SPECIAL SECTION ON NEW TECHNOLOGIES FOR SMART FARMING 4.

0:
RESEARCH CHALLENGES AND OPPORTUNITIES

Received July 7, 2019, accepted July 19, 2019, date of publication August 1, 2019, date of current version September 23, 2019.
Digital Object Identifier 10.1109/ACCESS.2019.2932609

Internet-of-Things (IoT)-Based Smart Agriculture:

Toward Making the Fields Talk
MUHAMMAD AYAZ 1 , (Senior Member, IEEE),
MOHAMMAD AMMAD-UDDIN 1 , (Senior Member, IEEE),
ZUBAIR SHARIF2 , ALI MANSOUR3 , (Senior Member, IEEE),
AND EL-HADI M. AGGOUNE1 , (Senior Member, IEEE)
1 SensorNetworks and Cellular Systems Research Center, University of Tabuk, Tabuk 71491, Saudi Arabia
2 CS Department, COMSATS University Islamabad, Sahiwal 57000, Pakistan
3 Lab-STICC, UMR 6285 CNRS, ENSTA Bretagne, 29806 Brest, France

Corresponding author: Muhammad Ayaz (ayazsharif@ut.edu.sa)

This work was supported in part by the SNCS Research Center and in part by the Deanship of Scientific Research at the University of
Tabuk, Tabuk, Saudi Arabia.

ABSTRACT Despite the perception people may have regarding the agricultural process, the reality is that
today’s agriculture industry is data-centered, precise, and smarter than ever. The rapid emergence of the
Internet-of-Things (IoT) based technologies redesigned almost every industry including ‘‘smart agriculture’’
which moved the industry from statistical to quantitative approaches. Such revolutionary changes are
shaking the existing agriculture methods and creating new opportunities along a range of challenges.
This article highlights the potential of wireless sensors and IoT in agriculture, as well as the challenges
expected to be faced when integrating this technology with the traditional farming practices. IoT devices
and communication techniques associated with wireless sensors encountered in agriculture applications are
analyzed in detail. What sensors are available for specific agriculture application, like soil preparation, crop
status, irrigation, insect and pest detection are listed. How this technology helping the growers throughout
the crop stages, from sowing until harvesting, packing and transportation is explained. Furthermore, the use
of unmanned aerial vehicles for crop surveillance and other favorable applications such as optimizing crop
yield is considered in this article. State-of-the-art IoT-based architectures and platforms used in agriculture
are also highlighted wherever suitable. Finally, based on this thorough review, we identify current and future
trends of IoT in agriculture and highlight potential research challenges.

INDEX TERMS Food quality and quantity, Internet-of-Things (IoTs), smart agriculture, advanced agricul-
ture practices, urban farming, agriculture robots, automation, future food expectation.

I. INTRODUCTION tion predicted to be urban until 2050 (currently 49%) [3].

To improve the agricultural yield with fewer resources Furthermore, income levels will be multiples of what they are
and labor efforts, substantial innovations have been made now, which will drive the food demand further, especially in
throughout human history. Nevertheless, the high population developing countries. As a result, these nations will be more
rate never let the demand and supply match during all these careful about their diet and food quality; hence, consumer
times. According to the forecasted figures, in 2050, the world preferences can move from wheat and grains to legumes
population is expected to touch 9.8 billion, an increase of and, later, to meat. In order to feed this larger, more urban,
approximately 25% from the current figure [1]. Almost the and richer population, food production should double by
entire mentioned rise of population is forecasted to occur 2050 [4], [5]. Particularly, the current figure of 2.1 billion
among the developing countries [2]. On the other side, tons of annual cereal production should touch approximately
the trend of urbanization is forecasted to continue at an 3 billion tons, and the annual meat production should increase
accelerated pace, with about 70% of the world’s popula- by more than 200 million tons to fulfill the demand of
470 million tons [6], [7].
The associate editor coordinating the review of this article and approving Not only for food, but crop production is becoming equally
it for publication was Kun Mean Hou. critical for industry; indeed crops like cotton, rubber, and gum

VOLUME 7, 2019 This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see http://creativecommons.org/licenses/by/4.0/ 129551
 M. Ayaz et al.: IoT-Based Smart Agriculture: Toward Making the Fields Talk

