Content uploaded by Sujata Dash
Author content
All content in this area was uploaded by Sujata Dash on Jan 13, 2023
Content may be subject to copyright.
CHAPTER
Internet of things (IoT) and data
analytics in smart agriculture:
Benefits and challenges 1
Biswaranjan Acharya
a
, Kyvalya Garikapati
d
, Anuradha Yarlagadda
b
, and Sujata Dash
c
School of Computer Engineering, KIIT Deemed to be University, Bhubaneswar, Odisha, India
a
Gayathri Vidhya
Parishad College of Engineering (A), Visakhapatnam, AP, India
b
Department of Computer Application, North Orissa
University, Baripada, Odisha, India
c
School of Law, KIIT Deemed to be University, Odisha, India
d
1. Introduction
With its diverse climate and soil, India has for many years been categorized as an agrarian economy,
with 49% of the workforce truly based on agriculture and the allied sector. Indian agriculture feeds
around 1.3 billion people. The government from the first five-year plan onward has consistently en-
couraged agriculture by introducing many policies. Despite this, this sector has not progressed as it
could because of fluctuations in the climate and many other reasons; this has resulted in fluctuations
in supply and poor productivity. Many nonsustainable practices have been adopted such as the use of
dangerous pesticides, inorganic fertilizers, unorganized resource management, poor water manage-
ment, small unit cropping, nonadaptation of technical developments etc. The Committee on Doubling
Farmer Income headed by Ashok Dalwai researched the problems and suggested measures to increase
productivity. The committee submitted the final report in September 2018, recommending a technol-
ogy based on sensor drones and big data analytics, GIS mapping technology, and vessel mapping
technology.
“The ability to filter a multitude of information into the ‘right and relevant’ will require extensive
deploying of technologies such as artificial intelligence (AI). This will feed on big data from agri-
culture and rely on the Internet of Things (IoT) and a Web of Things (WOT) to connect between
people-people, people-devices, and devices-devices. Robots and intelligent sensors are already mak-
ing a foray into the agricultural landscape. A sampling of technologies and the possible areas of use
(are) listed in Chapters 7 and 8 of Volume XII” [1].
The committee has also recommended organizing regular competitions and research programs in non-
agricultural universities across India to encourage innovations in software and hardware technologies
as well as application systems, including sensors, automation, robotics, AI, drones, alternative
technologies, etc.
The finance minister Nirmala Sitharaman in the budget session of 2018–19 directed NITI Aayog to
develop a national strategy for AI. Accordingly NITI Aayog submitted a report (discussion paper) in
AI, Edge and IoT-based Smart Agriculture. https://doi.org/10.1016/B978-0-12-823694-9.00013-X
Copyright #2022 Elsevier Inc. All rights reserved.
3
June 2018 after thorough research on AI [2]. NITI Aayog in this research collaborated with the Wadh-
wani Institute for AI, VideoKen, MIT Media Labs, INTEL, IBM, Nvidia, etc. This extensive research
paper covered the scope of AI in many fields such as healthcare, agriculture, education, smart city and
infrastructure, smart mobility, and transportation.
In India, more than 30 million farmers own smartphones, and this figure may increase three times by
2020. Around 315 million people may be using the Internet by 2020. A study conducted by Accenture
says that AI-based digital farming and connected farm services may impact 70 million farmers in India
by 2020, which may cause an income increase of $9 billion. These predictions are not completely hy-
pothetical, as they are in play now. This vast digital ecosystem includes traditional original equipment
manufacturers (OEM), global software companies, open-source platforms, cloud providers, research
and development institutions, certain start-ups, and others.
1.1 Understanding AI
Machine learning means “the ability to learn without being explicitly programmed”(Artur Samuel,
1959). Machine learning involves the use of algorithms to parse data and learn from that, and making
a determination or prediction as a result. Instead of hand coding software libraries with well-defined
specific instructions for a particular task, the machine gets “trained” using large amounts of data and
algorithms, and in turn gains the capability to perform specific tasks.
Deep learning is a technique for implementing machine learning. Deep learning was inspired by the
structure and function of the brain, specifically the interconnecting of many neurons.
“A computer would deserve to be called as intelligent if it would deceive a human into believing
that it was human.”–Alan Turing (British mathematician, 1912–54).
AI is the science of making systems intelligent. There is no one definition for AI because the word
intelligent is so abstract in nature. What was intelligent in the 18th century is not intelligent in later
ages. But there are certain tests to know about intelligence, including the Turing test. According to
this test, if a human being was allowed to chat with a human being and a computer program without
knowing who is who, and if that human being failed to discover which one was the computer program,
then such a computer system is treated as intelligent. So, human judgment is used as a variable to find
the level of intelligence of a computer program. The other way to weigh the intelligence is comparing
the efficiency and ability of a system program with human efficiency and ability. So, AI systems are
systems that perform functions like a well-trained human being. There are two kinds of AI: weak AI
and strong AI.
