Computers and Electronics in Agriculture 226 (2024) 109386

Contents lists available at ScienceDirect

Computers and Electronics in Agriculture

journal homepage: www.elsevier.com/locate/compag

Precision farming for sustainability: An agricultural intelligence model

Vinod Chandra S.S. a , Anand Hareendran S. b , Ghassan Faisal Albaaji a,c ,∗
a
Machine Intelligence Research Laboratory, Department of Computer Science, University of Kerala, Thiruvananthapuram, India
b
Department of Computer Science, Chinmaya Vishwavidyapeeth, Kochi, India
c
Directorate of Agriculture, Wasit Governorate, Iraq

ARTICLE INFO ABSTRACT

Keywords: Digital cultivation is emerging as one of the most promising fields that helped in creating an ecosystem for
Artificial intelligence smart farming. Precision farming, modernized techniques, and creating smart agriculture supply chains are
Agricultural intelligence the need of the hour for high-quality yield. Artificial Intelligence (AI) helps create a framework and plays an
Smart farming
important role in making decisions by analysing various data points. There are countries where more than
Sustainable agriculture
70% of the population depends on agriculture for their living, technological advancements help to improve
Digital farming
crop yields and get better farming results through sustainable ways. Each stage in agriculture, starting from
preparation of land, crop selection, type of fertilizer to use, to the kind of watering needed for the crops; can
be monitored and regulated by technological advancement. Farmers can also make decisions and implement
the best practices in their field by using AI and allied technologies. Disruptive technologies such as blockchain,
the Internet of Things, remote sensing, imaging technologies, and drones can transform the primitive way of
agriculture. Market analysis and user demands can also be foreseen, which helps farmers to get better yields.
Another important sector where technology can play a big role in disease control and pest management.
Artificial Intelligence-based farming creates high productivity and better yield, increasing individual farmers’
profit. In this study, the authors would like to throw light on AI and allied technologies, which can make
agricultural productivity increase significantly. In a post-pandemic situation, high-yield and more productive
farming will have a major impact. An agricultural intelligence framework model for self-sustained farming
is proposed in this work. The proposed framework will help achieve self-sustained growth with increased
economic stability. An end-to-end supply chain ensures customers are provided with quality products and
farmers are not financially looted. Technology-driven farming will also push the next generation to take up
agricultural jobs. The various advancements and strategies we propose in this study aim to build a better
ecosystem for transforming Artificial Intelligence into agricultural intelligence.

1. Introduction ecosystem to maximize farm productivity. The role of Artificial Intel-

ligence and expert systems in contributing to a nation’s agricultural
Agriculture is the backbone of the economy in nearly all developing development cannot be underestimated. As a field of study, Artificial
countries, with approximately 70%–75% of their populations directly Intelligence involves simulating human intelligence in machines, en-
or indirectly dependent on it for their livelihoods (Chandra, 2023).
abling them to think and act like humans (León-Bravo et al., 2019). This
Despite its critical role, it is disheartening to observe that the methods
simulation encompasses the capacity to reason and learn to navigate
and markets of agriculture often lag behind the necessary standards.
Farmers lack proper guidelines and scientific farming methods, leav- various scenarios successfully.
ing them unaware of the latest advancements and technologies. Even The challenge is to extend the boundless potential of AI and al-
agricultural research institutes seem to overlook the development of lied technologies to the ordinary person. Even though the intricacies
technologies that could benefit farmers. In stark contrast, Israel is of artificial intelligence may be beyond the comprehension of the
an exemplary model of advanced farming methods that yield high average individual, its impact on society is profound and pervasive,
productivity (Albaaji and Vinod Chandra, 2023). In this discussion, playing an indispensable role in modern life. For instance, imagine an
we will explore various technological advancements in farming and automated system examining a tomato seedling’s yellow flower and
outline a comprehensive blueprint for establishing an end-to-end smart

∗ Corresponding author.
E-mail addresses: vinod@keralauniversity.ac.in (Vinod Chandra S.S.), anandhareendrans@mgits.ac.in (Anand Hareendran S.), ghassan.albaaji88@gmail.com
(G.F. Albaaji).

https://doi.org/10.1016/j.compag.2024.109386
Received 11 December 2023; Received in revised form 10 June 2024; Accepted 23 August 2024
Available online 6 September 2024
0168-1699/© 2024 Elsevier B.V. All rights are reserved, including those for text and data mining, AI training, and similar technologies.
Vinod Chandra S.S. et al. Computers and Electronics in Agriculture 226 (2024) 109386

Fig. 1. Cyber physical system framework.

accurately predicting the duration until it ripens into a fruit ready for noting that Artificial Intelligence must be entered into the agricultural
harvesting, packaging, and distribution to grocery stores (Sharafudeen sector meticulously. Challenges such as unreliable market access, com-
and Vinod Chandra, 2023). This is not a future vision; it is today’s modity price fluctuations, low incomes, and high debts have made
reality. As we gaze into the next decade, we must focus on high- agriculture high-risk. As we reflect on the historical transition from
tech agricultural tools, intelligent transportation systems, consistent the Industrial Revolution, which reduced the need for human labour,
field analytics for farmers, and innovative pest control mechanisms. AI- to the present era, it is evident that it is time to introduce intelligent
powered advancements increase productivity, alleviate farmers’ debt machines (Mor et al., 2021).
burdens through reduced production costs and losses, and minimize Summing up these insights, the central objective of technologi-
the overall environmental impact (Ernst, 2022). Infusing science and cal advancements should be to establish a cyber–physical framework
technology into agriculture is pivotal for combating severe poverty and that allows farmers to maximize their profits by efficiently harnessing
famine. Microsoft Corporation, in collaboration with the International their available resources. Cyber-Physical Systems (CPS) represent a
Crops Research Institute for the Semi-Arid Tropics (ICRISAT), has novel engineered system class that seamlessly integrates computation
embarked on a groundbreaking initiative in Andhra Pradesh, India, and physical processes in dynamic environments. CPS encompasses
since 2018, providing services and solutions to 175 farmers (Albaaji various technologies, including cybernetics, mechatronics, design and
and Chandra, 2024). They developed an Artificial Intelligence sewing embedded systems, the Internet of Things (IoT), big data, Artificial
application that leverages machine learning and business intelligence Intelligence (AI), and more. These systems are intelligent, autonomous,
from the Microsoft Cortana Intelligence Suite, utilizing cloud-based and efficient, and they are poised to drive innovation across vari-
predictive analytics to empower farmers with essential information ous sectors, including agriculture, irrigation, energy, transportation,
and insights. This comprehensive system covers land preparation, seed infrastructure, security, health, and manufacturing (Jararweh et al.,
sowing, climate adaptation, fertilizer application, and the use of organic 2023).
nutrients, among other aspects. The result has been a remarkable 30% The distinguishing characteristics of a cyber–physical system in-
increase in average yield per hectare compared to previous harvests. clude cyber capabilities within every physical component, networking
Predictive analytics are instrumental in assisting farmers in making at multiple scales, dynamic reconfiguration, logical and physical cou-
informed decisions regarding crop sowing, fertilizer application timing, pling of constituent elements, and high levels of automation with
and harvest timing (Ahmad and Sharma, 2023). This involves analysing control loops operating at various scales. A comprehensive framework
climate data, historical conditions, market conditions, and individual that outlines the supporting technologies, requirements, features, and
farming information (Rodríguez et al., 2020). However, it is worth high-level applications is presented in Fig. 1. The expanded framework

