [email protected]

Abstract— Despite the intricate nature of the microclimate any mobile or stationary device with a secure internet
in the greenhouse, AI is now accessible and helpful to the connection by using wireless sensors and IoT-enabled
farmer. In order to simulate AI for food sustainability and the equipment [3].
hazards that people experience, it is necessary to develop models
and technologies that enable data collection. A microcontroller In this work, we present an experimentation-based
and sensors for various variables, such as temperature, automation system for greenhouses that was created in our lab.
humidity, and soil moisture, manage the microclimate. An Additionally, we present a mobile application and web-based
effective platform for gathering and analyzing this data is the backend architecture that make use of the sensor data gathered
Internet of Things. Web-based and mobile apps provide on the greenhouse microclimate to enable automation and
interactive, visible features that might have an effect on the control. We've broken the document into a number of sections
greenhouse environment. Based on user input or the controller's to accomplish this. We highlight pertinent research and
ideal threshold settings, the fan or irrigation pump may turn on technological advancements pertaining to the greenhouse
in real-time. In this paper, we propose a control system for microclimate in Section 2. We outline the greenhouse
increasing agricultural yield that will also provide a dataset parameters that are used to regulate the microclimate in
demonstrating the performance of ML models in the future. Section 3. A thorough overview of the many technologies
utilized to build the prototype and platform is given in Section
Keywords—greenhouse, control, microclimate, IoT.
4. The results are then discussed in Section 5. Finally, in
I. INTRODUCTION Section 6, we emphasize the unrealized potential of our
forward-looking ideas for this dataset and summarize the
The state-of-the-art greenhouses are designed to give findings of this work.
plants shelter from the elements outside and a climate that is
favorable for intensive food production. However, because of II. RELATED WORKS
the swift shift to precision agriculture and the ensuing increase
Recent research in precision agriculture emphasizes the
in food security while protecting natural resources,
significance of managing and monitoring the microclimate
greenhouses are undergoing a fundamental transformation.
within and outside greenhouses. The techniques involved in
This is a result of enhanced metering, communication, control,
controlling the environment include gathering, analyzing, and
and monitoring systems [1].
processing data, while also visually identifying climatic
Agriculture in closed-field conditions is now more parameters to alert of potential issues. All these measures are
sustainable because of the management and automation of aimed at increasing agricultural productivity and reducing
microclimate control in greenhouses. This is accomplished by input costs. The importance of contemporary technologies
using fewer resources, such as energy, water, and fertilizer such as wireless sensor nodes, embedded devices, and IoT-
while boosting output and profitability [2]. The microclimate, based and cloud-based data collection platforms in achieving
sometimes referred to as the optimal growing environment, is these goals cannot be overstated.
impacted by elements including temperature, humidity, light
The use of novel technology such as IoT in the domain can
intensity, and air movement. An intelligent greenhouse can
be seen in [4]. In this paper, the authors use a model-based
produce the ideal environment for plant development by
implementation in Matlab Simulink to analyze a greenhouse
precisely regulating and managing these variables.
microclimate utilizing IoT, sensors, and Node MCU
Monitoring and modifying environmental parameters are part
(ESP8266) as controller. The goal of this study is to better
of the microclimate control process in order to provide plants
understand how to interpret microclimate data variances about
the ideal habitat. This can be done manually or by utilizing
plant needs at various growth stages. The model-based
cutting-edge technology including sensors, automated
analysis offered a way to accept these uncertainties due to the
systems, and computer software.
nonlinear dynamics of the system (i.e., solar radiation and
The highest greenhouse agricultural yield and production wind speed) and reduce the production risk despite even
levels to date have been made possible by recent though the precise behavior of the greenhouse environment
advancements in sensor technology. Thanks to digital from raw data and its impact on the plants.
technologies like the Internet of Things (IoT), automation
The authors of [5] designed and implement a remote
professionals may now take benefit of customised solutions
monitoring and control system for the greenhouse based on
created especially for greenhouse applications. Greenhouse
IoT. The greenhouse model is divided into four sections for
conditions may be monitored and managed in real time from
planting various plants. Different levels of soil moisture are
maintained in each area. This is used as the basis for III. PARAMETERS OF MICROCLIMATE IN GREENHOUSE
monitoring and analyzing plant development. The To create the best climatic conditions for crop
microclimatic conditions in the greenhouse model are development while using the least amount of fuel and energy
controlled by an automated control greenhouse system, which possible, a variety of factors are used both inside and outside
is made up of various separate subsystems (watering, lighting, the greenhouse. Solar radiation, air temperature, relative
temperature control, air quality subsystem, etc.). The humidity, air flow rate, and carbon dioxide concentration are
greenhouse IoT component uses the ThingSpeak cloud. a few of these variables. Nevertheless, monitoring and climate
In [6], the objective was to gather information on air control are made more difficult by the interaction of the
temperature, air humidity, and light levels using a group variables that affect one another and by their dependency on
control device (ESP) in a greenhouse microclimate. To the shifting ambient environment.
achieve this, an automated microclimate management system
A. Temperature
was developed using cloud architecture. This system
comprises six fundamental components: a cloud server, a local Temperature's main function is to ensure that crop leaves
object server, a mobile application, a group control device, a develop at a young age. Nelson states in [10] that while
device module, and a sensor. The greenhouse complex's greenhouse plants do best at temperatures between 15 and 30
dependability is increased, and the possibility of data loss due degrees Celsius, most plants grow well at temps between 10
to network failures is decreased, by replicating data from a and 24 degrees. Nelson [10] found that the temperature
local device to the cloud. Based on the test results, the user difference needed for crop development between day and