are playing important roles in the economies of many nations. constantly with higher accuracy and are able to, most impor-
Furthermore, the food-crops-based bioenergy market started tantly, detect early stages of unwanted state. This is the reason
to increase recently. Even before a decade, only the produc- why modern agriculture involves the usage of smart tools
tion of ethanol utilized 110 million tons of coarse grains and kits, from sowing to crop harvesting and even during
(approximately 10% of the world production) [7], [8]. Due storage and transportation. Timely reporting using a range of
to the rising utilization of food crops for bio-fuel production, sensors makes the entire operation not only smart but also cost
bio-energy, and other industrial usages, food security is at effective due to its precise monitoring capabilities. Variety of
stake. These demands are resulting in a further increase of autonomous tractors, harvesters, robotic weeders, drones, and
the pressure on already scarce agricultural resources. satellites currently complement agriculture equipment. Sen-
Unfortunately, only a limited portion of the earth’s surface sors can be installed and start collecting data in a short time,
is suitable for agriculture uses due to various limitations, like which is then available online for further analyses nearly
temperature, climate, topography, and soil quality, and even immediately. Sensor technology offers crop and site-specific
most of the suitable areas are not homogenous. When zoom- agriculture, as it supports precise data collection of every site.
ing the versatilities of landscapes and plant types, many new Recently, the Internet-of-Things (IoT) is beginning to
differences start to emerge that can be difficult to quantify. impact a wide array of sectors and industries, ranging
Moreover, the available agricultural land is further shaped by from manufacturing, health, communications, and energy
political and economic factors, like land and climate patterns to the agriculture industry, in order to reduce inefficiencies
and population density, while rapid urbanization is constantly and improve the performance across all markets [12]–[16].
posing threats to the availability of arable land. Over the past If looking closely, one feels that the current applications are
decades, the total agriculture land utilized for food production only scratching the surface and that the real impact of IoT and
has experienced a decline [9]. In 1991, the total arable area its uses are not yet witnessed. Still, considering this progress,
for food production was 19.5 million square miles (39.47% of especially in the near past, we can predict that IoT technolo-
the world’s land area), which was reduced to approximately gies are going to play a key role in various applications of the
18.6 million square miles (37.73% of the world’s land area) agriculture sector. This is because of the capabilities offered
in 2013 [10]. As such, the gap between demand and supply by IoT, including the basic communication infrastructure
of food is becoming more significant and alarming with the (used to connect the smart objects—from sensors, vehicles,
passage of time. to user mobile devices—using the Internet) and range of
Further examination showed that every crop field has dif- services, such as local or remote data acquisition, cloud-
ferent characteristics that can be measured separately in terms based intelligent information analysis and decision making,
of both quality and quantity. Critical characteristics, like soil user interfacing, and agriculture operation automation. Such
type, nutrient presence, flow of irrigation, pest resistance, capabilities can revolutionize the agriculture industry which
etc., define its suitability and capability for a specific crop. probably one of most inefficient sectors of our economic
In most of situations, the differentiations of characteristics value chain today. To summarize this discussion, figure 1 pro-
can exist within a single crop field, even if the same crop is vides the main drivers of technology, while figure 2 highlights
being cultivated in entire farm; hence, site-specific analyses
are required for optimal yield production. Further, adding the
dimension of time, specific crops in the same field rotate
season-to-season and biologically reach different stages of
their cycle within a year in areas where locational and tempo-
ral differences result in specific growth requirements to opti-
mize the crop production. To respond to these demands with a
range of issues, farmers need new technology-based methods
to produce more from less land and with fewer hands.
Considering the standard farming procedures, farmers
need to visit the agriculture sites frequently throughout the
crop life to have a better idea about the crop conditions.
For this, the need of smart agriculture arises, as 70% of
farming time is spent monitoring and understanding the crop
states instead of doing actual field work [11]. Considering
the vastness of the agriculture industry, it incredibly demands
for technological and precise solutions with the aim of sus-
tainability while leaving minimum environmental impact.
Recent sensing and communication technologies provide a
true remote ‘‘eye in the field’’ ability in which farmers can
observe happenings in the field without being in the field.
Wireless sensors are facilitating the monitoring of crops FIGURE 1. Key drivers of technology in agriculture industry.

129552 VOLUME 7, 2019

M. Ayaz et al.: IoT-Based Smart Agriculture: Toward Making the Fields Talk

spoilage, climate changes, environmental pollution, and

urbanization.
• Strategies and policies that need to be considered when
implementing IoT-based technologies
• Critical issues that are left to solve and possible solutions
that are further required, while suggestions are provided
considering these challenges.
This article is a compendium of knowledge that can help
the researchers and agriculture engineers implementing the
IoT-based technologies to achieve the desired smart agri-
culture. The rest of this document is organized as follows.
Section II provides a deep overview of major applications
of IoT in agriculture and what we can achieve by utilizing
these technologies. Section III gives insight regarding the
role of IoT in advanced agriculture practices, like vertical
farming (VF), hydroponics, and phenotyping, to manage the
issues of increased urban population. Section IV highlights
various technologies and equipment, like sensors, robots,
tractors, and communication devices, being used to imple-
FIGURE 2. Major hurdle’s in technology implementation for smart ment IoT in this industry. Accepting the worth of UAVs in
agriculture. precision agriculture, Section V caters application achieve-
ments that are not possible even using other latest technolo-
the major hurdles of technology implementation in smart gies. Food safety and transportation are other critical areas
agriculture. requiring focus to overcome the hunger issues which did not
Researchers and engineers around the globe are propos- get the attention of researchers as it deserves. Section VI
ing different methods and architectures and based on that supplies the role of the IoT to ensure food quality for longer
suggesting a variety of equipment to monitor and fetch the periods and to deliver to remote areas. Section VII identifies
information regarding crop status during different stages, current and future trends of this technology in the crop indus-
considering numerous crop and field types. Focusing on try by highlighting potential research challenges. Finally,
the market demand, many leading manufactures are provid- Section VIII concludes this article.
ing a range of sensors, unmanned aerial vehicles (UAVs),
robots, communication devices, and other heavy machinery II. MAJOR APPLICATIONS
to deliver the sensed data. In addition, various commissions, By implementing the latest sensing and IoT technologies
food and agriculture organizations, and government bodies in agriculture practices, every aspect of traditional farming
are developing polices and guidelines to observe and regulate methods can be fundamentally changed. Currently, seamless
the use of these technologies in order to maintain food and integration of wireless sensors and the IoT in smart agricul-
environment safety [17]–[20]. ture can raise agriculture to levels which were previously
There are reasonable efforts that highlight the role of the unimaginable. By following the practices of smart agricul-
IoT in the agriculture industry, but most of the published ture, IoT can help to improve the solutions of many traditional
work focuses only on applications [10], [21], [22]. Most of farming issues, like drought response, yield optimization,
the existing articles either provide no insight or show limited land suitability, irrigation, and pest control. Figure 3 lists a
focus on the various IoT-based architectures, prototypes, hierarchy of major applications, services and wireless sensors
advanced methods, the use of IoT for food quality, and being used for smart agriculture applications. While, major
other future issues considering the latest facts and figures. instances in which the advanced technologies are helping
This manuscript examines the trends in IoT-based agricul- at various stages to enhance overall efficiency are discussed
ture research and reveals numerous key issues that must be below.
addressed in order to transform the agriculture industry by
utilizing the recent IoT developments. The major contribution A. SOIL SAMPLING AND MAPPING
of this article is to provide real insight regarding: Soil is the ‘‘stomach’’ of plants, and its sampling is the
• Expectations of the world from the agriculture industry first step of examination to obtain field-specific information,
• Very recent developments in IoT, both scholarly and in which is then further used to make various critical decisions
industry are highlighted and how these developments are at different stages. The main objective of soil analysis is to
helping to provide solutions to the agriculture industry. determine the nutrient status of a field so that measures can
• Limitations, the agriculture industry is facing. be taken accordingly when nutrient deficiencies are found.
• Role of IoT to cope these limitations and other issues Comprehensive soil tests are recommended on an annual
like resources shortage and their precise use, food basis, ideally in Spring; however, based on soil conditions and