The most recent achievements of AI are completely based on machine learning from large datasets.
The algorithms of machine learning enable predictions and recommendations by processing data and
experience, not by just programming instructions. These algorithms can even adopt new data and ex-
periences and work better over a period of time.
Today, AI is a global strategy. Almost all countries have national AI strategies, including the United
States in December 2016, France in January 2017, Japan in March 2017, China in July 2017, and the
United Kingdom (UK) in November 2017. All these countries identified agriculture as a potential de-
velopment for the use of AI. All these countries also have allotted funds for the research and devel-
opment of AI. Universities all over these countries published research papers on AI between 2010
and 2016. US universities such as the Massachusetts Institute of Technology, Carnegie Mellon Uni-
versity, and Stanford University have contributed to the development of AI through teaching research.
4 CHAPTER 1 IoT and data analytics in smart agriculture
Many countries have come up with structured governance. The National Science and Technology
Council in the United States [3], the Strategic Council for AI Technology in Japan, and the AI Council
in the UK are part of the planning and designing of AI initiatives. These bodies are working with in-
dustry experts. The UK has a separate department that will be collaborating with all other departments;
China and Japan also have separate ministries for AI. Other than private investment in AI, government
salso hope to capitalize on the benefits from AI, such as economic benefits, social benefits, and
leadership.
2. IoT ecosystem in agriculture
Sustainable agriculture refers to all sorts of developments related to farming to marking or distribution
and reaching the end-user without any wastage. Pretty and Jules [1] defined the concepts, principles,
and evidence on this; the interested reader can go through this article for a detailed study.
The use of today’s advanced technology gives increased productivity in optimized resources uses.
That is possible if the farmer cultivates in a controlled and transparent way. A smart farm is an example
of sustainable agricultural productivity increases [4]. It not only gives shows the importance of in-
creased productivity, but it also emphasizes environmental sustainability along with data analytics,
where the task is to discover the requirements and predict the suitability of crops and fertilizers accord-
ing to the soil and environment. One of the challenges is plant diseases, which increase the loss to
farmers. The researchers Wang et al. [5] proposed using the IoT to detect plant disease and spray pes-
ticide where necessary. This would be achieve by data mining and modern machine learning ap-
proaches which given that article and also experimented. After the successful deployment of IoT in
urbanization now the matter of sustainability uses IoT in rapid urbanization and also in rural areas.
So now comes the environmental IoT, which provides sustainable smart city development that can
be used in every sector of human need; it also gives a future direction of Zero-IoT or Zero-SIT [5].
Agricultural production in India is decreasing daily for many reasons. To overcome this, farmers must
have adequate knowledge. Mohanraj et al. [6] defined in their article e-agricultural application based
on the framework of Knowledge Model (KM) using knowledge base that is the iterative model and it
has monitoring model which monitor agriculture environment requirement and crop. It gives better
result in terms of optimized cost, including manpower and increased production, so it works as a path-
finder to rural farmers.
2.1 Management techniques/systems (IoT and big data)
To implement smart agriculture, a wireless sensor network plays a vital role. A sensor senses the en-
vironment and passes that information on to the actuator for further processing; we can make an as-
sumption of smart irrigation. In a smart irrigation system, the sensor detects the need for watering by
sensing the moisture threshold level. There are may be so many works in this context in the different
procedure; Hari Ram and Vijay et al. defined how a mobile phone controls an irrigation system using
wireless technology [7]. Not only is an irrigation or watering system possible, but the whole agricul-
tural system will be modernized, which is known as smart agriculture, by using the IoT. In 2018, Elijah
et al. defined how the IoT combines data analytics to the great advantage of smart agriculture [8].Fur-
ther, that paper also describes an IoT ecosystem for smart agriculture, future trends, and challenges.
52 IoT ecosystem in agriculture
Both the integration of radio frequency identification (RFID) and sensors using IoT to monitor agri-
cultural field environment and uses resources optimistic way in the term of less manpower and getting
better productivity. Wasson, Tarushi et al. in 2017 defined this RFID integration into agriculture and
gave the results [9]. Due to the developments in technology, the terrain classification and estimation are
possible. This has helped to increase production by predicting which soil and environment are appro-
priate for which crops [9]. Big data and IoT nowadays don’t just facilitate smart agriculture, but they
also automate the nutrition-based recommendations and distribute the production. The readers can go
through the works from Ahrary et al. [10] for procedures and implementation. Worldwide, 50% of pro-
duction didn’t reach the customer or end user due to wastage at the farm itself. This is because of a lack
of transportation at the right time and mismanagement; it can be improved using IoT starting from reg-
ulating and controlling all the cultivation processes to reaching the end user [10]. A recent advancement
is that a surveillance system was added in the server room with an integrated automated farming system
where the user will monitor the whole system [11]. “Surprise: Agriculture is doing more with IoT In-
novation than most other industries” by Jahangir Mohammed and it can better result in agri-food busi-
ness chain [12]. It is also used in food firm automation, personalized nutrition, and for new solutions to
communication problems.
