Solar-smart hydroponics farming with IoT-

based AI controller with mobile app

2023 IEEE Sustainable Smart Lighting World Conference & Expo (LS18) | 979-8-3503-4699-2/23/$31.00 ©2023 IEEE | DOI: 10.1109/LS1858153.2023.10170491

1st Pallavi Khare 2nd Abhishek Koti 3rd Akhil Khare

ECE dept CSE dept. CSE dept..
MVSR Engineering College MVSR Engineering College Matrusri Engineering College
Hyderabad, India Hyderabad, India Hyderabad , India
[email protected] [email protected] [email protected]

Abstract— Only Hydroponics uses less water and other Climate change, soil erosion, drought, and ecological
resources than traditional soil-based farming. Hydroponics degradation make it more difficult and costly to produce
using mechanical aid. Controversies with hydroponics. In the sustenance. Not less. By 2050, according to a well-known
realm of gardening, hydroponics is king. Water, temperature, prediction, the global population will approach 9 billion. IoT
and pH are all controlled by it. Typical household current is agricultural applications and burgeoning technology hold
used. We seldom see a spike in electricity use. The investigation great potential. Researchers estimate that 75 million IoT
resulted in two potential courses of action. Hydroponics with devices will be utilised in agriculture by 2023, valuing the
environmental temperature control. Hydroponics solar power hydroponic system industry at $23.14 billion USD.
plant second. Light intensity, battery voltage, DC-to-AC
Connecting the actual world to computer systems, IoT enables
converter current, and solar panel voltage are all tracked by the
smart solar power plant. Server-side telemetry of water storage
remote sensing and control of physical objects. IoT connects
tank's temperature, acidity, depth, brightness, and power sensor-equipped devices [3,4]. Ethiopia, India, and others are
supply. In this research, a solar-powered hydroponic agriculturally dependent. Traditional agriculture is profitable
automation system was created, allowing for precise regulation but inhibits agricultural production. Threatening rural
of acidity/ alkalinity/ temperature/ water communities are climate change, pollution, soil erosion,
level/light/current/voltage. Hydroponics farming is difficult to urbanisation, and the loss of agricultural land, among others.
monitor because of the many parameters, nutrition, and plant Intelligent agriculture increases crop yields. New and
diagnosis systems. Recent technological advances help solve traditional agricultural techniques require technology to
these problems by adopting AI-based controlling algorithms in address the food problem. Hydroponics, vertical farming, and
agriculture. The work focuses on implementing a mobile polyhouses may be able to solve these problems. Technology
application integrated AI- smart hydroponics system with IoT. improves hydroponics.
The proposed smart hydroponic system with IoT consists of
three phases. II. PROBLEM STATEMENT
Existing Hydroponics systems facing difficulties such as
Keywords— Smart Hydroponics, AI-Based Controlling, Deep
seedling (wilting, dead roots), system blockage, infestation
learning, Nutrient Level Prediction, Plant Diseases
(algae, bug), and nutrient deficits[7]. Solar energy might
I. INTRODUCTION improve grid resilience. Two major traits may give excellent
solar help for toughness. Solar power generation may be
IoT in agriculture has the potential to transform the distributed. This means homeowners may use solar panels on
equation for humans and the entire planet [1]. Solar their land to generate electricity. Communities may create a
hydroponics uses nutrient-rich water to grow plants. This micro-grid using solar panels, storage, and other technologies
approach grows plants without soil. This sophisticated plant to power important infrastructure. Large power plants satisfy
growing method is used worldwide to grow flowers and crops. demand. Homes and businesses get electricity through power
Solar hydroponics would help the system succeed. This plant lines. Damage to the electrical distribution system might
growing technique is energy-efficient and environmentally create issues. Power may be generated near users and stored
friendly. Solar-powered hydroponics systems outperform in regional energy hubs in case the centralised system fails.
those that use LEDs or electricity. Solar-powered hydroponics Second, unlike diesel generators, which are typically used as
is made possible by the sun's nuclear fusion rays. Energy is emergency backup power, solar energy may be stored and
produced in abundance by large panel plates.It is the released without energy labour, boosting resilience. Solar and
photovoltaic effect that powers solar hydroponics. solar cells energy storage will eventually power homes and businesses at
are devices that harness the sun's rays to generate power.The night. Farmers may closely monitor hydroponic systems.
solar panels are tilted so that they face the sun. Separate They manage pH and nutrition. This ensures that plants get the
readings for temperature, humidity, pH, and luminosity are right nutrients. These methods reuse water in plants. Growers
provided by this instrument. This helps to keep fertiliser water can optimise crop productivity by controlling temperature and
from escaping.Hydroponic solar systems allow for the illumination inside. Solar-powered hydroponics feeds plants
cultivation of soil-less, nutrient-rich crops. The yield of crops via their roots. The solar-powered hydroponics system is
is increased by using this method. Electricity reliability is cutting-edge and efficient. Solar-powered hydroponics
enhanced via solar hydroponics.Hydroponic solar systems systems help grow more plants in less area.
provide plants with more oxygen at their roots and feed them
directly with liquid nutrients. The addition of these nutrients It might be challenging to get seedlings off to a good start
increases crop yields. when using hydroponics. Leaves dry and wilt as a plant loses
its vigour. Wilting may be brought either by overheating or a