VOLUME 7, 2019 129553

 M. Ayaz et al.: IoT-Based Smart Agriculture: Toward Making the Fields Talk

absorption rate, which ultimately help to minimize erosion,

densification, salinization, acidification, and pollution (by
avoiding excessive use of fertilizer). Lab-in-a-Box, a soil
testing tool kit developed by AgroCares, is considered a com-
plete laboratory in itself based on its offered services [24].
By using this, any farmer, without having any lab experience,
can analyze up to 100 samples per day (overall, more than
22,000 nutrient samples a year) without visiting any lab.
Drought is a major concern which limits the productivity
of crop yield. Most of the regions around the globe face
this issue with various intensities. To deal with this issue,
especially in very rural areas, remote sensing is being used to
obtain frequent soil moisture data which helps to analyze the
agricultural drought in far regions. For this purpose, the Soil
Moisture and Ocean Salinity (SMOS) satellite was launched
in 2009 which provides global soil moisture maps every, one
to two days. Authors in [25] used SMOS L2 to calculate the
Soil Water Deficit Index (SWDI) in Spain in 2014. In this
effort, they followed different approaches to obtain the soil
water parameters in order to compare with the SWDI acquired
from in situ data. In [26], authors used the moderate resolu-
tion imaging spectroradiometer (MODIS) sensor to map vari-
ous soil functional properties to estimate the land degradation
risk for sub-Saharan Africa. The soil maps and field survey
data, which covered all major climate zones on the continent,
were used to develop the prediction models.
Sensors and vision based technologies are helpful to decide
the distance and depth for sowing the seed efficiently. Like
in [27], sensor and vision based autonomous robot called
Agribot is developed for sowing seeds. The robot can perform
on any agricultural lands on which the self-awareness of
the robot’s placement is ascertained through the global and
local maps generated from Global Positioning System (GPS)
while the on-board vision system is paired with a personal
computer. Advancing further, various non-contact sensing
FIGURE 3. General hierarchy of possible applications, services and
methods are proposed to determine the seed flow rate as
sensors for smart agriculture. in [28] where the sensors are equipped with LEDs; consist
of infrared, visible light and laser-LED as well as an element
weather consents, it may be done in in Fall or Winter [23]. as a radiation receiver. The output voltage varies based on the
The factors that are critical to analyze the soil nutrient lev- movement of the seeds through the sensor and band of light
els include soil type, cropping history, fertilizer application, rays, and falling of shades on the elements of receiver. The
irrigation level, topography, etc. These factors give insight signal information, linked to the passing seeds, is used to
regarding the chemical, physical, and biological statuses of measure the seed flow rate.
a soil to identify the limiting factors such that the crops can
be dealt accordingly. Soil mapping opens the door to sowing B. IRRIGATION
different crop varieties in a specific field to better match soil About 97% of Earth’s water is salt-water held by oceans and
properties accordingly, like seed suitability, time to sow, and seas, and only the remaining 3% is fresh water—more than
even the planting depth, as some are deep-rooted and others two-third of which is frozen in the forms of glaciers and
less. Furthermore, growing multiple crops together could also polar ice caps [29], [30]. Only 0.5% of the unfrozen fresh
lead to smarter use of agriculture, simply making the best use water is above the ground or in the air, as the rest lies under-
of resources. ground [31]. In short, humanity relies on this 0.5% to fulfill
Currently, manufacturers are providing a wide range of all its requirements and to maintain the ecosystem, as enough
toolkits and sensors that can assist farmers to track the soil fresh water must be kept in rivers, lakes, and other similar
quality and, based on this data, recommend remedies to avoid reservoirs to sustain it. It is worth mentioning that solely the
its degradation. These systems allow for the monitoring of agriculture industry uses approximately 70% of this accessi-
soil properties, such as texture, water-holding capacity, and ble fresh water [32], [33]. In many countries, situation rises