interface's reaction time can be deemed acceptable when it night should be around 8 to 10 °C under sunny situations.
takes less than two seconds from the time that data packets are Depending on the type of plant, different temperatures are
received until they are seen on the user's screen. Data required for optimal plant development.
availability and system dependability can be increased by B. Humidity
duplicating functionality and replicating databases across on-
premises object servers and cloud services. The appropriate humidity levels for a plant are strongly
dependent on water stressors, harsh weather conditions, the
The system designed here [7] is extremely complex, risk of pestilence/insects, ripeness, and the crop growth phase
cutting-edge, and capable of measuring six microclimatic [11, 12]. As a consequence, optimal humidity levels are
parameters in three separate greenhouse sections. A decision- required for plant development, which is expressed in terms
making control system can be activated to adjust required of the vapor pressure deficit, which is defined as the difference
parameter values by providing ventilation or by turning on and between the water vapor pressure at saturation and the actual
off processes of mist foggers in required areas of the water vapor pressure at the greenhouse temperature [11].
greenhouse after comparing the collected data with expected
values. In addition to this, the system continuously records to C. Air flowrate
create databases. The system presented demonstrates its utility Poor air circulation inhibits plant activity and might cause
for measuring and revealing distinctions in the greenhouse humidity and disease control issues. The greenhouse's air
microclimate in terms of the temperature, humidity, and soil circulation should be between 0.2 and 0.7 meters per second.
moisture parameter monitored. Plant development is hampered if carbon dioxide levels are
not maintained. The exchange of air is what ventilation is all
The primary contribution of this study [8] is the about. Moving the air breaks down this "stratified" air,
development of two-way data transmission based on the reducing heating costs and keeping your plants comfy.
Internet of Things (IoT) utilizing Firebase Realtime Database
as the platform for handling and tracking the environment of D. Light intensity
the smart greenhouse. Sensors and micro-controller are This is the point at which photosynthesis is at its peak and
present as smart nodes, which may broadcast microclimate plant growth is at its peak. Growth is slowed when the amount
readings to a distant firebase to store the dataset. Users will be of light is lowered. Light saturation is the point at which an
able to remotely operate the actuators in the smart greenhouse increase in light intensity no longer increases photosynthesis.
using the data that is provided on the Laravel-based website
interface. By measuring the latency and packet loss data E. Carbon dioxide concentration
values, the architecture of a smart greenhouse-based two-way The advantage of CO2 enrichment is mostly seen in
data communication system was analyzed and tested. increased crop yield due to improved photosynthesis. CO2
The main goal of the study [9] is to create an IoT-based enrichment in the greenhouse is an important characteristic
system capable of producing temperature, humidity, and soil because it has a beneficial influence on crop development,
moisture sensors that are location-specific microclimatic provided that other growth parameters, such as water supply,
characteristics. With the aid of the Cropwat program, which is are satisfied. CO2 levels should be provided to greenhouse
based on macroclimatic parameters, the data is further crops to compensate for the significant drop in CO2 caused by
examined and verified. By tracking various microclimatic photosynthesis, especially when adequate ventilation is not
indicators, farmers can assess their irrigation water available.
requirements. This involves using sensors to measure soil F. Solar radiation
moisture and temperature, which are then fed into the LoRA
The first and most important meteorological characteristic
system. The system analyzes this data to estimate
for determining a region's eligibility for protected cropping is
evapotranspiration. The Mcguinnes-Bordne formulation is
solar radiation [13]. The primary source of heat gain is direct
used to evaluate the sensor data, and the results of this study
solar radiation intercepted in the greenhouse, and it
effort open the way for the estimate of evapotranspiration in
contributes the most to the increase in the daytime temperature
the microclimate setting.
of the protected cropping environment. Furthermore,
significant amounts of solar energy are collected and stored in expressive and clean syntax, the API resources are specified
the soil before being released at night [14]. and set up.
IV. METHOD AND MATERIALS Since the server is a basic device located in the lab with a
public IP address. The server has a dual-core CPU, 2 Go of
We utilized advanced technologies to gather crucial Memory, and 500 Go of hard drive space. We would rather
indicators for greenhouse growth, such as temperature and install a Linux operating system, and we choose Debian 11,
humidity levels inside and outside the greenhouse, soil which is set up as a web server by installing apache2 and
moisture, light intensity, and airflow. These parameters were phpMyAdmin to build the database.
collected to develop comprehensive databases. Furthermore,
we employed remote operation techniques in compliance with The “Fig. 2”, illustrates the greenhouse model which is
recommended procedures and effective technology. This built of wood and wrapped in translucent plastic foil. The
allowed us to remotely operate various devices, including seedling starter in the center has sensors for measuring the soil
fans, lighting, and irrigation pumps. The primary objective of moisture humidity. The irrigation system consists of a pump,
this study is to optimize the variables that create an ideal a water level sensor, and two tanks. Moreover, there are three
environment for plants to thrive in a greenhouse. We fans, where two create air and the other withdraws it. The light
evaluated three control subsystems for these variables and subsystem is made up of LEDs designed specifically for
provided recommendations for their potential application in seedling growth and LDR sensors to gauge light intensity.
an automated or autonomous sustainable greenhouse farming
system.
Users can choose between an Android application or a
Web application, both of which provide access to the database
through a REST API in a Laravel project. The data is delivered
and received in JSON (JavaScript Object Notation) format.
Additionally, the app includes a user interface that allows for
controlling the actuators.
The EPS32 controller, connected to the Wi-Fi network,
sends the measured parameters to the local server every five
seconds. More detailed design information regarding the built
IoT-based data transmission system can be found in “Fig. 1”.
Fig. 2. The modal of the greenhouse constructed.