2
Vinod Chandra S.S. et al. Computers and Electronics in Agriculture 226 (2024) 109386

Fig. 2. IoT sensor model.

aims to foster a network where consumers and producers coexist, that by 2050, agricultural productivity must surge by at least 70% to
facilitating the discovery of demand and supply. It is important to em- cater to the escalating global population demands. The pressing con-
phasize that the proposed framework goes beyond the impact of AI in cerns of natural resource depletion and diminishing arable land have
agriculture; it indirectly promotes a holistic approach to agro-ecology, compellingly mandated the enhancement of farm yields. The Internet
agro-forestry, and climate conservation agriculture. This holistic ap- of Things (IoT) is the networking of physical objects equipped with
proach underscores the pivotal role of technology in ensuring food sensors, processors, software, and other technologies that allow them
security, poverty reduction, and environmental sustainability (Talaviya to connect and share data over the Internet or similar communication
et al., 2020). networks. In the agricultural landscape, IoT integration necessitates
the deployment of diverse sensors for data collection, subsequently
2. Technologies in agriculture archived within sensor-attached memory units for subsequent analysis.
However, this paradigm proves less efficacious when confronted with
Built-in sensors and new technologies that use data have become live data mandating real-time responses.
very important in making many areas of agriculture more efficient.
The intrinsic limitations of memory capacity within individual sen-
These include checking the quality of seeds, keeping an eye on the
sors often lead to data inundation, thus advocating a paradigm shift
soil, predicting the weather, analysing crops, tracking market changes,
towards cloud-based analytics. Cloud-centric analytics entail transmit-
and setting up distribution networks (Tewari et al., 2020). Integrating
ting data gleaned from diverse sensors—such as those measuring tem-
cloud computing infrastructure, robust data ecosystems, the Internet
perature, soil moisture, and mineral levels—to dedicated cloud-based
of Things (IoT), and machine learning is a promising avenue towards
servers for comprehensive analytics. The resultant analytical outcomes
enhancing farming productivity. These technologies are intricately fo-
are disseminated through customized local websites or mobile inter-
cused on optimizing various facets of agricultural practices, ranging
from resource-efficient farming methods, precise irrigation techniques, faces. For instance, leveraging data obtained from soil moisture and
intelligent chemical application, and streamlined harvesting method- temperature sensors, an automated drip irrigation system can be metic-
ologies to effective pest control strategies. Notably, the advent of ulously devised, ensuring optimal crop hydration regardless of prevail-
Artificial Intelligence-based, AI-based automated irrigation systems has ing climatic conditions. As shown in Fig. 2, implementation methods
showcased a remarkable surge in the efficacy of the contemporary often use technologies like Arduino or microcontrollers. This is shown
agricultural supply chain. The need for food grains and the demand in a schematic layout that shows an example of an IoT Sensor Model:
for freshwater resources have increased, underscoring AI’s crucial role a network of sensors used in agriculture to check on temperature, soil
in facilitating intelligent irrigation, monitoring water levels, soil tem- moisture, and mineral levels. These sensors are connected to a cloud-
perature, nutrient content, and predictive weather forecasting (Abioye based server via wireless communication, enabling real-time data trans-
et al., 2022). mission and storage. This setup allows continuous monitoring of field
Models must be meticulously crafted to meet future agricultural conditions, which is essential for precision agriculture. The collected
demands to fulfil these requisites (Sanga et al., 2020). The United data is processed and analysed in the cloud, where advanced algorithms
Nations Food and Agriculture Organization (UNFAO) has stipulated provide actionable insights for farmers. This cloud-centric approach