129554 VOLUME 7, 2019

M. Ayaz et al.: IoT-Based Smart Agriculture: Toward Making the Fields Talk

to 75% e.g. Brazil, further in some underdeveloped countries, movement [38]. Any sort of nutrients deficiency or apply-
even it exceeds 80% [34]. The main reason for this high water ing them improperly can be seriously harmful for the plant
consumption is the monitoring procedure as even in 2013, health. More importantly, excessive use of fertilizer not only
crops visual inspection for irrigation decision-making was results in financial losses but also creates harmful impacts
very common, as nearly 80% of farms in United States were to the soil and environment by depleting the soil quality,
observed by this [10], [35]. According to the UN Convention poisoning ground water, and contributing to global climate
to Combat Desertification (UNCCD) estimates in 2013 show changes. Overall, crops absorb less than half the nitrogen
that there were 168 countries affected by desertification and applied as fertilizer, while remaining either emitted to the
by 2030, almost half of the world population will be living atmosphere or lost as run off. Unbalanced use of fertilizer
in areas with high water shortages [36]. Considering the fig- leads to an imbalance in both soil nutrient levels and global
ures of water crises around the globe, same time its increasing climate as, reportedly, around 80% of the world’s deforesta-
demands in agriculture and many other industries, it should be tion has occurred due to agricultural practices alone [39].
provided to places only where it is needed, most importantly, Fertilization under smart agriculture helps to precisely
in required quantities. For this purpose, increased aware- estimate the required dose of nutrients, ultimately minimize
ness has been implemented to conserve the existing under- their negative effects on the environment. Fertilization
stress water resources by employing more efficient irrigation requires site-specific soil nutrient level measurements based
systems. on various factors, such as crop type, soil type, soil absorption
Various controlled irrigation methods, like drip irrigation capability, product yield, fertility type and utilization rate,
and sprinkler irrigation, are being promoted to tackle the weather condition, etc. The reason is that the measurement
water wastage issues, which were also found in traditional of soil nutrient level is not only expensive but also time
methods like flood irrigation and furrow irrigation. Both the consuming, as, typically, investigations of soil samples at
crop quality and quantity are badly affected when facing each location are required. To better depict this discussion,
water shortage, as irregular irrigation, even excess, leads to figure 4, summarizes the major inputs, processes and resul-
reduced soil nutrients and provokes different microbial infec- tant outputs of smart agriculture.
tions. It is not a simple task to accurately estimate the water New IoT-based fertilizing approaches help to estimate
demand of crops, where factors like crop type, irrigation the spatial patterns of nutrients requirements with a higher
method, soil type, precipitation, crop needs, and soil moisture accuracy and minimum labor requirements [40], [41].
retention are involved. Considering this fact, a precise soil For example, the Normalized Difference Vegetation Index
and air moisture control system using the wireless sensors not (NDVI) uses aerial/satellite images to monitor crop nutrient
only makes an optimal use of water but also leads to better status [42], [43]. Basically, NDVI is based on the reflection
crop health. of visible and near-infrared light from vegetation and is used
The current situation of irrigation methods is expected to estimate the crop health, vegetation vigor, and density,
to be changed by adopting the emerging IoT technologies. further contributing to assess the soil nutrient level. Such
A significant increase in crop efficiency is expected with precise implementation can significantly improve the fertil-
the use of IoT based techniques, such as crop water stress izer efficiency, simultaneously reducing the side effects to the
index (CWSI)-based irrigation management [10], [35]. For environment. Many recent enabling technologies, like GPS
this, attaining crop canopy at different periods and air tem- accuracy [44], geo mapping [45], Variable Rate Technology
perature are needed for the calculation of CWSI. A wireless (VRT) [46], [47], and autonomous vehicles [48], are strongly
sensors based monitoring system where all the field sensors contributing to IoT-based smart fertilization.
are connected to collect the mentioned measurements, further Other than precision fertilization, fertigation [49] and
transmit to processing center where corresponding intelli- chemigation [50], [51] are other benefits of IoT. In these
gent software applications are used to analyze the farm data. methods, water-soluble matters, such as fertilizers, soil
Not only this but information from other sources including amendments, and pesticides, can be applied through the
weather data and satellite imaging is applied to CWSI models irrigation system. Although, these methods are not new to
for water need assessment, and finally specific irrigation agriculture and have been applied over last three decades,
index value is produced for each site. A prominent example is their precise use with real results has been witnessed only
VRI (Variable Rate Irrigation) optimization by CropMetrics with IoT integration [52], [53]. Based on recent outcomes,
[37], which works according to topography or soil variability, fertigation is considered as the best management practice to
ultimately improves the water use efficiency. improve the effectiveness of many agriculture matters; most
importantly, it can be integrated with IoT-based smart farming
C. FERTILIZER infrastructure seamlessly.
A fertilizer is a natural or chemical substance that can
provide important nutrients for the growth and fertility of D. CROP DISEASE AND PEST MANAGEMENT
plants. Plants mainly need three key macronutrients: nitrogen The Great Famine, also known as the Irish Potato Famine,
(N) for leaf growth; phosphorus (P) for root, flowers, and in which approximately one million Irish people died
fruit development; potassium (K) for stem growth and water around 1950, resulted due to crop failure and yield

VOLUME 7, 2019 129555

 M. Ayaz et al.: IoT-Based Smart Agriculture: Toward Making the Fields Talk

Generally, the reliability of crop disease monitoring and

pest management depends on three aspects: sensing, evaluat-
ing, and treatment. The advanced disease and pest recognition
approaches are based on image processing in which raw
images are acquired throughout the crop area using field
sensors, UAVs, or remote sensing satellites. Usually, remote
sensing imagery covers large areas and, hence, offers higher
efficiency with lower cost. On the other hand, field sensors
are capable to support more functions in collecting data,
like environment sampling, plant health, and pest situations,
in every corner throughout the crop cycle. For example,
IoT-based automated traps [62], [63] can capture, count, and
even characterize insect types, further uploading data to the
Cloud for detailed analysis, which is not possible through
remote sensing.
Approaches like vehicle precise spray and automatic VRT
chemigation [64], commonly used under smart fertilization,
can also be utilized for disease treatment and other pes-
ticide applications. Moreover, the advancement of robotic
technology offers new solutions. When equipping an agricul-
tural robot with multispectral sensing devices and precision
spraying nozzles, it can locate and deal with pest problems
more precisely under the manipulation of a remote IoT dis-
ease management system. This IoT-based pest management
system has many advantages, as it can reduce the overall
expenditures while, at the same time, support the restoration
of the natural climate. For example, recently, it has been found
that yields of many crop types are facing severe threat due to
the lack of pollination [65], [66]. In fact, the pollination is
being affected due to bee colony collapse disorder resulting
from the uncontrolled pesticides.
FIGURE 4. Some key inputs, processes involved and possible outputs of
smart farming.
E. YIELD MONITORING, FORECASTING, AND HARVESTING
Yield monitoring is the mechanism used to analyze various
reduction caused by ‘‘potato blight’’ disease [54]. Even today, aspects corresponding to agricultural yield, like grain mass
corn growers in the US and southern Canada are facing flow, moisture content, and harvested grain quantity. It helps
an economic loss of approximately one billion USD due to accurately assess by recording the crop yield and moisture
to ‘‘southern corn leaf blight’’ disease [55]. The Food and level to estimate, how well the crop performed and what to do
Agriculture Organization (FAO) estimates that 20–40% of next. Yield monitoring is considered an essential part of pre-
global crop yields are lost annually due to pests and dis- cision farming not only at the time of harvest but even before
eases [56]. To control such vast production losses, pesticides that, as monitoring the yield quality plays a crucial role. Yield
and other agrochemicals became an important component of quality depends on many factors, e.g. sufficient pollination
the agriculture industry during the last century. It is estimated with good quality pollen especially when predicting seed
that, in each year, around half a million tons of pesticide are yields under changing environmental conditions [67]–[69].
used in the US alone, while more than two-million tons are Currently, when we are dealing with more open markets,
used globally [57]. Most of these pesticides are harmful to buyers around the world become more particular about fruit
human and animal health, leaving severe, even irreversible, quality; hence, effective production depends on the right fruit
impact to the environment, ultimately causing significant size to the right market at the right time [13].
contamination to entire ecosystems [58], [59]. Crop forecasting is an art to predict the yield and produc-
Recent IoT based intelligent devices, such as wireless tion (tons/ha) before the harvest takes place. This forecasting
sensors, robots and drones are allowing the growers to slash helps the farmer for near-future planning and decision mak-
pesticide uses significantly by precisely spotting crop ene- ing. Furthermore, analyzing the yield quality and its maturity
mies. Compared to traditional calendar or prescription based is another critical factor which enables the determination of
pest control procedures, modern IoT-based pest management the right time for harvesting. This monitoring covers various
provides real-time monitoring, modeling, disease forecasting, development stages and uses fruit conditions like its color,
hence proving more effective [60], [61]. size, etc., for this purpose. Predicting the right harvesting time