979-8-3503-4699-2/23/$31.00 ©2023 IEEE

Authorized licensed use limited to: Chulalongkorn University provided by UniNet. Downloaded on September 06,2024 at 07:04:54 UTC from IEEE Xplore. Restrictions apply.
 lack of watering[8]. Roots may be killed by overwatering, lighting and sampling settings are analysed to estimate
high EC levels, or high water temperatures in dense soils. information. P. S. Chatterjee et al. created a smart agricultural
Roots may decay due to infections. Clogging is the most technology system using AI and the Internet of Things. AI
frequent problem with drip-style hydroponic systems. It's analyses the data to predict crops. In real time, this strategy
common for tubes to get clogged because to particulates in consumes electricity. I. L. Maldonado et al. created a
growth medium. Damage to crops is caused when water hydroponic automated plant growth analysis system from
circulation is disrupted by obstructions. Check the electrical germination to harvest [25]. ESP32 also controls several
conductivity, temperature, and pH of the feeding solution sensors and actuators. The LOTUS app now tracks humidity,
before making a diagnosis [9]. Autonomous, AI-powered temperature, and irrigation. This approach, however, has a
robots with a variety of hardware controllers and industrial considerable computational cost. The [4] authors set up a
robots are needed for hydroponic plant minoring. They were hybrid system with numerous parameters for monitoring,
unable to solve the problems because of a lack of multi-sensor including nutrient concentration, water pH, and water
monitoring. Agricultural yields are reduced by factors such as temperature. To further automate these parameters in
plant diseases and development, which affect both accordance with reference water levels generated by the
conventional farming and hydroponics. Diseases are nutrient film technique, a machine learning approach based
identified by farmers by personal examination, while on the K-Nearest Neighbor (KNN) algorithm is employed.
pathologists provide laboratory diagnoses. Time-consuming On the other hand, when compared to deep learning models,
and skill-intensive, conventional methods have always been its reliability and efficiency are quite low. Automated
the norm. Hydroponics efficiency might be improved with hydroponics systems are used. Control and monitoring are
early illness identification and control. incorporated. EC, pH, and microcontroller sensors make up
the monitoring system. These sensors monitor plant roots in
III. OBJECTIVE: nutrient-rich solutions. Plants need various nutrients. This
A. Design and implementation of user-friendly environment nutritious drink is certified for pH and electrical conductivity.
for farmers using Agri-Hydroponic application, which Electrical conductivity and pH may indicate nutritional
provides hybrid monitoring and controlling of content. The microcontroller checks sensor data to verify
hydroponics farm field whether enough food is being eaten. If EC and pH are
excessively high or low, pumps can correct them. These
B. Implementation of AI framework for predictive analysis pumps pump pH up, pH down, EC up, and water from their
of sensor data and alerting the end users (farmers). storage tanks. This system's microcontroller modulates pump
output based on sensor data. Monitoring and controlling
values will be sent and stored in real-time. Google's open-
Hardware System (Raspberry
Agri-Hydroponic source Fire Base hosts it.
Pi connected with all sensors,
Android application
M o d e o f o p e r a t io n