3
Vinod Chandra S.S. et al. Computers and Electronics in Agriculture 226 (2024) 109386

facilitates centralized data management, scalability, and accessibil- blockchain-enabled access to comprehensive crop details, including
ity, empowering informed decision-making for optimized irrigation, seed origins, planting specifics, fertilization, harvest timelines, and
crop health monitoring, and overall farm management, thus enhancing export details, culminating in QR code-enabled information at the point
productivity and sustainability (Sadowski and Spachos, 2020). of sale, offering a transparent end-to-end view of the supply chain.
Moreover, ground and aerial-based drones have emerged as indis- Leveraging plant growth images and videos is a cornerstone for
pensable tools within agriculture, amplifying a spectrum of agricultural deploying intelligent models underpinned by machine learning algo-
practices encompassing crop health assessment, irrigation, monitoring, rithms trained on vast datasets to facilitate real-time predictions. This
spraying, planting, harvest operations, soil, and field analysis. Smart ingrained intelligence finds widespread application across diverse agri-
greenhouses, augmented through IoT frameworks, exemplify intelligent cultural activities, encompassing drone and satellite imagery analysis,
climate monitoring and regulation, obviating manual intervention (Eli- disease and pest detection, weather forecasting, and optimal timing
Chukwu, 2019). Predictive analytics, integral to intelligent farming, aid for fertilizer application. A specialized facet of AI is tasked with pre-
farmers in strategic decision-making about crop production, storage, dicting optimal chemical spray usage and addressing variations in
optimal marketing techniques, and risk mitigation. Integrating sensor- harvest times and product yields, even under uniform fertilization
derived data, encompassing soil parameters, temperature, pressure, conditions. The intricate complexities of amalgamated farming issues
rainfall, and humidity, via dashboards or bespoke mobile applications find resolution through developing and utilizing specialized built-in
furnishes farmers with precise soil insights (Ramesh and Vydeki, 2020; intelligence models, leveraging in-depth agricultural data, growth pat-
Faisal et al., 2023). A recent case study on implementing an AI- terns, and disease prevalence, culminating in actionable market and
based pest management system at a large-scale agricultural farm in harvest forecasts (Furstenau et al., 2023).
California underscores the transformative potential of such technology.
Before this intervention, the farm experienced significant crop yield 2.1. Advanced vs. Traditional farming methods
reductions due to persistent pest infestations. A substantial change in
their pest management strategy was introducing an AI-driven solution Advanced agricultural technologies significantly enhance farming
by an agricultural technology startup. This advanced system leveraged efficiency and sustainability compared to traditional methods. Precision
machine learning algorithms alongside real-time data collection from farming utilizes GPS-guided machinery and data analytics to optimize
sensors and drones to accurately identify and predict pest outbreaks. inputs like water and fertilizers based on specific field conditions, lead-
Consequently, the farm achieved a 40% reduction in pesticide usage, ing to higher crop yields and more efficient resource use. This approach
improved overall crop health, and realized a 25% increase in produc- contrasts with traditional farming’s broad and often inefficient input
tivity within a year. This case exemplifies AI’s profound impact on pest application, which relies heavily on manual labour (Kırkaya, 2020).
management practices, promoting greater sustainability and efficiency Intelligent sensors and IoT devices provide real-time data on soil mois-
in modern agriculture (Victor et al., 2024). ture, temperature, and crop health, enabling timely decision-making
Artificial Intelligence heralds a multitude of avenues for augmenting and adjustments that prevent crop diseases and optimize irrigation.
farmer incomes and bolstering agricultural productivity, notably by Unlike traditional periodic manual inspections, these technologies of-
leveraging diverse AI frameworks. Determining soil types and nutrition, fer continuous monitoring, improving resource efficiency and crop
crucial crop selection, and quality determinants can be systematically management (Sreekantha and Kavya, 2017).
facilitated through diverse AI frameworks. To safeguard cultivated AI and machine learning further revolutionize agriculture by analysing
crops from threats posed by animals and intruders, AI-driven surveil- vast data sets to offer predictive analytics, crop modelling, and de-
lance systems adeptly monitor real-time video feeds from agricultural cision support. These technologies help farmers optimize planting
fields, promptly identifying breaches and issuing timely alerts. AI also schedules, predict yields, and manage risks more effectively than
uses real-time sensor data and visual analytics enabled by drones to traditional decision-making processes, which rely on experience and
predict and find pest infestations before they happen. It also programs intuition (Virnodkar et al., 2020). By comparing these advanced tech-
autonomous robots to deliver fertilizer precisely, which saves money nologies with traditional methods, it is evident that precision farming,
and increases crop yields. Accurate crop price forecasting based on smart sensors, IoT, and AI collectively enhance agricultural sustain-
yield projections holds immense potential in devising cogent pricing ability and productivity through targeted, efficient, and data-driven
strategies and minimizing waste, empowering agricultural stakeholders resource management.
to negotiate optimal harvest prices. AI’s pervasive role across these
domains fortifies ‘‘Agriculture as a Service’’, empowering farmers to 2.2. Challenges in adopting technologies in agriculture
automate farming practices, adopt precision cultivation techniques, and
garner higher crop yields while optimizing resource utilization (Liakos Challenges loom large within the ambit of futuristic agricultural
et al., 2018). Table 1 shows the multifaceted utilities of AI delineated sectors, notably the concerted thrust across the supply chain to amplify
across eight principal domains. capacity and slash costs, which might undercut the farmers. A rural
Connecting these AI applications to the evolving agricultural land- farming community in Sub-Saharan Africa exemplifies this. Farmers
scape dynamics to discern demand and supply nuances The burgeoning endeavoured to integrate IoT sensors for precision agriculture in this
demand for data-driven farming and predictive analytics, especially context but were confronted with severe challenges, such as inadequate
in the post-pandemic landscape, mandates a proactive adoption of internet connectivity and restricted access to data networks. The in-
precision farming methodologies buoyed by AI and machine learn- sufficiency of reliable connectivity severely hindered their capacity to
ing frameworks. These frameworks underpin comprehensive crop and transmit real-time data and access crucial information necessary for in-
livestock management, mitigating conjecture through yield forecasting, formed decision-making. Because of this, using IoT sensors did not have
adept supply chain management, and risk assessment (Chandra and the desired effects, showing how vital robust internet infrastructure
Albaaji, 2023). This knowledge-driven strategy is the perfect example is for making advanced agricultural technologies work well (Ahmad
of the paradigm shift that automated analytics has enabled. Blockchain and Sharma, 2023). The integration of built-in intelligent technologies,
technology, characterized by its decentralized and immutable ledger while carrying unique advantages, grapples with several hurdles in
system, is poised to revolutionize agricultural supply chains by en- its adoption within the agricultural domain, particularly in ASEAN
suring digital assets’ transparent and traceable provenance. Applying countries, as shown in Table 2.
blockchain within agriculture facilitates seamless traceability, from Additionally, the landscape of AI-based research in Southeast Asian
seed identification to harvest (Kim et al., 2020). Consumers, increas- countries presents its own set of challenges. Even in the most developed
ingly concerned about food quality and sourcing, stand to benefit from laboratories, a small community of 50 to 75 professionals operates