129556 VOLUME 7, 2019

M. Ayaz et al.: IoT-Based Smart Agriculture: Toward Making the Fields Talk

not only helps to maximize the crop quality and production and pesticides. Soon it was realized that these conven-
but also provides an opportunity to adjust the management tional ways were not adequate enough to fit this demand
strategy. Although, harvesting is the last stage of this process, gap; hence, agriculture scientists have begun thinking of
proper scheduling can make a clear difference. To obtain the other alternatives, like bioengineered (BE) foods. BE foods,
real benefits from crops, farmers need to know when these also known as genetically modified (GM) or genetically
crops are actually ready to harvest. Figure 5 represents a engineered (GE) foods, are foods produced by introduc-
snapshot of a farm area network (FAN) that can portrait the ing changes into their DNA using the methods of genetic
whole farm to the farmer in real time. engineering. However, several studies highlight their serious
effects on human health, including infertility, disruption in
immune system, accelerated aging, faulty insulin regulations,
etc. [75], [76]. All these and many other similar technologies
did not receive much popularity and acceptance in society
because people prefer bio and organic food. In this regards,
massive research has been conducted for decades in which
sensors and IoT-based technologies are helping to improve
conventional agriculture processes to enhance yield produc-
tion without, or with minimum, effect on its originality.
For this purpose, new sophisticated and more controlled
environments are projected to tackle the above-mentioned
issues. The importance and involvement of new technologies
is more critical, as we are moving toward more cultured and
urban farming. In fact, it would not be incorrect for one to
say that the success of these advanced practices is in doubt
without using sensor-based technologies.

FIGURE 5. An IoT based farm area network (FAN).

A. GREENHOUSE FARMING
A yield monitor, developed in [70], can be installed on any Greenhouse farming is considered the oldest method of smart
harvester combine and linked with the mobile app FarmRTX, farming. Although, the idea of growing plants in controlled
which displays live harvest data and uploads it automatically environment is not new as found since Roman times but it
to the manufacturer’s web-based platform. This app has the gained popularity in 19th century where largest greenhouses
ability to generate high-quality yield maps and share these were built in France, Netherlands and Italy. Further, the prac-
maps with an agronomist, and the farmer also has the option tice was accelerated in the mid-20th century and highly pro-
to export to other farm management software to analyze them. moted in countries that facing harsh weather conditions [77].
To estimate the production and quality of yield precisely, Crops grown indoors are very less affected by environment;
the measurement of fruit growth can be highly beneficial. most importantly, they are not limited to receiving light only
This idea is used in [71], where authors considered the fruit during the daytime. As a result, the crops that traditionally
growth as the most basic and relevant parameter to estimate could only be grown under suitable conditions or in cer-
that how well the crop is progressing. Satellite images can tain parts of the world are now being growing anytime and
be a good option to monitor the yield of crops with vast anywhere. This was the actual time in which sensors and
areas. This method is utilized in [72], where authors used communication devices started to support various agriculture
Sentinel-1A Interferometric images to map the rice crop yield applications genuinely.
and intensity in Myanmar. As we mentioned earlier in this The success and production of various crops under such
section, fruit size always plays a critical role to estimate controlled environment depend on many factors, like accu-
its maturation, making decisions regarding harvesting, and racy of monitoring parameters, structure of shed, covering
targeting the right market, for this purpose, color (RGB) depth material to control wind effects, ventilation system, decision
images are used in [73] to track the different fruit conditions support system, etc. A detailed analysis is provided in [78],
in mango farms. Similarly, multiple optical sensors are used where all these factors, their impacts, and how wireless
in [74] to monitor the shrinking of papayas, especially during sensors can help for all this are considered. Precise mon-
drying conditions. itoring of environment parameters is the most critical task
in modern greenhouses, where several measurement points
III. ADVANCED AGRICULTURAL PRACTICES of various parameters are required to control and ensure the
Adopting the novel methods to enhance the quality and local climate. In [79], an IoT-based prototype is proposed
quantity of food is not something new, as humans have to monitor the greenhouses where MicaZ nodes are used to
been doing this for centuries. Initially, we tried to enhance measure the inside parameters like humidity, temperature,
the crop production by focusing on seed variety, fertilizers, light, and pressure.