camera & actuators ) V. PROPOSED METHODOLOGY

This hydroponic farming system is an integrated system that
includes a mobile application, the Raspberry Pi computer,
and an environment for the Internet of Things. When farmers
IOT Cloud Server (take utilise programmes that are capable of manual and automated
sensors data from modes of operation, it is much simpler for them to monitor
Raspberry Pi and hybrid and exercise control over their fields and crops. This is
deep learning algorithm because the applications are compatible with both modes.
The data gathered by these sensors is sent to an Internet of
Things ecosystem that is hosted in the cloud using a hardware
Fig1 : Basic Block Digram solution that is based on a Raspberry Pi controller. A deep
IV. . LITERATURE REVIEW learning-based artificial intelligence system is deployed
across many cloud locations. This system is backed by
V. Bhatnagar et al. [14] wrote "sustainable development in distributed computing. This system does a continuous
agriculture: past and present scenario of Indian agriculture." analysis of sensor data as well as the condition of plant
The authors use the IoE, a good basis for IoT development. diseases and then sends appropriate notifications to farmers.
Upgraded soil sensors monitor farms. The suggested The last step, but perhaps not the least important, is for the
technique saves energy but is less efficient when computing farmer to operate his hydroponics farm field using the manual
component heat index to monitor crop development. "An mode. This ensures that the plants get their nutrients in the
efficient IoT-based smart farming system using machine quantities that the farmer has determined. In addition, when
learning algorithms" by N. G. Rezk et al. will be deployed in the system is set to operate in automated mode, the plants will
South Africa in 2021 for autonomous water management to get the appropriate amount of fertiliser based on a reference
decrease water wastage. This project involves sediment and standard. This occurs when the system is in operation.
water analysis. However, data transmission range and Because it makes the most of the water that is readily
parameter current usage decrease. "Internet of things-based accessible, the hydroponic system is efficient; as a result, the
smart farming monitoring system for bolting reduction in final product has a greater nutritional quality. This efficiency
onion farms" by Z. Khan, S. Ali, et al. [17] used also adds to the system's relatively modest influence on the
computational intelligence for wheat crop nitrogen surrounding environment. Because of this, it is now
prediction. Crop photos collected in real time under various conceivable to cultivate plants without making use of any soil