4
Vinod Chandra S.S. et al. Computers and Electronics in Agriculture 226 (2024) 109386

Table 1
The multifaceted utilities of AI across eight principal domains.
Domain Description
1. IoT-driven growth for crop nourishment AI analyses data from IoT sensors to optimize crop nourishment, offering insights into soil conditions,
nutrient levels, and water needs. This helps farmers make data-driven decisions on fertilization and
irrigation to boost crop growth and yield.
2. Image-based analysis for monitoring crop health AI algorithms process images from drones or sensors to assess crop health, detecting diseases, pests,
and nutrient deficiencies. This early detection enables timely corrective actions to safeguard and
enhance crop health.
3. Formulation of tailored agricultural product blends for AI customizes fertilizer blends based on soil composition, crop type, and growth stage, ensuring
superior fertilization optimal nutrient delivery for healthy growth and maximum yield.
4. Augmented crop health monitoring for maximal yield AI-powered systems continuously monitor crop indicators like growth rates, leaf colour, and moisture
levels, enabling farmers to adjust practices to maximize yield potential.
5. Empowering farmers to optimize irrigation for efficient AI analyses soil moisture, weather forecasts, and crop water needs to optimize irrigation schedules,
resource utilization conserve water, reduce costs, and ensure optimal crop hydration.
6. Implementation of autonomous technological interventions in AI-driven technologies, such as robotic systems and automated machinery, enhance planting,
agriculture harvesting, and spraying efficiency, reducing labour costs and improving farm operations.
7. Strategic direction for value-added products, manufacturing, AI aids in developing value-added products, optimizing manufacturing, and implementing marketing
and enterprise marketing strategies by analysing market trends and consumer preferences, helping farmers diversify offerings
and increase profitability.
8. Perceptive comprehension of the market AI analyses market data, consumer behaviour, and supply chain information to provide insights into
demand fluctuations and supply chain dynamics, guiding production planning, pricing, and market
positioning.

Table 2
Challenges in integrating advanced technologies into agricultural practices.
Challenge Description
1. Connectedness and data availability The fundamental challenge is ensuring robust connectivity and data availability across agricultural setups.
2. Financial stability The financial underpinning for integrating advanced technologies remains a hurdle, especially for smaller farming
communities.
3. Reliability of systems and technology Ensuring the dependability and resilience of integrated systems and technology poses a significant challenge.
4. Security and acceptance of information Concerns about the security of data and the broader acceptance of information among stakeholders hinder seamless
adoption.
5. Data Privacy, Storage, and Use The ethical and regulatory aspects of data privacy, secure storage, and appropriate utilization present formidable
challenges.
6. Lack of technical expertise A shortage of technical know-how and expertise impedes the effective implementation of sophisticated technologies.
7. Social acceptance and recognition Garnering social acceptance and recognition for these advanced systems remains challenging.
8. Cost-Effectiveness Ensuring the cost-effectiveness of integrating these technologies into agricultural practices is a significant concern.
9. Ease of use and training Ensuring ease of use and providing adequate training for stakeholders remain essential for successful adoption.
10. Unethical Stakeholders Mitigating the impact of unethical stakeholders poses a significant challenge in integrating these technologies.
11. The unsuitable facilities The lack of infrastructure limits modern farming technology. Without solid infrastructure, these technologies cannot
maximize their benefits, resulting in inefficiencies and poor agricultural yield and sustainability.

within this domain, hindering significant advancements. Factors such 2.3. Artificial intelligence-based framework
as limited space, resource scarcity, administrative barriers, research
methodologies, subpar computing facilities, inadequate quality data A 2020 economic survey that emphasized the creation of farm-
interpretation, and insufficient fund allocation in national budgets ing models based on sizeable datasets highlighted the necessity of
underscore the predicaments afflicting Artificial Intelligence research data-based farming to achieve high yields (Gaitan et al., 2015). Inte-
in agriculture. Translating these technologies from the laboratory envi- grating Artificial Intelligence into agriculture and the agricultural mar-
ronment to real-world scenarios is imperative to propagate the growth ket presents potential advancement. This implementation has already
of Artificial Intelligence applications in agriculture, ensuring that all found practical application in various developed countries (Javaid
stakeholders derive benefits. The deployment of recommendation sys- et al., 2023).
tems benefits consumers and farmers alike, streamlining the entire Fig. 3 depicts a model framework called Systems of Sedentary
behaviours (SoS), illustrating rainfall prediction in specific areas using
supply chain swiftly and efficiently. Integrating Artificial Intelligence-
Artificial Intelligence techniques like artificial neural networks. Such
driven methodologies across diverse farming stages is poised to usher
methodologies, developed by meteorological departments, contribute
in an era of flourishing agricultural resources, necessitating the simulta-
significantly to rainfall and climate change forecasting. When incor-
neous advancement of multiple research methodologies. Furthermore,
porated into the SoS framework, these technologies yield maximal
the escalating importance of agriculture and the surge in demand for agricultural profits and higher crop yields for farmers (Priyadarshini
food grains in the post-pandemic era underline the need for novel meth- and Abhilash, 2020), imparting a sense of direction to the forthcoming
ods and AI-based frameworks in agriculture. We must push for new generations.
technological frameworks and improvements in farming methods and Artificial Intelligence in agriculture encapsulates various subfields
marketing strategies to get around problems like limited transportation, such as natural language processing, robotics, machine learning, auto-
scarce raw materials, and trouble deploying skills. This will help the matic reasoning, knowledge representation, expert systems, computer
national economy grow. vision, speech recognition, automated data analytics, cloud computing,

5
Vinod Chandra S.S. et al. Computers and Electronics in Agriculture 226 (2024) 109386

Fig. 3. Artificial intelligence-based SoS.

statistical computing, and deep learning (Gaitan et al., 2015). The tech- ensuring higher yields (Gill et al., 2017). Furthermore, farmers’ depen-
nological software that farmers use integrates many of these subfields. dency on external factors and the perishable nature of agro-products
For instance, standalone Q&A systems in diverse languages empower necessitate AI-driven solutions for optimal storage, transportation, and
farmers. At the same time, deep learning-based agriculture mapping demand forecasting. Predicting market needs during specific events or
from satellite images provides intelligent forecasting systems for soil festivals can also aid in foreseeing demand–supply dynamics, thereby
science, crop yield, pest management, and harvesting. creating new market opportunities (Lin et al., 2017).
The SoS framework relies fundamentally on data and analytics. A case study of a precision irrigation system in a Napa Valley
It incorporates diverse environmental factors—dew, sunlight, rainfall, vineyard demonstrates the substantial benefits of advanced agricultural
and other climatic changes—analysed via sensors and satellite images. technology. The vineyard implemented a Vititech Solutions system that
The subsequent data processing by an analytical engine, which causes uses real-time data from soil moisture sensors and weather forecasts to
trained models to start particular tasks, emphasizes how dependent deal with water scarcity and inconsistent grape quality. This AI-driven
prediction and automation are on data availability. The knowledge system optimizes water distribution, ensuring each vine receives the
acquired is instrumental in developing decision support systems ben- necessary amount. Consequently, the vineyard achieved a 30% reduc-
efiting agriculture and allied services like water and livestock manage- tion in water usage and a 20% improvement in grape quality within
ment. two seasons. This case underscores the potential of precision irrigation
Precision agriculture emphasizes enhancing overall harvest quality to enhance sustainability and productivity in viticulture (Seyar and
and accuracy through AI methods. AI can be used in agriculture to Ahamed, 2023). Thus, the application of technology, particularly AI,
find weeds with sensors connected to herbicides and make intelli- holds promise for farmers with limited landholdings, aiding in weather
gent predictions about where to store seeds and how to distribute prediction at the local level and guiding sustainable farming practices.
grains (Mahmud et al., 2023). Crop diseases pose a severe challenge, This technological integration assists in pest management, robotics for
leading to diminished productivity. Machine learning techniques fa- multi-crop harvesting, demand prediction, and meeting regional and
cilitate early identification and remedial solutions for crop diseases, export requirements.