The automated control system is intended to regulate the

greenhouse model's microclimate environment. As depicted
previously, the model is made up of several autonomous
subsystems, including temperature control, irrigation,
lighting, remote monitoring and control, and IoT subsystem.
The functional wiring diagram for all components of the
greenhouse model is shown in “Fig. 3”.

Fig. 1. The constructed IoT-based data transfer infrastructure.

The system utilizes an ESP-WROOM-32 controller,

which is a dual-core 32-bit microcontroller with built-in Wi-
Fi (802.11/b/g/n/e/i) and Bluetooth capabilities. The controller
is programmed using the Arduino IDE through an ESP32 Dev
Module add-on. It is configured as a client access point,
allowing it to obtain an IP address from the gateway, connect
to the Wi-Fi network, collect the sensed parameters, and
transmit them to the server for storage.
The information sent by the gateway to the API address
serves to specify the connection for the MySQL database's
data storage. The information is then used on a mobile Fig. 3. Wiring diagram of the hardware used in the model.
application interface built using Android Studio and a
website-based interface created with Laravel as front-end and
back-end. Using a web application framework (Laravel) with
 The watering subsystem in “Fig. 4” is implemented with a of the software or hardware compatibility issues and can be
water pump submerged in a large tank beneath the greenhouse supported by any device available to users. For example, we
to collect excess water droplets, which has a level sensor to use the Java programming language in Android Studio to
detect when there is insufficient water and alert the user. A create mobile applications, and JavaScript on the web to
second tank is placed at a steeper angle to ensure that the access this JSON format.
automatic micro-irrigation (drip) will be filled by the needs of
plants as determined by the tree soil moisture planted in a
different pot. The slope will automatically give watering to the
crops when the second tank is full or as needed. Experience
has taught us that it takes a certain amount of time to fill the
tank since the pump must be stopped so that there isn't an
excess of water.

Fig. 4. The structural irrigation subsystem.