Authorized licensed use limited to: Chulalongkorn University provided by UniNet. Downloaded on September 06,2024 at 07:04:54 UTC from IEEE Xplore. Restrictions apply.
 at all, which represents a big advancement in the progression The PH level of the water, which shows the degree to which
of plant farming. The liquid is administered to the plant roots the water is acidic or basic, may also be determined with the
in such a manner that it makes a contribution to the use of a turbidity sensor if one is available. In addition, a soil
development of ionic and inorganic forms of sustenance. This moisture sensor may be replaced with an NPK sensor, which
is done in order to maximise the effectiveness of the is capable of measuring the concentrations of nitrogen,
treatment. To achieve this goal, it is necessary to make certain phosphorous, and potassium in liquids. The pH of the liquid
that the liquid is administered in the appropriate method. The will need to be measured in order to do this. In addition to this,
production of crops via the use of this method is not only hydrostatic pressure level sensors are used in order to achieve
more effective but also more efficient when compared to the this objective. Additionally, camera modules are used to take
agricultural approaches that have been used traditionally. images of the plants at the predetermined intervals. It is
When compared to traditional farming, this approach offers a possible to get images of the plants. After the Raspberry Pi has
few distinct benefits. completed gathering all of the sensor readings and
photographs, the data is sent to a cloud server, where a Deep
Camera Learning model is trained utilising the information that was
provided by the Raspberry Pi. By comparing the current
nutrient concentration to previously taught reference
Sun light sensor
concentrations, the Prediction-Deep Learning model is able to
Display unit
accurately identify plant nutrient deficiencies and anticipate
Air Temperature & the standard nutrient levels. This is accomplished via the
Drain Pump
Humidity Sensor process of "deep learning." Because of this, the model is able
Arduino / Fresh water Pump
to accurately predict the amounts of the standard nutrients.
Water Temperature
Sensor Raspberry Pi Because nutrient deficiencies in plants have the ability to
Cooler produce disease, it is crucial that outbreaks be discovered as
PH sensor
soon as they appear. This is because disease may be induced
Turbidity Sensor
by nutritional deficiencies. Because of this, the Classification-
Deep Learning Neural Network Model is used in order to
Soil moisture Sensor discriminate between the many plant diseases that may be seen
in photos. This is done in order to improve diagnostic
Solar panel
accuracy. After that, this model will transmit the data that it
has gathered to the Agri-Hydroponic app, where the farmer
Fig2 : Proposed system hardware Section will have the opportunity to choose the method of cultivation
that best suits their needs. The farmer has full control over the
A. Hardware Implementation:
functioning of his hydroponic farm when it is set to the manual
A large variety of sensors were integrated into the gadget mode, and he is the one who decides how much of each
before it was completed, and the Raspberry Pi was used as the nutrient the plants get. The mode of operation that is created
controller of the system. In this setting, a large number of by the environment of the deep learning neural network is
sensors are used in order to carry out an investigation on the used to drive the operation of the Raspberry Pi controllers,
features of hydroponic farming. This system regulates a range which then directs the operation of the output devices in turn.
of variables, including temperature, water level, sunlight, and This cycle continues until all output devices have been
systems for the disposal of wastewater and cooling systems. directed by the mode of operation that was produced by the
The word "fresh water" refers to water that has lost some of environment of the deep learning neural network. As a direct
its nutritious value, and this system also manages these result of this, the mode of operation makes it possible to
aspects. In addition, the cloud environment that is based on the exercise direct control over output devices while the mobile
Internet of Things is continually updated with the most recent application is being used. Motors and pumps are two examples
sensor information. This takes place in real time. In order to of the types of equipment that fall within this category. Both
carry out an investigation on the effects of solar radiation, a regular water and nutrition water are pushed from this area
grove sunlight sensor is now being used. The properties of sun using their own separate pumps. The plants continue to
rays, such as the quantity of infrared, ultraviolet, and visible receive both types of water until the required equilibrium of
light, are produced as a result of this process. We are able to minerals and nutrients is attained. A heater output device is
calculate the rate of photosynthesis that is going place inside utilised so that the temperature of the water and air inside the
the leaf by using the information that is supplied by the hydroponics system may be better regulated and maintained.
properties of these sensors. The SHT-20 sensor is an This allows for more precise temperature regulation. The solar
imperative must for any effort that requires the measurement panel that is integrated into the system is the component that
of environmental factors such as temperature and humidity. Its is in charge of the charging process for the system's batteries.
use may be found in a wide variety of applications. Because
water is the only medium via which nutrients may be delivered B. Software Approach :
in hydroponics farming, it is vital to constantly monitor the Both types of deep learning models are stored in the
mineral content of the water in real time. Because of this, the cloud, which is also where the system retrieves them and
physical architecture of a hydroponic system has to consist of makes use of them. These models provide the basis for how
a much greater number of water sensors. In order to get a the system functions. Both "Prediction based deep learning"
precise measurement of the temperature of the water, a probe and "Classification based deep learning" are names for similar
sensor that may be completely submerged in the water is used. models. In this case, we use a deep learning model trained on