6
Vinod Chandra S.S. et al. Computers and Electronics in Agriculture 226 (2024) 109386

Table 3 combined with digital image processing help identify diseases and find
Southeast Asian global rankings for connectedness and data flows.
the right treatments (Subeesh and Mehta, 2021). This lets farmers find
Country Connectedness index Dataflow GDP/Capita rank Reference problems quickly and fix them, preventing possible epidemics.
Singapore High High 1 world bank (2023)
Malaysia Upper-middle Medium 61 ITU (2023)
4. Results
Thailand Upper-middle Medium 93 world bank (2023)
Vietnam Lower-middle Low 126 world bank (2023)
Indonesia Lower-middle Low 170 WDI (2023) The paradigm of modern agriculture is evolving by integrating
Philippines Lower-middle Low 132 world bank (2023) cutting-edge technologies into farm operations. The proposed AI-based
model for agriculture, which shows the stages of its implementation
in Fig. 4, incorporates advanced technologies into a cohesive system
designed to enhance productivity and sustainability. Initially, data is
3. Agricultural intelligence model
collected through sensors and IoT devices, capturing real-time infor-
mation on soil moisture, temperature, humidity, pest activity, and
This study is centred on Southeast Asian countries that rely heavily
crop health. Fig. 5 represents samples of sensors used in this stage.
on agriculture for economic stability. Comprising eleven nations—
This data is then transmitted and stored using edge computing for
Brunei, Burma, Cambodia, Timor-Leste, Indonesia, Laos, Malaysia,
immediate processing and cloud storage for centralized management.
Philippines, Singapore, Thailand, and Vietnam—this region reflects
Subsequently, machine learning algorithms and big data analytics anal-
remarkable diversity in culture, technology, farmland, and economy.
yse the data, uncovering patterns and generating actionable insights.
Notably, Malaysia, Philippines, Singapore, Thailand, Indonesia, and
These insights feed into a decision support system (DSS), which pro-
Vietnam contribute over 70% of their income through agri-food, show-
vides AI-driven recommendations through a user-friendly interface,
casing their technological vibrancy and extensive farmland. The fo-
cal point of this manuscript on these six countries stems from their facilitating informed decision-making for farmers. Automated systems
economic stability, technological adaptability, and ample farmland and alert mechanisms are then activated to execute necessary actions,
availability for multiple crops. Their remarkable economic upsurge such as adjusting irrigation or deploying pest control measures. The
substantiates these nations’ significant emphasis on their agricultural model incorporates a feedback loop that ensures continuous learning
sectors in recent years. Critical to these countries is ensuring food secu- and improvement by integrating outcome data and farmer input, thus
rity. In 2019, the agri-food sector contributed a staggering $717 billion adapting to evolving conditions and practical insights.
to the collective economies of Indonesia, Thailand, the Philippines, and Fig. 6 shows an advanced agriculturally intelligent farm that uses an
Vietnam, marking a 30% increase from 2015. Moreover, this sector Application Programming Interface (API) and a network of technolog-
employed 48% of the total workforce across these countries, generating ical parts, including a crop-connected sensor network, to run the farm.
127 million jobs. The Association of Southeast Asian Nations (ASEAN) This integration is the backbone for comprehensive data analysis, em-
emerges as a globally productive agricultural region, and integrating powering the farm’s operations. The orchestration of multiple sensors,
technological solutions will catalyze exponential market growth. activated by diverse alarms, is a cornerstone of ensuring the operational
Connectivity and data flow are instrumental in fostering growth and efficiency of this system; Fig. 5 represents samples of sensors that were
enabling an agricultural boom in these nations. Fortunately, prominent used.
players in this sector rank within the top 50 regarding connectedness, Notably, these sensors play distinct roles, with soil moisture sensors
indicating a promising landscape if these countries receive adequate triggering precise watering mechanisms and image analysis to evaluate
technological training. The connectedness index of ASEAN countries crop health and instigating interventions such as medication, pest con-
is depicted in Table 3 (Source: International Monetary Fund, World trol, and fertilizer administration. Furthermore, the adaptive regulation
Economic Outlook Database). of the greenhouse, responding to UV exposure by opening or closing
‘‘New standards for new-age farming’’ epitomizes Agriculture Rev- as needed, is a testament to the system’s dynamic responsiveness. The
olution 5.0. The diverse standalone technologies mentioned serve as meticulous recording of each analysis into a blockchain repository,
crucial aids for farmers. Over 40 critical farming activities require made possible by node entry, is a crucial component of this agricultural
analysis before making decisions, ensuring optimal yields and returns framework. This approach meticulously traces the entire life cycle of
on investment (Chohan et al., 2020). Farmers’ obstacles include diffi- crops, commencing from seed selection to harvest and eventual retail.
culties such as low-quality inputs, the unpredictability of the weather, Subsequently, these meticulously documented data traverse to cloud
incorrect forecasting rules for crucial phases of crop development, analytics, presenting predictions readily accessible to farmers through
and intermediate problems such as a lack of available labour and web-based platforms or applications installed on their smart devices.
inadequate insurance coverage. Efficient access to real-time data en- The accuracy and reliability of these predictions are contingent upon
compassing weather, crop characteristics, soil, and pest data, coupled the robustness of the sensor networks and the efficacy of the data
with new and emerging technologies like drones, cloud computing, and transmission frameworks employed. Delineating the sensors essential
image analytics, has transformative potential for traditional agriculture. to intelligent farming reveals a multifaceted toolkit integral to farm
This convergence of technology and new-age products could redefine management. Acoustic sensors have emerged as versatile tools, con-
traditional agriculture into data-driven precision farming, ensuring tributing significantly to soil cultivation, weeding, and fruit harvesting.
profitability and sustainability. Their adaptability and swift response rates, mainly when employed
Digital agriculture is a panacea for the multifaceted problems haunt- with portable equipment, render them cost-effective solutions across
ing the agricultural domain. Digitization promises farmers enhanced various farm management facets. Adding ultrasonic ranging sensors
access to markets, inputs, advisory services, and credit. However, the and a camera for finding weeds expands the range of functions even
lack of timely and accurate data remains a bottleneck, hindering the more. This lets you keep track of tank levels, evaluate crop canopy,
establishment of a demand-driven, efficient supply chain. Mitigating and measure spray distance. Also, adding airflow sensors helps fig-
these challenges necessitates synthesizing data for informed decision- ure out things about the soil, like how absorbent it is to air, how
making by various stakeholders, facilitated by digital technologies and much water is in it, and its structure, which makes it possible to
numerous AgriTech startups. Addressing pest control remains a signifi- tell the difference between different soil types. Elective sensors can
cant challenge in this sector. Although not entirely automated, technol- record electrical conductivity and transient electromagnetic responses
ogy, notably AI facilitated through drones, can substantially enhance very well. This makes variable-rate applications possible, which is
pest control efficiency, ensuring safer, quicker, and cost-efficient solu- essential in situational farming situations. Additionally, deploying Eddy
tions for farmers (Li et al., 2020). AI-powered deep learning algorithms covariance-based sensors quantifies gas and energy exchanges between