VOLUME 7, 2019 129557

 M. Ayaz et al.: IoT-Based Smart Agriculture: Toward Making the Fields Talk

B. VERTICAL FARMING When combining hydroponics with VF, a farm of 100 sq.
The world needs more farmable lands to fulfill increased meters can produce the crop equivalent to 1 acre of traditional
food demands, but reality is that one-third arable land was farm, most importantly upto 95% less water and fertilizers
lost during the last four decades due to erosion and pollution utilization and without pesticides/herbicides [87]. Currently,
[80], [81]. Unfortunately, current agricultural practices based available systems and sensors e.g. [88], [89] are not only
on industrial farming are damaging the soil quality far faster used to monitor a range of parameters and take readings at
than nature can rebuild it. Overall, it is estimated that erosion predefined intervals but, also, the measurements are stored so
rates from cultivated fields is 10 to 40 times greater than the that can be used to analyze and diagnostic purpose later on.
soil formation rates [82]. Considering the reduction of arable Under this application, the precision of nutrient measure-
land issues, it could be a disaster for food production in the ments is crucial, as such, a highly reliable wireless control
near future with current agriculture practices. Further, as we system for tomato hydroponics is proposed in [90] in which
mentioned, 70% of fresh water is only used for agriculture they focused on various communication standards that are
purpose, which can increase the burden on existing limited least effected by plants’ presence and their growth. The
water reservoirs. Vertical Farming (VF) is an answer to meet monitoring of solution contents and their precision is most
the challenges of land and water shortages. critical under this method; for this purpose, many systems
VF in the form of urban agriculture offers an opportunity are offered to check the presence of contents considering the
to stack the plants in a more controlled environment result- plant demands. In [91] a wireless-sensor-based prototype is
ing in, most importantly, significant reduction in resource proposed to deliver a turn-key solution for the hydroponic
consumption. By following this method, we can increase the cultivation which offers real-time measurements for soilless
production multiple times, as only a fraction of ground sur- indoor growing. Further, a compact sensor module is pre-
face is required (depending on the number of stacks) as com- sented in [92], which uses oscillator circuits to measure the
pared to traditional agriculture practices. Not only for ground presence and concentrations of various nutrients and water
surface, this system is highly efficient in terms of other levels.
resources, as well. For example, according to Mirai, a Japan
based indoor farm developer presented the figures regarding D. PHENOTYPING
a Japanese farm comprised of 25,000 square meters. The fig- The previously discussed smart methods look more promis-
ures are highly encouraging, as it is producing 10,000 heads ing for the future of agriculture, as they are already being
of lettuce per day (double the production when compared used to produce different crop products under precise envi-
with traditional methods) and is, most importantly, consum- ronments. Other than these, a few advanced techniques are
ing 40% less energy and up to 99% reduced water consump- under experiment to further enhance the crop capabilities by
tion compared to outdoor fields [83]. Aerofarms, a leader controlling their limitations with the help of advanced sensing
in VF, growing agricultural products with upto 390 times and communication technologies. Among these methods, the
higher yields while utilizing 95% less water at Newark [84]. more prominent is phenotyping, which is based on emerging
Under this farming method, many parameters are impor- crop engineering, which links plant genomics with its eco-
tant, but CO2 measurements are most critical; hence, physiology and agronomy, as shows in Figure 6. The progress
non-dispersive infrared (NDIR) CO2 sensors play a crit- in molecular and genetic tools for various crop breeding was
ical role to track and control the conditions in vertical significant in the last decade. However, a quantitative analysis
farms. Boxed Gascard, developed by Edinburgh Sensors [85], of the crop behavior, e.g. grain weight, pathogen resistance,
is especially designed by considering such an environment, etc., was limited due to the lack of efficient techniques and
which employs a pseudo dual beam NDIR measurement sys- technologies that we can now enjoy.
tem to enhance the stability and reduced optical complexity.
Human hands are not required to touch the crops at any stage
when following the IoT-connected vertical farm; this is the
claim made by Mint Controls [86] developers who offer a
wide range of solutions, like waste containers and sensors
and their integration for various VF applications.

C. HYDROPONIC
In order to enhance the benefits of greenhouse farming, agri-
culture experts moved forward another step and provided
the idea of hydroponic, a subset of hydroculture in which
plants are grown without soil. Hydroponic is based on an FIGURE 6. The process of phenotyping [96].
irrigation system in which balanced nutrients are dissolved
in water and crop roots stay in that solution; in some cases, Research investigations, completed in [93], conclude that
roots can be supported by medium like perlite or gravel. plant phenotyping can be highly beneficial to investigate

129558 VOLUME 7, 2019

M. Ayaz et al.: IoT-Based Smart Agriculture: Toward Making the Fields Talk

the quantitative characteristics, such as those are responsible Based on these facts, during the period of 2017 to 2022, the
for its growth, yield quality and quantity, and resistance global smart farming market is predicted to rise at a growth
capabilities to handle various stresses. Similarly, the role rate of 19.3% per year to touch $23.14 billion in 2022 [97].
of sensing technologies and image-based phenotyping are Here it is worth mentioning that UAV/drones are generating
highlighted in [94] and describes how these solutions can and further expected to generate the highest revenue amongst
help to boost the progress not only for screening numerous all agricultural robots utilized in smart farming (UAVs are
biostimulants but also their role in understanding the mode discussed in Section V). Evergreen demand for higher crop
of actions. Furthermore, an IoT-based phenotyping platform, yield, increased incorporation of information and commu-
CropQuant, is designed to monitor the crop and relevant trait nication technology (ICT) in farming and the rapid global
measurements that can provide facility for crop breeding and climatic changes are some of the major drivers resulting to
digital agriculture [95]. Here, an automatic in-field control such high market growth.
system was developed to process the data generated by plat- Manufacturers in the market offer a variety of products and
form. The provided trait analyses algorithms and machine- solutions, mostly based on sensors and efficient communica-
learning modeling help to explore the relation among the tion for a range of applications; a few are shown in figure 7.
genotypes, phenotypes, and environment where it grows. The key technologies and equipment’s that are currently
available for this purpose are discussed in following.

IV. MAJOR EQUIPMENT AND TECHNOLOGIES

Different from ancient farming, most of the tasks in modern,
large-scale agriculture are being done by heavy and urbane
equipment, such as tractors, harvesters, and other robots
which are fully or partially supported by remote sensing and
other communication technologies. In precision agriculture,
when tasks like sowing, fertilizing, irrigation, and harvesting
are being performed, the operating vehicles are equipped
with GPS and GIS facilities so that they can work precisely,
site-specifically, and autonomously. In fact, the idea of site-
specific crop management is not possible without involving
the recent advanced technologies. The success of precision
agriculture is based on the accuracy of collected data, which
is usually done in two ways [10]. The first entails the usage
of multifunctional imaginary devices equipped with remote
sensing platforms, such as satellites, agriculture airplanes,
balloons, and UAVs; the second is from various types of
sensors—those that are mostly deployed for specific purpose
across various sites of our interest. The gathered data is
identified with the precise location information by using GPS
devices so that the site-specific treatment can be provided FIGURE 7. Selected IoT based products and prototypes for smart
agriculture.
afterwards.
Agriculture has transformed during the last few decades
from small/medium farming operations to highly industri- A. WIRELESS SENSORS
alized and commercial farming. This transition allows the Among all the equipment for smart farming currently avail-
leading corporations to treat agriculture like other industries, able in the market, wireless sensors are the most crucial
e.g., manufacturing where the measurements, data, and con- and play a key role when it comes to collecting the crop
trol are very important to provide a balance between costs conditions and other information. Wireless sensors are being
and production in order to boost the profits. Accordingly, used standalone wherever required, further integrated with
every aspect of agriculture that can be automated, digitally almost every portion of advanced agricultural tools and heavy
planned, and managed will benefit from IoT technologies machinery, depending on application requirements. In the
and solutions. Based on this fact, efforts are being focused following, major sensor types are discussed according to their
to offer more sophisticated tools such as agricultural robots working procedure and purpose and the benefits they offer.
to perform a range of activities, like planting, watering,
weeding, picking, thinning, fertilizing, spraying, packing and 1) ACOUSTIC SENSORS
transporting. This revolution is being driven not only due to Acoustic sensors offer a miscellaneous appliance in farm
the advancement of technology, but is also a result of factors management, including soil cultivation, weeding, fruit har-
like fear of losing the low-cost labor, most importantly need vesting, etc.; the main benefit of this technology is its low-
for better and cheaper food. cost solutions with fast response, especially when considering