When conditions are darker than during the day, the

lighting subsystem is used. That is made possible using a
special LED meant to promote rapid development and larger
yields (ultra-purple light due to the short wavelength), linked
to the controller, and it will turn on when the intensity sensor
notices that there is no daylight left. To give greenhouse plants
the appropriate light during the day, the user can interact with
the interface of the light subsystem at night if there is a need
for light.
A group of fans was utilized to create another subsystem
with a DHT22 and MQ2 gas sensor to regulate the indoor
Fig. 5. The mobile and Web-based interface of the application.
temperature and monitor the quality of the air. Cooling fans
are activated to pull or generate air when the temperature, Our main objective is to help farmers achieve higher crop
humidity, or air quality is below or upper the average required yields while conserving energy and resources, especially
value. reducing water loss, by controlling the internal microclimate
V. RESULTS AND DISCUSSIONS using this technology. In addition, we aim to collect and
prepare the dataset for future research by integrating a deep-
The model performs well with the relatively easy learning algorithm to predict microclimate parameters.
interactive user interface of the mobile and web-based
applications shown in “Fig. 5”, where the user can find a VI. CONCLUSION
variety of make-up that aid them in keeping track of the The field of smart agriculture is anticipated to grow
greenhouse's internal microclimate and provide them with quickly as a result of the usage of digital tools and computer
some information about the condition of the plants in the form approaches. Our approach for monitoring and controlling
of sweet notifications, as seen in “Fig. 5”. If necessary, a different elements that have an impact on the microclimate
system can automatically perform a single action or a set of within a greenhouse has been successful in helping us gather
actions, such as turning on a pump for irrigation when the soil information to describe these elements. Future research in this
moisture senses a need for it or turning on a fan to remove air. area will, however, need to include further data validation.
For the dataset's development, it is quick and simple since The use of contemporary tools and technologies, such as
every 5 seconds, a new record of the humidity (internal and AI, cloud platforms, edge computing, and fog computing, has
external), temperature (internal and external), airflow, the become normal practice to save costs and time when
amount of water in the tank, and the status of actuators (on/ processing and organizing such massive volumes of data
off) are recorded. All of these measurements are time and date because traditional approaches are unfeasible.
stamped.
REFERENCES
The use of APIs and JSON format (Requests/Responses) [1] A. Ouammi, "Model predictive control for optimal energy management
demonstrates our application’s ability to co-use several of connected cluster of microgrids with net zero energy multi-
programming languages and platforms for the collection and greenhouses," Energy, vol. 234, p. 121274, 2021.
manipulation of data. Thus, our system is not limited by most
[2] R. R Shamshiri et al., "Advances in greenhouse automation and
controlled environment agriculture: A transition to plant factories and
urban agriculture," 2018.
[3] S. M.-e. Rezvani et al., "IoT-based sensor data fusion for determining
optimality degrees of microclimate parameters in commercial
greenhouse production of tomato," Sensors, vol. 20, no. 22, p. 6474,
2020.
[4] R. R. Shamshiri et al., "Model-based evaluation of greenhouse
microclimate using IoT-Sensor data fusion for energy efficient crop
production," Journal of Cleaner Production, vol. 263, p. 121303, 2020.
[5] U. Drakulić and E. Mujčić, "Remote monitoring and control system for
greenhouse based on IoT," in Advanced Technologies, Systems, and
Applications IV-Proceedings of the International Symposium on
Innovative and Interdisciplinary Applications of Advanced
Technologies (IAT 2019), 2020: Springer, pp. 481-495.
[6] O. Kachanova and D. Levonevskiy, "Cloud-Based Architecture and
Algorithms for Monitoring and Control of an Automated Greenhouse
Complex," in Data Science and Intelligent Systems: Proceedings of 5th
Computational Methods in Systems and Software 2021, Vol. 2, 2021:
Springer, pp. 910-921.
[7] M. M. Kolapkar and S. B. Sayyad, "Greenhouse Microclimate Study
for Humidity, Temperature and Soil Moisture Using Agricultural
Wireless Sensor Network System," in Recent Trends in Image
Processing and Pattern Recognition: Third International Conference,
RTIP2R 2020, Aurangabad, India, January 3–4, 2020, Revised
Selected Papers, Part II 3, 2021: Springer, pp. 278-289.
[8] A. Sofwan, F. R. Adhipratama, and K. Budiraharo, "Data
Communication Design Based on Internet of Things Architecture for
Smart Greenhouse Monitoring and Controlling System," in 2022 5th
International Conference on Information and Communications
Technology (ICOIACT), 2022: IEEE, pp. 205-209.
[9] T. E. Mathew, A. Sabu, S. Sengan, J. Sathiamoorthy, and A. Prasanth,
"Microclimate monitoring system for irrigation water optimization
using IoT," Measurement: Sensors, p. 100727, 2023.
[10] P. V. Nelson, Greenhouse Operation and Management ( Prentice Hall:
Hoboken). 2003.
[11] M. Amani, S. Foroushani, M. Sultan, and M. Bahrami,
"Comprehensive review on dehumidification strategies for agricultural
greenhouse applications," Applied Thermal Engineering, vol. 181, p.
115979, 2020.
[12] M. Sultan, T. Miyazaki, B. B. Saha, and S. Koyama, "Steady-state
investigation of water vapor adsorption for thermally driven adsorption
based greenhouse air-conditioning system," Renewable Energy, vol.
86, pp. 785-795, 2016.
[13] A. Papasolomontos, W. Baudoin, and N. Lutaladio, "1. Regional
Working Group on Greenhouse Crop Production in the Mediterranean
Region: History and development," Good Agricultural Practices for
greenhouse vegetable crops, vol. 46, no. 3, p. 1, 2013.
[14] M. Buchholz, "The new generation of greenhouses," Unlocking the
Potential of Protected Agriculture in the Countries of the Gulf
Cooperation Council—Saving Water and Improving Nutrition, pp. 97-
132, 2021.