7
Vinod Chandra S.S. et al. Computers and Electronics in Agriculture 226 (2024) 109386

Fig. 4. The fundamental phases of agricultural intelligence model.

Fig. 5. Examples of sensors specifically designed for data collection in precision farming.

the Earth’s surface and the atmosphere, adding depth to environmental operations. Diverse sensors, activated by specific alarms, play crucial
monitoring within agricultural contexts. roles in optimizing various aspects of farm management. Soil man-
The contributions of artificial intelligence technologies to the agri- agement sensors encompass a range of functionalities, from enabling
cultural sector appear in many aspects, incorporating data such as precise watering based on soil conditions using moisture sensors to
a 30% reduction in water usage and a 20% improvement in grape utilizing image analysis for crop health assessment, thereby triggering
quality due to AI-driven precision irrigation, as well as the $717 billion interventions like medication, pest control, and fertilizer administra-
economic contribution of the agri-food sector in Southeast Asia, which tion. Additionally, the adaptive response of greenhouse sensors to
has increased by 30% since 2015 and created 127 million jobs, would UV exposure dynamically optimizes growing conditions, showcasing
significantly enhance the validity and impact of the findings, further- the system’s responsiveness to environmental cues. The meticulous
recording of analyses in a blockchain repository ensures secure and
more the pervasive utilization of Light Detection and Ranging (LIDAR)
traceable records of the crop’s entire life cycle. At the same time,
technology across agricultural applications is notable, encompassing
data transmission to cloud analytics facilitates the delivery of accurate
activities such as land mapping, determination of soil types, 3D farm
predictions accessible to farmers through web pages or smart device
modelling, erosion monitoring, and yield forecasting. Similarly, remote
applications. The accuracy of these predictions relies significantly on
sensing sensors are essential in storing and analysing geographic data
robust sensor networks and efficient data transmission frameworks.
and find extensive application in crop assessment, yield modelling, and An array of essential sensors in intelligent farming includes acoustic
forecasting endeavours (Weiss et al., 2020). sensors with versatile functionalities in soil cultivation, weeding, and
fruit harvesting. Their cost-effectiveness and swift response rates, es-
4.1. Discussion pecially with portable equipment, render them valuable for diverse
farm management tasks. Further integration of cameras and ultrasonic
In this study, the advanced agriculturally intelligent farm operates sensors aids in weed detection, tank level tracking, crop canopy as-
through a sophisticated network of technologies, primarily a crop- sessment, and spray distance measurement. Moreover, airflow sensors
connected sensor network linked to the farm’s API. This integration ascertain critical soil characteristics, such as air permeability, mois-
is the cornerstone for comprehensive data analysis and efficient farm ture percentage, and structure, enabling the differentiation between

8
Vinod Chandra S.S. et al. Computers and Electronics in Agriculture 226 (2024) 109386

Fig. 6. Agricultural intelligence prototype.

Fig. 7. Predictive demand.

soil types. Electromagnetic sensors record electrical conductivity, fa- the deployment of artificial intelligence-based frameworks. To optimize
cilitating variable-rate applications in situational farming scenarios. agricultural yield, a crucial recommendation surfaces: an escalated im-
Additionally, eddy covariance-based sensors quantify gas and energy petus on bolstering research and development endeavours across allied
exchanges between the Earth’s surface and the atmosphere, offering domains of agriculture. Concurrently, enhancing interaction among
insights into environmental dynamics. The extensive application of stakeholders at multiple supply chain echelons is pivotal to benefiting
LIDAR technology in agricultural domains spans activities such as land farmers comprehensively. This envisioned work espouses the imper-
mapping, 3D farm modelling, erosion monitoring, and yield forecasting. ative of a self-sustaining ecosystem and introduces a comprehensive
Likewise, remote sensing sensors are pivotal in storing and analysing framework model underpinning agricultural intelligence. An automated
geographic data supporting crop assessment, yield modelling, and fore- farmland system stands out as the linchpin for ensuring profitability
casting. Looking ahead, the emergence of Agriculture 5.0 emphasizes and bolstering productivity within the agricultural domain. Adopting a
Systems of Sedentary Behaviours (SoS) framework is a cornerstone for
the necessity for self-sustaining autonomous farms, promoting ease,
establishing a robust and refined system architecture. At the crux of
viability, and heightened productivity. Given that the demand for
this paradigm shift lies the imperative to comprehend innovative and
food grains is expected to rise by 2050, almost doubling the current
intelligent processes from diverse agricultural universities and research
production, it becomes more apparent that we need to switch to self-
centres and discern their tangible impact on agricultural practices.
sufficient digital farming methods. Fig. 7 shows the predicted demand However, amid the integration of these sophisticated technological
for 2050 in metric tons, emphasizing the urgent need to take proactive paradigms, the challenge persists in empowering farmers and culti-
steps to transform agriculture. This transition is crucial to addressing vating their awareness regarding these transformative technologies. A
the anticipated agricultural challenges in the coming decades. concerted, collective effort emerges as the linchpin in addressing this
challenge and harnessing the dividends of technological advancements
5. Conclusion for the agricultural community. As an unintended but significant side
effect, the suggested method pushes for a more complete view that
The global landscape is witnessing a concerted effort towards ush- includes agro-ecology, agro-forestry, and expanding the range of live-
ering in next-generation agricultural strategies, prominently featuring stock markets. Notably, the roadmap to realizing Agriculture 5.0 hinges