Authorized licensed use limited to: Chulalongkorn University provided by UniNet. Downloaded on September 06,2024 at 07:04:54 UTC from IEEE Xplore. Restrictions apply.
 predictions to determine the appropriate nutrient levels based Fig 3: Complete Hardware Setup
on reference values. Because of this, we can make better
predictions. In addition, a classification-based deep learning
model is used to differentiate between the various plant
diseases. In order to provide the farmer with the most up-to-
date information possible, both models were run
simultaneously and the results were communicated to him
through his mobile device. In order to execute classification-
based Deep Learning, an input feature matrix is constructed
by merging the sensor data together. In the hydroponics
industry, the collected data is initially analysed using a
number of distinguishing features. This information is then
forwarded to the cloud, where IoT services are hosted, where
it is further analysed. The controller is running on a Raspberry
Pi.
Fig 4: Dashboard to monitor all the sensor values
Deep learning models are then used to create predictions Accuracy Recall Precision
based on the data collected by these sensors. The correct Method
(in %) (in %) (in %)
F-measure
nutrient concentrations may be predicted for a variety of EFIS 89.8 90.51 89.52 90.66
sensor readings when a Deep Learning model is trained using Support
the reference nutrient dataset. This opens up a wide range of Vector 92.8 93.11 89.11 90.234
potential uses for the model. This is why the model is so Machine
useful. Due to the many conditions under which the gathered Mic-Conv
93.26 94.52 91.88 92.69
Net
data was generated, a deep learning model trained on the task
R-CNN 94.21 95.34 92.66 94.67
of predicting new nutrient levels may do so with uncanny Prediction
precision. based deep 98.23 97.554 98.12 97.68
learning
VI. RESULTS: Table 1 : Performance estimation of Prediction Deep learning
In this section, we will discuss the outcomes of this IoT-based models
system. It also compares the proposed models' performance Accuracy Recall Precision
on standard nutrition and plant leaf datasets to that of state- Method F-measure
(in %) (in %) (in %)
of-the-art approaches using deep learning for prediction and K- nearest
87.45 83.02 86.77 82.34
deep learning for classification. Neighbour
ANN- Genetic
89.04 91.88 89.35 90.11
Algorithm
Hybrid CNN 91.78 93.478 89.10 91.078
Classification
based deep 96.78 97.67 97.12 97.5
learning -CNN

Table 2 : Performance estimation of Classification deep

learning models

(a)

Fig 5: Classification of Diseases in plant using classification

based deep learning

In Figure 5, we see the classification based deep learning

model's classification of different plant diseases. The
Prediction based deep learning answer for three different sets
of sensor data is shown in Table 1. In this case, we consider
data from three separate samples: one collected during the
(b) rainy season, one during the winter, and one during the