VOLUME 7, 2019 129559

 M. Ayaz et al.: IoT-Based Smart Agriculture: Toward Making the Fields Talk

portable equipment. It functions by measuring the change in 6) AIRFLOW SENSORS

the noise level as the tool interacts with other materials, e.g., These sensors are capable of measuring soil air permeability
soil particles [98]. Acoustic sensors are commonly used for and percentage of moisture and identifying soil structure to
pest monitoring and detection [99] and classifying the seed distinguish different types of soils. Measurements can be
varieties according to their sound absorption spectra [100]. made at singular locations or dynamically while in motion,
e.g., can be used on a fixed position or in mobile mode. The
2) FIELD-PROGRAMMABLE GATE ARRAY (FPGA)-BASED desired output is the pressure required to push a predeter-
SENSORS mined amount of air into the ground at a prescribed depth.
FPGA based sensors are starting to be used in agricul- It follows the procedure of various soil properties, including
ture recently due to their flexibility of reconfiguration. compaction, structure, and moisture levels, producing unique
The major options where these can be employed include identifying signatures [113].
measuring real-time plant transpiration, irrigation, and
humidity [101], [102]. However, their utilization in agricul-
7) ELECTROCHEMICAL SENSORS
ture is in the early stages due to their limitations, such as
These are mostly used to assess the significant soil character-
size, cost, and power consumption. These sensors require
istics to analyze the soil nutrient levels, such as pH [114]. The
more power; hence, they are not suitable for continuous
standard chemical soil analysis, which are mostly expensive
monitoring, further high in cost even compromising on per-
and time consuming, can be easily substituted with these
formance [103]. By overcoming these issues, FPGA-based
sensors. To be more precise, the macro and micro nutrients
sensors can offer satisfactory solutions according to specific
in the soil, salinity and pH [115], are measured using sensors
application requirements.
of this nature.
3) OPTICAL SENSORS
These sensors use light reflectance phenomena and help to 8) ELECTROMAGNETIC SENSORS
measure soil organic substances, soil moisture and color, Electromagnetic sensors are used to record electrical conduc-
presence of minerals and their composition, clay content, etc. tivity and transient electromagnetic response, identify elec-
[104], [105]. These sensors test the soil’s ability to reflect trical response, and adjust variable rate applications in the
light based on different parts of the electromagnetic spectrum. actual situation. Sensors based on this technology use electric
The changes occurred in wave reflections help to indicate the circuits to measure the capability of soil particles to con-
changes in soil density and other parameters. Fluorescence- duct or accumulate electrical charge, which is mostly done
based optical sensors are used for basic plant assessment, by the following two methods: contact or non-contact [116].
especially to supervise the fruit maturation [106]. Further, Residual nitrates and organic matter in the soil can also
when integrating optical sensors with microwave scattering, be measured using the electromagnetic sensors, as done
it can be used for characterizing grove canopies, like olives in [117].
and other similar crops [107].
9) MECHANICAL SENSORS
4) ULTRASONIC RANGING SENSORS
Mechanical sensors assess the soil mechanical resistance
Sensors of this category are considered a good choice being
(compaction) to indicate the variable level of compaction.
low cost, potential to operate in a variety of applications,
The mechanical sensors enter or cut through the soil and
and ease of use and adjustability, such as the sampling rate.
record the force assessed by strain gauges or load cells [118].
Common uses are tank monitoring, spray distance measure-
A pressure unit is used to measure the soil’s mechanical
ment (e.g., boom height and width control in order to per-
resistance, which is actually the ratio of the force needed to go
form uniform spray coverage, object detection, and collision
into the soil medium using the frontal part of the tool, actually
avoidance), and monitoring crop canopy [108], [109]. When
engaged with the soil.
combined with a camera, these sensors can then be used
for the weed detection [110], where the heights of plants
are identified using the ultrasonic sensors and the camera 10) MASS FLOW SENSORS
determines the weed and crop coverage. Category of this sensors are used for yield monitoring, as it
provides the yield information by measuring the amount of
5) OPTOELECTRONIC SENSORS grain flow, e.g., when passing through the combine harvester.
Optoelectronic sensors can differentiate plant type; hence, Sensing the mass flow of grain to determine the crop yield is
they help to detect weeds, herbicides, and other unwanted not new, as it has been performed the last two decades [119].
plants, especially in wide-row crops [111]. When combining, The mass flow sensor is the most critical component, but,
an optoelectronic sensor and location information, it can map overall, the yield monitoring system consists of several other
the weed distribution and resolution [112]. Optoelectronic modules, like the grain moisture sensor, data storage device,
sensors are also capable of differentiating between vegetation and an internal software to analyze the data,, which are within
and soil from their reflection spectra. the interface provided in the John Deere tractors [120].

129560 VOLUME 7, 2019

M. Ayaz et al.: IoT-Based Smart Agriculture: Toward Making the Fields Talk

11) EDDY COVARIANCE-BASED SENSORS and mobile platforms worldwide [135]. Moreover, automatic
This type of sensors can be used for quantifying exchanges packet reporting system (APRS) is being integrated to report
of carbon dioxide, water vapor, methane or other gases, and telemetry data through satellite communication [136].
energy between the surface of the earth and the atmosphere. Table 1 lists a few sensors to provide the idea about their
This method offers an accurate way to measure surface- possible uses and the environment where they can be placed.
atmosphere fluxes of energy and trace gas fluxes over a
variety of ecosystems for, most importantly, agricultural TABLE 1. Some selected sensors and their possible uses in IoT based
agriculture.
applications [121]. Currently, the sensors based on this tech-
nology are preferred over other similar options, like the close
chamber, due to high precision and its ability to measuring
continuous flux over large areas [122].