9
Vinod Chandra S.S. et al. Computers and Electronics in Agriculture 226 (2024) 109386

on the collective adoption of an integrated, multi-dimensional research Chandra, V.S., 2023. Role of artificial intelligence in indian agriculture: A review. Agric.
approach spanning agriculture and its allied realms. This holistic amal- Rev. 44 (4), 558–562. http://dx.doi.org/10.18805/ag.R-2296.
gamation of advanced technologies and progressive methodologies sets Chandra, S.S.V., Albaaji, G.F., 2023. Automatic orange fruit classification using convo-
lutional neural network. Alkut Univ. Coll. J. 2023 (The 6th Scientific Conference),
the stage for a promising and vibrant Agriculture 5.0 era.
254–263.
Chohan, M., Khan, A., Chohan, R., Katpar, S.H., Mahar, M.S., et al., 2020. Plant
CRediT authorship contribution statement disease detection using deep learning. Inte. J. Recent Technol. Eng. 9 (1), 909–914.
http://dx.doi.org/10.35940/ijrte.A2139.059120.
Vinod Chandra S.S.: Writing – review & editing, Writing – origi- Eli-Chukwu, N.C., 2019. Applications of artificial intelligence in agriculture: A review.
nal draft, Visualization, Validation, Supervision, Software, Resources, Eng. Technol. Appl. Sci. Res. 9 (4).
Project administration, Methodology, Investigation, Funding acquisi- Ernst, E., 2022. The AI trilemma: Saving the planet without ruining our jobs. Front.
Artif. Intell. 5, 886561. http://dx.doi.org/10.3389/frai.2022.886561.
tion, Formal analysis, Data curation, Conceptualization. Anand Ha-
Faisal, G., Sreelakshmi, S., Chandra SS, V., 2023. Crop yield prediction for smart
reendran S.: Writing – review & editing, Writing – original draft,
agriculture with climatic parameters using random forest. In: International Con-
Visualization, Validation, Supervision, Software, Resources, Project ad- ference on Advances in Computing and Data Sciences. Springer, pp. 367–376.
ministration, Methodology, Investigation, Funding acquisition, Formal http://dx.doi.org/10.1007/978-3-031-37940-6_30.
analysis, Data curation, Conceptualization. Ghassan Faisal Albaaji: Furstenau, L.B., Rodrigues, Y.P.R., Sott, M.K., Leivas, P., Dohan, M.S., López-
Writing – review & editing, Writing – original draft, Visualization, Robles, J.R., Cobo, M.J., Bragazzi, N.L., Choo, K.-K.R., 2023. Internet of things:
Validation, Supervision, Software, Resources, Project administration, Conceptual network structure, main challenges and future directions. Digit.
Commun. Netw. 9 (3), 677–687. http://dx.doi.org/10.1016/j.dcan.2022.04.027.
Methodology, Investigation, Funding acquisition, Formal analysis, Data
Gaitan, N.-C., Gaitan, V.G., Ungurean, I., 2015. A survey on the internet of things
curation, Conceptualization. software architecture. Int. J. Adv. Comput. Sci. Appl. 6 (12), 25–31.
Gill, S.S., Chana, I., Buyya, R., 2017. IoT based agriculture as a cloud and big data
Declaration of competing interest service: the beginning of digital India. J. Organ. End User Comput. (JOEUC) 29
(4), 1–23. http://dx.doi.org/10.4018/JOEUC.2017100101.
The authors declare that they have no known competing finan- ITU, 2023. Measuring digital development facts and figures.
cial interests or personal relationships that could have appeared to Jararweh, Y., Fatima, S., Jarrah, M., AlZu’bi, S., 2023. Smart and sustainable agricul-
ture: Fundamentals, enabling technologies, and future directions. Comput. Electr.
influence the work reported in this paper.
Eng. 110, 108799. http://dx.doi.org/10.1016/j.compeleceng.2023.108799.
Javaid, M., Haleem, A., Khan, I.H., Suman, R., 2023. Understanding the potential
Data availability applications of artificial intelligence in agriculture sector. Adv. Agrochem 2 (1),
15–30. http://dx.doi.org/10.1016/j.aac.2022.10.001.
Data will be made available on request. Kim, W.-S., Lee, D.-H., Kim, Y.-J., 2020. Machine vision-based automatic disease
symptom detection of onion downy mildew. Comput. Electron. Agric. 168, 105099.
Acknowledgements http://dx.doi.org/10.1016/j.compag.2019.105099.
Kırkaya, A., 2020. Smart farming-precision agriculture technologies and practices. J.
Sci. Perspect. 4 (2), 123–136. http://dx.doi.org/10.26900/jsp.4.010.
The authors appreciate the researchers and staff of the Machine León-Bravo, V., Caniato, F., Caridi, M., 2019. Sustainability in multiple stages of the
Intelligence Research Lab in the Department of Computer Science, food supply chain in Italy: practices, performance and reputation. Oper. Manage.
University of Kerala, and the directorate of agriculture in Wasit Gov- Res. 12, 40–61. http://dx.doi.org/10.1007/s12063-018-0136-9.
ernorate in Iraq for supporting all the resources and help required to Li, Y., Cao, Z., Lu, H., Xu, W., 2020. Unsupervised domain adaptation for in-field cotton
accomplish such a study. boll status identification. Comput. Electron. Agric. 178, 105745. http://dx.doi.org/
10.1016/j.compag.2020.105745.
Liakos, K.G., Busato, P., Moshou, D., Pearson, S., Bochtis, D., 2018. Machine learning in
Funding
agriculture: A review. Sensors 18 (8), 2674. http://dx.doi.org/10.3390/s18082674.
Lin, Y.-P., Petway, J.R., Anthony, J., Mukhtar, H., Liao, S.-W., Chou, C.-F., Ho, Y.-F.,
This research was funded by the Indian Council for Cultural Rela- 2017. Blockchain: The evolutionary next step for ICT e-agriculture. Environments
tions ICCR with Ref.No.: RBIS0PFMS01. 4 (3), 50. http://dx.doi.org/10.3390/environments4030050.
Mahmud, M.S., Zahid, A., Das, A.K., 2023. Sensing and automation technologies for
Ethics approval ornamental nursery crop production: current status and future prospects. Sensors
23 (4), 1818. http://dx.doi.org/10.3390/s23041818.
Mor, S., Madan, S., Prasad, K.D., 2021. Artificial intelligence and carbon footprints:
This manuscript does not contain any studies with animals, and no
Roadmap for Indian agriculture. Strateg. Change 30 (3), 269–280. http://dx.doi.
harm has been done during any phase of the work. org/10.1002/jsc.2409.
Priyadarshini, P., Abhilash, P.C., 2020. Policy recommendations for enabling transition
Consent for publication towards sustainable agriculture in India. Land Use Policy 96, 104718. http://dx.
doi.org/10.1016/j.landusepol.2020.104718.
All authors have provided written consent for publishing this Ramesh, S., Vydeki, D., 2020. Recognition and classification of paddy leaf diseases
using optimized deep neural network with jaya algorithm. Inf. Process. Agricult. 7
manuscript in this journal.
(2), 249–260. http://dx.doi.org/10.1016/j.inpa.2019.09.002.
Rodríguez, J.P., Corrales, D.C., Aubertot, J.-N., Corrales, J.C., 2020. A computer vision
References system for automatic cherry beans detection on coffee trees. Pattern Recognit. Lett.
136, 142–153. http://dx.doi.org/10.1016/j.patrec.2020.05.034.
Abioye, E.A., Hensel, O., Esau, T.J., Elijah, O., Abidin, M.S.Z., Ayobami, A.S., Yer- Sadowski, S., Spachos, P., 2020. Wireless technologies for smart agricultural monitoring
ima, O., Nasirahmadi, A., 2022. Precision irrigation management using machine using internet of things devices with energy harvesting capabilities. Comput.
learning and digital farming solutions. AgriEngineering 4 (1), 70–103. http://dx.
Electron. Agric. 172, 105338. http://dx.doi.org/10.1016/j.compag.2020.105338.
doi.org/10.3390/agriengineering4010006.
Sanga, S., Mero, V., Machuve, D., Mwanganda, D., 2020. Mobile-based deep learn-
Ahmad, U., Sharma, L., 2023. A review of best management practices for potato
ing models for banana diseases detection. http://dx.doi.org/10.48550/arXiv.2004.
crop using precision agricultural technologies. Smart Agric. Technol. 100220. http:
03718, arXiv preprint arXiv:2004.03718.
//dx.doi.org/10.1016/j.atech.2023.100220.
Albaaji, G.F., Chandra, S., 2024. Blockchain technology in agriculture: Digitizing the Seyar, M.H., Ahamed, T., 2023. Development of an IoT-based precision irrigation
Iraqi agricultural environment. Environ. Dev. Sustain. 1–12. http://dx.doi.org/10. system for tomato production from indoor seedling germination to outdoor field
1007/s10668-024-04623-4. production. Appl. Sci. 13 (9), 5556. http://dx.doi.org/10.3390/app13095556.
Albaaji, G.F., Vinod Chandra, S.S., 2023. Artificial intelligence SoS framework for Sharafudeen, M., Vinod Chandra, S.S., 2023. Multimodal siamese framework for
sustainable agricultural production. Comput. Electron. Agric. 213, 108182. http: accurate grade and measure estimation of tropical fruits. IEEE Trans. Ind. Inform.
//dx.doi.org/10.1016/j.compag.2023.108182. http://dx.doi.org/10.1109/TII.2023.3316182.