Agriculture is one of the important factors for sustainability. Besides production, nothing is really
sustained, so we must emphasize better productivity at low cost using smart methods, along with
analyzing and/or predicting the environment, crop, and farmer and consumer needs. In the year
2015 Baek, and Park [13] gave a sustainable development plan for Korea through Green-IT. It de-
scribes the use of big data and IoT in the agricultural sector within all the plan and policies for sus-
tainable city as a case study the author taken as Korea city. Really, without the sustainable
development of agriculture in rural areas, we dare not dream of sustainable development. The authors
Nomusa, and Kalezhi [14] took a rural area of South Africa and integrated many things such as crop
predicting or controlling, weather forecasting, animal and husbandry farming, rural financing, busi-
ness place identification, etc., with IoT for poverty reduction as well as better productivity with mod-
ern technology.
Sometimes may ask if it is helpful to the sustainable development agenda or it has some limitations
also. It hampers to earning sources of labor involved in the agricultural sector, technical factors or or-
ganizational factor. In this perspective as per a case study in the Chinese agricultural sector, Dlodlo
et al. [14] developed a framework named technology-organization-environment (TOE). It analyzed
the effect factor, not a single point it effects and also what are the reader of the benefits can find from
there and also recommends to use this better productivity with low cost in a smart and intelligent way.
Controlled irrigation is another challenge for eco-friendly production as it supports less use of fertilizer
or more use of organic fertilizer, which saves water, energy, and other natural resources. In the year of
2016 Srisruthi et al. [15] defined the use of green sensor technology which facilitates eco-friendly sus-
tainable agriculture which also remotely monitor and control the environment with less cost, less man-
power, and the most important thing is with any harm to the soil as well as the environment, that is, the
optimized use of resources.
2.2 Smart information systems (SIS) in agriculture
From different resources, large datasets will be collected and analyzed in agricultural big data analyt-
ics. The combination of big data and AI is called a smart information system. Its role is expected to be
6 CHAPTER 1 IoT and data analytics in smart agriculture
very important for the proposed growth in agriculture. Now, almost all agribusinesses are concentrating
more on data-driven agricultural solutions. The integration of agriculture and SIS deals with data re-
trieval, analysis, and prescription. Data retrieval is of several types such as soil moisture, rainfall and
climate changes, growth patterns of the crops, animal moments and grazing patterns, and market pric-
ing (Bennett, 2015). The proposed areas to develop through SIS are quality of life, food quality, bio-
diversity, food security, output, air quality, and water quality.
3. Benefits of IoT in agriculture
The introduction of AI and IoT to agriculture was aimed at a global increase of more than 50% to cater
to an additional 2 billion people by 2050. This will potentially address problems and challenges such as
poor demand prediction, improper assured irrigation, and unnecessary overuse of chemicals and fer-
tilizers while at the same time increasing yield. It can improve the harvest quality, tackle labor chal-
lenges, and forecast weather data (Fig. 1).
3.1 Remote sensing as a major tool in agriculture
The use of remote sensing is based on gathering information from far away. The whole process in the
IoT is controlled by remote sensing tools. To get better productivity in agriculture, the farmer needs to
monitor field from time to time. Jha et al. described how a farmer remotely monitors the whole process
that is, by using IoT [16]. In 2017, Takashi et al. proposed an integrated approach for environmental
monitoring and plant growth measurement that gives an affordable cost for smart and precise
FIG. 1
Different views of smart farming.
73 Benefits of IoT in agriculture
agriculture [17]. It senses the environmental soil, water level, and moisture for watering while also
estimating plant growth by depth and height through capturing photos with a high-resolution camera
using an IoT-based drone; if necessary, the farmer can replace some plants. In this way, the farmer or
observer observes all the plants and not a single plant is neglected or destroyed. So it increases pro-
duction with an affordable cost. Nowadays, due to underground sensor networks, real-time soil sensing
and monitoring are possible. So, it also helps precision agriculture. Vuran et al. defined the Internet of
Underground Things (IOUT), which is used for precision agriculture [18].
Researchers are tirelessly working to improve the system, and as a result, farmers can capture im-
ages and videos for processing in real time. This means that it can be controlled and monitored re-
motely. The details are described by Zhang et al. in their 2011 paper [19].
3.2 Weather forecasting as a prime IoT in agriculture
Different remote sensing applications are in use in agriculture. These remote sensing satellites are in
space to convey data on weather phenomena. Different meteorological institutions are working in the
field of weather forecasting. Advance knowledge on weather forecasting as well as warnings on up-
coming natural hazards such as floods, droughts, cyclones, etc., are based on the success of weather
forecasting. The growth of modern technology helps to analyze the huge amount of agricultural data
in accordance with accurate weather forecasting for precision agriculture [19]. In this system, an in-
tegrated IoT also automates the necessary steps according to the needs of that crop.