Authorized licensed use limited to: Chulalongkorn University provided by UniNet. Downloaded on September 06,2024 at 07:04:54 UTC from IEEE Xplore. Restrictions apply.
 summer. These sensor readings were analysed by the [7] E. Nigussie, T. Olwal, G. Musumba, T. Tegegne, A. Lemma, and F.
Mekuria, “IoT-based irrigation management for small- holder farmers
Prediction based deep learning model, which produced in rural sub-Saharan Africa,” Procedia Com- puter Science..
accurate predictions of nutrient needs. The results of using [8] A. Sharma, M. Georgi, M. Tregubenko, A. Tselykh, and A. Tselykh,
the proposed Classification based deep learning for disease “Enabling smart agriculture by implementing arti- ficial intelligence
detection and classification are shown in Table 2. In this case, and embedded sensing,” Computers & Indus- trial Engineering, vol.
the proposed method achieved better results across the board 165, article 107936, 2022.
than state-of-the-art methods like K-nearest neighbour, [9] M. E. Karar, A. H. Abdel-Aty, F. Algarni, M. F. Hassan, M. A. Abdou,
and O. Reyad, “Smart IoT-based system for detecting RPW larvae in
ANN-Genetic algorithm, Hybrid-CNN. date palms using mixed depthwise convolu- tional networks,”
Alexandria Engineering Journal, vol. 61, no. 7, pp. 5309–5319, 2022.
VII. CONCLUSION [10] H. Pang, Z. Zheng, T. Zhen, and A. Sharma, “Smart Farming,”
The completion of this system as a whole will benefit International Journal of Agricultural and Environmental Infor- mation
tremendously from the expansion of solar hydroponics, which Systems (IJAEIS), vol. 12, no. 1, pp. 55–67, 2021.
are powered by the sun. Growing veggies and other plants [11] A. Khan, U. Nawaz, A. Ulhaq, and R. W. Robinson, “Real-time plant
health assessment via implementing cloud-based scal- able transfer
with this approach is healthy to the environment and uses less learning on AWS DeepLens,” PLoS One, vol. 15, no. 12, p. e0243243.
energy than traditional methods. Hydroponic systems that are
[12] F. Sabrina, S. Sohail, F. Farid, S. Jahan, F. Ahamed, and S. Gordon,
fueled by the sun provide solar energy, which is superior than “An interpretable artificial intelligence based smart agriculture
that produced by LEDs or electrical supply. For the purpose system,” Computers, Materials & Continua, vol. 72, no. 2, pp. 3777–
of controlling and monitoring the development of plants, an 3797, 2022.
Arduino-based solar hydroponics monitoring system has been [13] D. Adidrana, A. R. Iskandar, A. Nurhayati et al., “Simulta- neous
developed. In this setup, Wi-Fi is employed for the purpose of hydroponic nutrient control automation system based on Internet of
Things,” JOIV: International Journal on Infor- matics Visualization,
communication, while solar modules are used to produce
vol. 6, no. 1, pp. 124–129, 2022.
power. The system is able to automatically detect and control
[14] V. Bhatnagar and R. C. Poonia, “"sustainable development in
the temperature, humidity, pH level, and light on its own by agriculture: past and present scenario of Indian agriculture." research
using a wide range of sensors. When compared to systems anthology on strategies for achieving agricultural sus- tainability,” IGI
based on soil, it results in less wasted water and fertiliser while Global, pp. 1342–1364, 2022.
simultaneously increasing crop output. The data from the [15] N. G. Rezk, E. E.-D. Hemdan, A.-F. Attia, A. El-Sayed, and M. A. El-
sensor is sent to the Arduino Uno microcontroller, and Rashidy, “An efficient IoT based smart farming sys- tem using machine
connection between the Android smartphone and the Arduino learning algorithms,” Multimedia Tools and Applications, vol. 80, no.
1, pp. 773–797, 2021.
Uno is established.
[16] P. Wang, B. A. Hafshejani, and D. Wang, “An improved mul- tilayer
This article details the design and execution of a Raspberry perceptron approach for detecting sugarcane yield pro- duction in IoT