10
Vinod Chandra S.S. et al. Computers and Electronics in Agriculture 226 (2024) 109386

Sreekantha, D., Kavya, A., 2017. Agricultural crop monitoring using IOT-a study. In: Victor, N., Maddikunta, P.K.R., Mary, D.R.K., Murugan, R., Chengoden, R.,
2017 11th International Conference on Intelligent Systems and Control. ISCO, IEEE, Gadekallu, T.R., Rakesh, N., Zhu, Y., Paek, J., 2024. Remote sensing for agriculture
pp. 134–139. http://dx.doi.org/10.1109/ISCO.2017.7855968. in the era of industry 5.0–A survey. IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens.
Subeesh, A., Mehta, C., 2021. Automation and digitization of agriculture using artificial http://dx.doi.org/10.1109/JSTARS.2024.3370508.
intelligence and internet of things. Artif. Intell. Agric. 5, 278–291. http://dx.doi. Virnodkar, S.S., Pachghare, V.K., Patil, V., Jha, S.K., 2020. Remote sensing and machine
org/10.1016/j.aiia.2021.11.004. learning for crop water stress determination in various crops: a critical review.
Talaviya, T., Shah, D., Patel, N., Yagnik, H., Shah, M., 2020. Implementation of artificial Precis. Agric. 21 (5), 1121–1155. http://dx.doi.org/10.1007/s11119-020-09711-9.
intelligence in agriculture for optimisation of irrigation and application of pesticides WDI, 2023. World bank WDI - ICT access.
and herbicides. Artif. Intell. Agric. 4, 58–73. http://dx.doi.org/10.1016/j.aiia.2020. Weiss, M., Jacob, F., Duveiller, G., 2020. Remote sensing for agricultural applications:
04.002. A meta-review. Remote Sens. Environ. 236, 111402. http://dx.doi.org/10.1016/j.
Tewari, V., Pareek, C., Lal, G., Dhruw, L., Singh, N., 2020. Image processing based rse.2019.111402.
real-time variable-rate chemical spraying system for disease control in paddy crop. world bank, 2023. World bank digital development report.
Artif. Intell. Agric. 4, 21–30. http://dx.doi.org/10.1016/j.aiia.2020.01.002.

11