Integrated and Judicious Disaster Management Procedure Such as, Cyclonic, Flood and Draught
forecasting and mitigation. Rapid Information Processing, and Disbursal also now trying to solve using
IoT technique for easy, fast, and error free.
3.3 Agriculture drones
Now, some developed countries are using various types of drones inagriculture. This has many advantages,
as drones gathers knowledgethrough their cameras of the entirefield of crops from plowingto production in
the last stage. They gather information regarding soil moisture, weather conditions such as humidity, at-
mospheric pressure, clouds, rains, etc., associated with particular crops. Drones with integrated high-
resolution cameras, appropriate sensors, and proper microcontrollers have greater impact in precision ag-
riculture. They touch every part of the land appropriately, so this helps in the betterment of farming [20].
3.4 Crop monitoring
Crop monitoring is a whole process until the crop is yielded from the land. The process includes the
appropriate use of seeds, the moisture needed for a crop, and the amount of pesticides and insecticides
for that crop. In an agricultural system, there are many stages of crop farming and there are many
threats. The farmers need to check on the soil, moisture, weather conditions, pesticide control, level
of watering, etc. This can be achieved using sensors integrated with microcontrollers, which enables
the appropriate services [21]. According to the soil type or the fertility of soil, modern technology pro-
vides different data mining techniques to classify the soil type and predict which crop will give better
production [22]. Combination of image processing and IoT provide better plant monitoring which
completely implementation procedure defined by Kapoor et al. [23].
8 CHAPTER 1 IoT and data analytics in smart agriculture
3.5 Smart irrigation
Smart irrigation systems tailor watering schedules and run times automatically to meet specific landscape
needs. These controllers significantly improve outdoor water use efficiencies. A real-time smart irrigation
model controlled by a mobile phone through IoT and all sensor data generated stored cloud base system
which can be found in literature [23].Lietal.[24] proposed an intelligent way for sustainable irrigation
through a mobile phone, it integrates decision making using unsupervised machine techniques and data
analysis through cloud storage of how much required for which soil andenvironment. So farmersgets good
production without any water waste while also saving money on unnecessary irrigation. An IoT architec-
ture using a service-oriented system designed for agricultural purposeswas proposed by Cambra et al. [25].
They developed a model for a low-cost irrigation controller that also takes differentreal-time parametersas
the constraints such as direction of wind flow, speed, and pressure level.
3.6 Greenhouse monitoring and automation system
A green house is a glass house and all its activities can be controlled well by the use of remote sensing
and soil nutrients inside the house. The house is well connected by different scientific equipment. It is
good initiative than any other internet of things in case of agriculture because it occupied a limited area
and hence use of technology in agriculture can be a well access. A well-controlled greenhouse model
for precision agriculture has been proposed where zero tolerance of environment harmless with in-
creasing productivity and remains signified beauty in both day and night [25].Again due to nature’s
imbalance upon the environment the farming sector also fall in the loss. So to overcome this situation,
a combination of IoT and a greenhouse shows tremendous development of production as well as an
environment by the ability to control irrigation as well as the requirements of moisture and nutrients.
The design system and how the mobiles are controlled over the cloud are well explained in the proposed
model and experimental results [26].The large datasets generated from agriculture and green ouse mon-
itoring that should be well analyzed to make the right decisions. Already, Li et al. [27] gave the concept
of embedded intelligence (EI) with different machine learning techniques to make a proper decision
that explained the proper management of a greenhouse. It is also designed in an integrated manner with
a wireless sensor network (WSN) with a decision-making support system (DMSS), which is based on
knowledge supporting technology for extracting new knowledge from existing data. On the other hand
the fast-growing technology has much negativity toward to environment, so not only development the
world toward to sustainable development. So Maksimovic [28] given in his article emphasizing the
term Greening the future through technology, greening the economy and main motto to became smart
agricultural system for better productivity through optimized resources using Green-IoT (G-IoT). Two
important needs for human beings are food and healthcare, and then others may come. So Nandyala
et al. [29] developed two separate models for these two sectors that provide less energy consumption
and are integrated into the cloud with IoT in both cases. Patil et al. [30] also defined uses of 3S such as
remote sensing (RS), GIS, and GPS or geometrics causes more greenery to agriculture as well as greater
sustainable environment. It also uses a combination of cloud and different sensor technology; RFID and
WSN are completely automated to the agricultural environment.
Both combinations of IoT and data analytics (DA) are implemented in Nigeria as case studies [30].
This solves the traditional problem there: they weren’t able to meet domestic needs and didn’t get quality
food. Adopting the proposed model solves both these problems. Th e IoT and big data analytics don’t only
93 Benefits of IoT in agriculture
control irrigation or plants, but they also have a large impact on the whole agricultural system. A case
study carried out by Popovi
c et al. [31] developed a model on a private IoT that provides smart irrigation,
plant control, pesticide monitoring, marine environment assessment, and fish/mussel farming monitor-
ing. The potato is one of the most influential foodstuffs for society and the economy, so many researchers
are working to increase productivity, distribution, and delivery to end users without waste. Ra d et al. [32]
took the potato as a case study and developed a cyberphysical system for sustainable agriculture. It pro-
vides an integrated approach of precision agriculture management with increasing production.