Pi, IoT, and mobile app-based AI-smart hydroponics system. based smart agriculture,” Microprocessors and Microsystems, vol. 82,
article 103822, 2021.
Agri-Hydroponic software helps farmers manage hydroponic
[17] Z. Khan, M. Zahid Khan, S. Ali et al., “Internet of things-based smart
farms. A Raspberry Pi controller gathers hydroponics farm farming monitoring system for bolting reduction in onion farms,”
sensor data. An IoT cloud service receives sensor data. Cloud- Scientific Programming, vol. 2021, 15 pages, 2021.
based deep learning algorithms analyse sensor data, monitor [18] E. T. Bouali, M. R. Abid, E. M. Boufounas, T. A. Hamed, and D.
plant health, and notify farmers via mobile app of potential Benhaddou, “Renewable energy integration into cloud & IoT-based
disease outbreaks. Finally, the farmer physically operates his smart agriculture,” IEEE Access, vol. 10, pp. 1175– 1191, 2022.
hydroponics farm field to provide nutrients. Automatically, [19] J. Arshad, M. Aziz, A. A. Al-Huqail et al., “Implementation of a
plants get predetermined nourishment. Optimisation and LoRaWAN based smart agriculture decision support system for
optimum crop yield,” Sustainability, vol. 14, no. 2, p. 827, 2022.
hybrid deep learning improve this system.
[20] M. Cicioğlu and A. Çalhan, “Smart agriculture with internet of things
REFERENCES in cornfields,” Computers and Electrical Engineering, vol. 90, article
106982, 2021.
[21] D. Manikandan, S. Prabhu, P. Chakraborty, T. Manthra, and M.
[1] W. Tao, L. Zhao, G. Wang, and R. Liang, “Review of the inter- net of Kumaravel, “IoT-Based Smart Irrigation and Monitoring System in
things communication technologies in smart agriculture and Smart Agriculture,” in Futuristic Communication and Network
challenges,” Computers and Electronics in Agriculture, vol. 189, article Technologies, pp. 45–56, Springer, Singapore, 2022.
106352, 2021. [22] H. N. Saha, R. Roy, M. Chakraborty, and C. Sarkar, “Develop- ment
[2] S. Qazi, B. A. Khawaja, and Q. U. Farooq, “IoT-equipped and AI- of IoT-based smart security and monitoring devices for agriculture,”
enabled next generation smart agriculture: a critical review, current Agricultural informatics: automation using the IoT and machine
challenges and future trends,” IEEE Access, vol. 10, pp. 21219–21235, learning, pp. 147–169, 2021.
2022. [23] P. S. Chatterjee, N. K. Ray, and S. P. Mohanty, “LiveCare: an IoT-
[3] D. Das, J. Sahoo, M. B. Raza, M. Barman, and R. Das, “Ongo- ing soil based healthcare framework for livestock in smart agricul- ture,” IEEE
potassium depletion under intensive cropping in India and probable Transactions on Consumer Electronics, vol. 67, no. 4, pp. 257–265,
mitigation strategies. A review,” Agronomy for Sustainable 2021.
Development, vol. 42, no. 1, pp. 1–26, 2022. [24] S. Biswas, L. K. Sharma, R. Ranjan, S. Saha, A. Chakraborty, and J. S.
[4] M. O. H. D. B. K. Putra and N. Nopriadi, “IOT based SMART Banerjee, “Smart farming and water saving-based intelligent irrigation
agriculture using fuzzy logic,” Computer and Science Industrial system implementation using the internet of things,” in Recent Trends
Engineering (COMASIE), vol. 6, no. 2, pp. 52–61, 2022. in Computational Intelligence Enabled Research, pp. 339–354,
[5] K. Lova Raju and V. Vijayaraghavan, “A self-powered, real- time, Academic Press, 2021.
NRF24L01 IoT-based cloud-enabled service for smart agriculture
decision-making system,” Wireless Personal Com- munications, pp. 1–
30, 2022.
[6] E. S. Mohameda, A. A. Belal, S. K. Abd-Elmabod, M. A. El- Shirbeny,
A. Gad, and M. B. Zahrana, “Smart farming for improving agricultural
management,” The Egyptian Journal of Remote Sensing and Space
Science, vol. 24, no. 3, pp. 971– 981, 2021.

Authorized licensed use limited to: Chulalongkorn University provided by UniNet. Downloaded on September 06,2024 at 07:04:54 UTC from IEEE Xplore. Restrictions apply.