4. Open issues and key challenges in the adoption of IoT in agriculture
The key challenges of IoT adoption in agriculture are: 1. Reliability of systems; 2. Security informa-
tion; 3. Data privacy and provenance; 4. Social acceptance; 5. Accessibility of reliable information; 6.
Cost effectiveness; and 7. User friendliness (ease of use). Other than these, there are certain issues for
policy makers in dealing with IoT in agriculture. Such issues are discussed below. Out of many, the
following four are the major challenges in IoT systems.
4.1 Reliability
Many doubts have been raised on the application of machine learning in agriculture because the algo-
rithms used are better suited to small datasets. These algorithms cannot process big data, so the result
will be not accurate enough. This limitation has to be answered by the SIS industry in a better way.
Otherwise, the outcomes may be misleading with wrong findings and analysis. Sometimes, the out-
come of these small data analyses may also be wrong because of environmental conditions. In some
cases, these will give the wrong output because of some technical problems such as radio signals used
for communication may be interrupted, perhaps by an animal being too close to the sensor and blocking
the radio signals. Some circumstances lead to false readings such as with temperature, humidity, etc. In
some cases, data interpretation is very difficult because of local usage differences, so the data analysis
should be very contextual and unbiased. The inaccuracy of data given to farmers may lead to loss of
product and damage to livestock, which will show a negative impact on their income.
4.2 Data privacy protection and issues of ownership
Many questions relating to ownership of data and the Intellectual property rights of such data owner is
also a big problem. The data belongs someone may be distributed to a third party these concerns about
distribution of data should be addressed properly. Because of this fear of data distribution and the usage
of such data against them will always be there in the minds of the farmers. This data transfer may be
prejudicial to them in future. They are also having some issues about the collection and circulation of
data to the regulatory bodies like government agencies which may result in increase of charges, fines
and levies. If the information caught in the hands of commodity traders, dealers and agents they will use
this against the farmers in fixing the prices. When the traders know about the availability, supply and
production exactly they will speculate the prices which will affect the farmers grossly [33].
Sometimes the data leak may leads sell unnecessary products back to them. Very big agricultural
farms data will influence the small farmers to buy a particular seed type, equipment or fertilizers, by
10 CHAPTER 1 IoT and data analytics in smart agriculture
which big agricultural farms may get some profit share from the concerned product company. So who
owns the data is very important concern.
Farmers own data this is the version of many ATPs but they may get a royalty free license to use the
data from the farmers so that they could use the data. If the farmers after sometimes wants to change
their ATP it may leads to breach of Contract which may attract huge penalties. So there is no protection
to the claim of ownership by the farmers. Sometimes the data provided by the farmers to the ATPs is
also not their own rather it is collective in such cases severe violations of Intellectual property rights
may cause [34].
4.3 Autonomy foreseeability and causation
The very important feature of AI is it functions very autonomously. AI is having a distinctive feature
that is forseeability and causation. It would do many complex functions very easily for example a driver
less car, performing and operation in the theatre, maintaining the portfolio following the market fluc-
tuations etc. The basic difference between the decision making capacity and process of humans and of
the AI may leads to completely make humans to predict the actions of the AI. Natural human beings are
limited and restricted by cognitive feature of the human brain. So the decisions of the human beings are
more of satisfactory solutions but not the optimal solutions. AI programs are having an ability to search
many possibilities which a human being cannot do in a given time. Therefore, AI systems can analyze
all potential solutions that human brains may not have considered and attempted to implement [32].
From the legal perspective, covering all the possibilities of AI actions in the legal framework is very
impracticable and highly imaginatory. In the legal principles of tortuous liability “an intervening force
or act that is deemed sufficient to prevent liability for an actor whose tortuous conduct was a factual
cause of harm” [35]. Here learning is not only the continuous learning of the AI but the information got
from the machine learning too which broadly effects on the final actions of the AI system.
4.4 Control
The problem with handling an autonomous AI is not only its foreseeability, but also its control. Some-
times to control in an AI technique which is very aunomous and where the actions are completely de-
pends on various factors is very difficult for the human beings. The loss of control may occur because of
various reasons that are not in the control of humans, such as malfunctions because of corrupted files or
some damage to input equipment, security trespass, flawed programming, etc. Loss of control may be
classified into local control and general control. The loss of local control refers to that situation where
the owner of such an AI system who is legally liable will no longer have control of its operation. Loss of
general control refers to a situation where the AI cannot be controlled by any person. This type of loss of
control causes greater harm to the public, but if the objectives of the AI system are aligned in such a way
where it would never cause public loss, then no public loss will occur. Unfortunately, controlling or
regulating the alignment of interests and objectives of AI is very difficult and next to impossible,
as determining human values is impossible to define or assess [36]. The world experts in AI, the ac-
ademicians and entrepreneurs, have warned that stronger forms of AI may resist the efforts of humans
at governing their actions, possibly leading to human risk. Hence, the more sophisticated AIs are the
toughest for human interference because they are more autonomous. These sophisticated AIs improve
their own hardware and programming to gain more cognitive abilities. To understand this, the best
114 Open issues and key challenges in the adoption of IoT in agriculture
example is the “Flash Crash 2010,” where the algorithms used in trading stocks caused a massive eco-
nomic impact in a short period of time.
4.5 Opaque research and development
From the legal side of AI, most of the problems are not its feature but rather a way in which the Re-
search and Development are done. All the research is done on a single component of that AI system that
might be very far from another. Separate components of AI are designed in different places at different
times and without any coordination. As the AI systems are susceptible to reverse engineering, all the
research will be kept very secret from others. The research in AI happens at different levels and dif-
ferent scales. Even a small college with very few facilities could research in AI. Most of the research is
by big business entrepreneurs with large set-ups and great secrecy. So, the regulators are at the diffi-
culty in finding who is doing this research and where it is being done and to predicting the public risk of
AIs being developed [37].
5. Legal issues in regulating AI in agriculture
Even with all the problems stated above, to go for a standard regulatory framework, there are many
legal mechanisms that can be used to reduce public risks. Several problems that were identified earlier
are the simple gaps in the existing law, which can be filled in many ways. Giving an exact definition for
AI is difficult but giving a precise legal definition is possible. But trying to give a precise definition to
an imprecise thing is a challenge for all legal systems. A legal definition for determining liability or
regulation likely would be under-or overinclusive.
5.1 Torts and contracts
A tort is a civil wrong that gives a right to the opposite party to file a civil suit for a breach of duty owed
by the persons under the law. A contractual liability talks about a duty voluntarily accepted; the parties
may limit their liability in the terms of the contract. A civil action is possible against the parties in both
cases but it requires the standard proof of balance of probabilities. The established legal principles in
tort and contract law, whether codified or not, help people get damages or a remedy against the owner
or user of AI.
Product defects are covered under tort law and consumer protection laws. These laws cover agri-
cultural products, including movable and immovable products. It covers finished products and unfin-
ished products, and raw materials are also part of the product. Safety issues will fall under defective
products. The standard of the defect is very objective. So, in determining the defectiveness of an AI
program, the protection provided by such a program is compared with human beings. A claim for dam-
ages will be filed by the person affected by the malfunction. The gross loss will be calculated to provide
proportionate damages to the victim. The relationship between the damage and defect should be proved
before the forum. The manufacturer of agribots may claim that the defect was not there when they were
supplied or some issue of malfunction, or he may defend by saying the agribot was used for a different
purpose altogether. The sellers or manufacturers may also defend themselves on the grounds that the
safety fault was because of a rule or regulation that prohibits some technology by which the risk may be
reduced.
12 CHAPTER 1 IoT and data analytics in smart agriculture
5.2 Crimes
The proceedings in crimes are initiated by the state as a crime against persons is a crime against society.
Here, the crime should be proved by the prosecution before the court beyond a reasonable doubt. The
punishments or penalties are stipulated by law. The concept of criminal corporate responsibility, which
imposes vicarious liability on the corporations for the crimes committed by the directors, managers,
and employees, is not well established in all countries. Some countries such as Germany do not impose
criminal liability on corporations. In India, the concept of vicarious liability is applicable on corporate
criminal liability; courts have limited criminal liability only to the extent of fines because a corporation
cannot be sentenced to imprisonment (Aneeta Hada vs. Godfather Travels and Tours Pvt. Ltd, [2012 5
(SCC 661).
5.3 Law relating to accidents, health, and safety
Unlike the UK, India does not have one single consolidated act to cover health and safety in the work-
place. There are many legislation talks about health and safety at workplace like Factories Act, Mines
Act, Workmen Compensation Act, Employees State Insurance Act, etc. Till today agriculture sectors in
developing countries are unorganized. The accidents caused in the workplace are covered by all these
enactments. If the accident is caused to a normal human by AI-programmed agribots, then there are
several claims available in civil and criminal court. Here, the owner of the agribots is liable. Sometimes
this claim may extend to the manufacturer of the agribot if the accident was caused by any defect in the
product. The Motor Vehicles Act also talks about accidents and claims. So, if any AI is attached to any
motor vehicle, then this act provides relief for the persons who suffered physical injury or damage to
their property. The government of India and the Ministry of Labor and Employment issued a national
policy on safety, health, and environment at the workplace by using a directive principle of state policy.
5.4 Accidents and negligence
A legal action for negligence could be possible against operators, farmers and their agents, agricultural
contractors, and manufacturers for physical injuries, loss, or damages to property resulting from the
negligent use of agribots or other kind of AI-based equipment. The difference between the claims
is based on strict liability (liability without fault). Negligence is the latter one, and fault must be proved
on the part of the person being sued. The duty of care and the breach of such duty should be proved by
the claimant. However, the duty of care may shift to different persons in different cases. For example, in
some cases the duty of care is on the manufacturer if there is a defect in the product. This duty of care
may be in some cases on the owner of the product, as the owner has a duty of care to hand over the
product to his agent or contractual worker who is actually using it.
5.5 Environmental laws
The environmental laws (Environment Protection Act, Air Act, Water Act, and Forest Act) govern ag-
riculture from time to time. The farmers and other stakeholders of agriculture have to follow all these
laws and the policies issued under such laws. These laws prohibit the use of pesticides that pollute the
air, water, and environment. The strict liability principle and the polluters pay principle are strictly
implemented on the agriculture sector. Hence, the use of AI-based robotics for application of pesticides
135 Legal issues in regulating AI in agriculture
and harmful fertilizers may affect the food chain, neighboring farms, and the staff; it will broadly affect
the water supply and the public environment as a whole.
Many other areas of laws will also be applied to AI in agriculture such as cyber law, intellectual
property law, and competition law (on manufacturers, traders, and suppliers).
6. Conclusion
The application of IoT-based systems in agriculture is the need of the hour against the backdrop of
increasing needs for better food security. As we have seen, IoT-based agriculture results in sustainable
development in the agricultural sector while also being cost-effective. In the wake of urbanization,
globalization, and technology growth, the use of modern AI systems is very effective. The efficiency
and effectiveness of these IoT-based systems have provided consistent growth in the income of farmers
and thereby resulted in the economic growth of countries. The smart farming and greenhouse mainte-
nance systems have also shown extraordinary growth. The vision of the leaders of countries, business
enterprises, academicians, and researchers in AI is becoming more realistic day by day. To be more
effective, the AI system should address ethical issues such as transparency and accountability. The
IoT should be used to serve the people and the planet. It should focus on the principles of fundamental
human dignity, privacy, freedom, cultural and gender diversity, etc. The research programs funded by
the government should necessarily have workshops for farmers to understand the systems. To address
and to fill the gaps in the regulatory mechanism on AI, governments should initiate better preventive
measures, keeping in view the vulnerability of small-scale farmers.
To address the legal issues and challenges, the government has to consult and collaborate with all
the stakeholders to bring an effective legislative and executive framework. As this area is so promising
and has the scope for many developments, looking into the legislative framework should be very flex-
ible to accommodate all the future technical developments.
References
[1] http://agricoop.nic.in/sites/default/files/DFI_Volume_14.pdf.
[2] https://niti.gov.in/national-strategy-artificial-intelligence.
[3] https://www.whitehouse.gov/ostp/nstc/.
[4] J. Pretty, Agricultural sustainability: concepts, principles and evidence, Philos. Trans. R. Soc., B 363
(1491) (2008) 447–465.
[5] X.F. Wang, et al., Monitoring and Discrimination of Plant Disease and Insect Pests based on agricultural IOT,
in: International Conference on Information Technology and Management Innovation (ICITMI 2015), 2015.
[6] I. Mohanraj, K. Ashokumar, J. Naren, Field monitoring and automation using IOT in agriculture domain,
Procedia Comput. Sci. 93 (2016) 931–939.
[7] R. Hari, V. Vijay, et al., Regulation of water in agriculture field using Internet of Things, in: Technological
Innovation in ICT for Agriculture and Rural Development (TIAR), IEEE, 2015, p. 2015.
[8] O. Elijah, et al., An overview of internet of things (IoT) and data analytics in agriculture: benefits and chal-
lenges, IEEE Internet Things J. (2018).
[9] Wasson, Tarushi, et al. "Integration of RFID and sensor in agriculture using IOT." Smart Technologies for
Smart Nation (SmartTechCon), 2017 International Conference On. IEEE 2017.
14 CHAPTER 1 IoT and data analytics in smart agriculture
[10] A. Ahrary, R.D.A. Ludena, Big data application to the vegetable production and distribution system, in:
Signal Processing & its Applications (CSPA), 2014 IEEE 10th International Colloquium on, IEEE,
2014.
[11] J. Shenoy, Y. Pingle, IoT in agriculture, in: Computing for Sustainable Global Development (INDIACom),
2016 3rd International Conference on, IEEE, 2016.
[12] R. Venkatesan, A. Tamilvanan, A sustainable agricultural system using IoT, in: Communication and Signal
Processing (ICCSP), 2017 International Conference on, IEEE, 2017.
[13] H. Baek, S.-K. Park, Sustainable development plan for Korea through expansion of green IT: policy issues for
the effective utilization of big data, Sustainability 7 (2) (2015) 1308–1328.
[14] N. Dlodlo, J. Kalezhi, The internet of things in agriculture for sustainable rural development, in: Emerging
Trends in Networks and Computer Communications (ETNCC), 2015 International Conference on, IEEE,
2015.
[15] S. Srisruthi, et al., Sustainable agriculture using eco-friendly and energy efficient sensor technology, in: Re-
cent Trends in Electronics, Information & Communication Technology (RTEICT), IEEE International Con-
ference on, IEEE, 2016.
[16] R.K. Jha, et al., Field monitoring using IoT in agriculture, in: Intelligent Computing, Instrumentation and
Control Technologies (ICICICT), 2017 International Conference on, IEEE, 2017.
[17] T. Okayasu, et al., Affordable field environmental monitoring and plant growth measurement system for
smart agriculture, in: Sensing Technology (ICST), 2017 Eleventh International Conference on,IEEE,
2017.
[18] M.C. Vuran, et al., Internet of underground things: sensing and communications on the field for precision
agriculture, IEEE 4th World Forum on Internet of Things (WF-IoT), IEEE, 2018, pp. 586–591.
[19] J. Zhang, et al., Research of real-time image acquisition system based on ARM 7 for agricultural environ-
mental monitoring, in: Remote Sensing, Environment and Transportation Engineering (RSETE), 2011 Inter-
national Conference on, IEEE, 2011.
[20] M.R. Bendre, R.C. Thool, V.R. Thool, Big data in precision agriculture: weather forecasting for future farm-
ing, in: Next Generation Computing Technologies (NGCT), 2015 1st International Conference on, IEEE,
2015.
[21] Saha, Arnab Kumar, et al. "IOT-based drone for improvement of crop quality in agricultural field." Comput-
ing and Communication Workshop and Conference (CCWC), 2018 IEEE 8th Annual. IEEE, 2018.
[22] D.K. Sreekantha, A.M. Kavya, Agricultural crop monitoring using IOT-a study, in: Intelligent Systems and
Control (ISCO), 2017 11th International Conference on, IEEE, 2017.
[23] A. Kapoor, et al., Implementation of IoT (Internet of Things) and Image processing in smart agriculture, in:
Computation System and Information Technology for Sustainable Solutions (CSITSS), International Confer-
ence on, IEEE, 2016.
[24] C. Li, et al., Mobile application based sustainable irrigation water usage decision support system: an intel-
ligent sensor CLOUD approach, Sensors (2013) 2013. IEEE.
[25] C. Cambra, et al., An IoT service-oriented system for agriculture monitoring, in: Communications (ICC),
2017 IEEE International Conference on, IEEE, 2017.
[26] R.K. Kodali, V. Jain, S. Karagwal, IoT based smart greenhouse, in: Humanitarian Technology Conference
(R10-HTC), 2016 IEEE Region 10, IEEE, 2016.
[27] Li, Li, et al. "Application of IoT technology in greenhouse management." 2012 Dallas, Texas, July
29–August 1, 2012. American Society of Agricultural and Biological Engineers, 2012.
[28] M. Maksimovic, Greening the future: green Internet of Things (G-IoT) as a key technological enabler of sus-
tainable development, in: Internet of Things and Big Data Analytics toward Next-Generation Intelligence,
Springer, Cham, 2018, pp. 283–313.
[29] C.S. Nandyala, H.-K. Kim, Green IoT agriculture and healthcare application (GAHA), Int J Smart Home 10
(4) (2016) 289–300.
15References
[30] V.C. Patil, et al., Internet of things (IoT) and cloud computing for agriculture: an overview, Proceedings of
agro-informatics and precision agriculture (AIPA 2012), 2012, pp. 292–296.
[31] T. Popovi
c, et al., Architecting an IoT-enabled platform for precision agriculture and ecological monitoring:
a case study, Comput. Electron. Agric. 140 (2017) 255–265.
[32] C.-R. Rad, et al., Smart monitoring of potato crop: a cyber-physical system architecture model in the field of
precision agriculture, Agric. Agric. Sci. Procedia 6 (2015) 73–79.
[33] Stuart Russell, Of Myths and Moonshine, EDGE, http://edge.org/conversation/jaron_lanier-the-myth-of-
ai#26015 [https://perma.cc/PLG8-RWBZ];.
[34] Nate Soares & Benja Fallenstein, Machine Intelligence res. Inst., Aligning Superintelligence With Human
Interests: A Technical Research Agenda 2(2014). https://intelligence.org/files/TechnicalAgenda.pdf
[https://perma.cc/2XQT-NEXV].
[35] M.B. Jack, The Path of Robotics Law, Calif. L. Rev. Circuit 45 (2015) 52. http://www.californialawreview.
org/wp-content/uploads/2015/06/Balkin-Circuit.pdf.
[36] Restatement (THIRD) OF TORTS: APPORTIONMENT OF LIABILITY §§ 10–17 (“Liability of Multiple
Tortfeasors for Indivisible Harm”); id. §§ 22–23 (“Contribution and Indemnity”);.
[37] L. Muehlhauser, N. Bostrom, Why We Need Friendly AI, 13 Think 41, 2014, pp. 41–43.
16 CHAPTER 1 IoT and data analytics in smart agriculture