12) SOFT WATER LEVEL-BASED (SWLB) SENSORS

SWLB sensors are being utilized in agriculture catchments
to characterize hydrological behaviors, such as water level
and flow, at adjustable time-step acquisitions. This is done by
measuring rainfalls, stream flows, and other water presence
options [11], [123].

13) LIGHT DETECTION AND RANGING (LIDAR)

This technology is widely used in a range of agriculture
applications, such as land mapping and segmentation, deter-
mining soil type, farm 3D modelling, monitoring erosion and
soil loss, and yield forecasting [124]–[126]. LiDAR is also
commonly used to obtain dynamic measurement information
regarding fruit-tree leaf area, and, when combined with GPS,
it can produce a 3D map [127]. Moreover, this technology is
often used when estimating the biomass of various crops and
trees [128].

14) TELEMATICS SENSORS

Telematics sensors support telecommunication between two
places—more precisely, among two vehicles when consid-
ering the agriculture-based applications. Telemetry sensors
are used to collect data from remote locations (especially
inaccessible points), operations of machines that report
on how the components are working, and record location
and travel routes to avoid visiting the same patch [129].
These services enable farm managers to record and store all
information related to farm operations automatically, which
maximizes the utilization of environmental benefits, further
can minimize threats like farm equipment theft as utilized
in [130], [131].

15) REMOTE SENSING

Sensors belong to this category are used to capture and B. IOT BASED TRACTORS
store the geographic information, further analyze, manipu- As rural labor resources have started to come under stress
late, manage and present all types of spatial or geographical due to the expansion of the crop industry, tractors and
data. Similar to LiDAR, these sensors also found significant other automatic heavy machinery started to enter the agri-
use in agriculture applications including crop assessment, culture sector. Where available, an average size tractor can
forecasting yield dates, yield modelling and forecasting, work 40 times faster with significantly less expenses than
identification of plants and pests, land cover and degradation traditional farm labor [157]. To fulfill the continuously
mapping etc [132]–[134]. Argos sensor is one of leading increasing demands, agricultural-based equipment manu-
example, a satellite-based sensor system used to collect, facturers, like John Deere, Hello Tractors, Case IH and
process and disseminates environmental data from fixed CNH (New Holland), have started to provide better solutions

VOLUME 7, 2019 129561

 M. Ayaz et al.: IoT-Based Smart Agriculture: Toward Making the Fields Talk

focusing on the grower’s requirements. With the advancement ultimately, its success. In some crops, this is done a single
of technology, most of these manufacturers are offering trac- time while, in some others, performed several times, even
tors with automatic-driven and even Cloud-computing capa- on a daily basis, as crop reaches a certain stage. Harvesting
bilities. This technology is not new, as self-driving tractors the crop at the right time is very critical, as doing so either
have been in the market even before semi-autonomous cars. early or late can affect the production significantly. When
One of the main advantages of self-driving tractors is their talking about the labor, it is estimated that the US faces a
ability to avoid revisiting the same area or row by reducing $3.1 billion decline in crop production on a yearly basis due to
the overlap even less than an inch. In addition, they can labor shortage [164]. Not only this, but, according to a study
make very precise turns without a driver’s physical presence. conducted by the United States Department of Agriculture,
This facility offers better precision with reduced errors, espe- overall 14 % of farm costs go to wages and labor costs,
cially when spraying insecticide or targeting weeds; those are while it can be upto 39% in some labor intensive farms [165].
mostly unavoidable when a human controls the machinery. Considering the worth of this stage and labor issues, farm
Although, at the moment, no fully autonomous tractor experts expect that involvement of agriculture robotics may
is available in market, many researchers and manufactures not only ease the labor pressure but also provide the flexibility
are hardly working to mature the technology. Based on to harvest whenever needed.
current progress and future demands of high-tech tractors, In order to automate the harvesting process and make
it is predicated that around 700,000 tractors equipped with it more precise, the role of robots has been increasing
facilities like autosteer or tractor guidance will be sold over the recent decades. Considering the robot services,
in 2028 [158], while the same study expects that around many researchers have done intensive research in order to
40,000 unmanned, fully-autonomous (level 5) tractors will be mature the sensitivity of fruit detection, its shape, size, color,
sold in 2038 [159]. and localization [166]–[169]. Automatic harvesting of fruits
When talking about such cultured machines, most farmers requires deep investigation of sophisticated sensors that are
can’t afford to own them while most of the tractor service capable of collecting precise and unambiguous information
providers and manufacturers operate well below their poten- of that particular crop and fruit. The task of detecting the
tial. Considering the challenge, Hello Tractor has developed right target in natural scenes is not simple since most of
a solution to sort these issues. The company has developed a the fruits are occluded partially—sometimes even fully—
low-cost monitoring device that can be placed on any tractor, under the leaves and branches or are overlapped with other
provides powerful software and analytics tools [160]. The fruits [170]. Here, most of the prominent studies found in
benefits of this device are twofold- on one side it ensures regard to this purpose are deeply based on computer vision,
that overall cost of tractor remains affordable for the most image processing, and machine learning techniques. This
of growers while at the same time it monitors the condition process needs very specialized and sophisticated tools to
of the tractor and reports if any problems occur. The soft- differentiate the fruit conditions, as there are more than sixty
ware connects tractor’s owner to farmers in need of tractor shapes, sizes, and colors for a pepper alone when it is ready to
services, just like Uber for tractors. Another major example harvest. Considering such complexity, many robots are being
is Case IH’s Magnum series [161] tractor which uses on developed for specific crops. Some of the leading robots
board video cameras and LiDAR sensors for object detection being used for crop harvesting include SW 6010 [171] and
and collision avoidance. Recently, Case IH used this tractor Octinion [172] for strawberries, SWEEPER robot [173] for
to plant soybeans by following the concept of autonomous peppers, and FFRobot [174] for tree-based fruits like apples
tractors. In another development made by standards group which can pick up to 10,000 fruits per hour.
ETSI where world’s first tractor connected to a car in France, Strawberries are one of the most consumed fruits, avail-
using IoT [162] to control the accidents due to farm vehicles. able mostly throughout the year while labor is the major
After collecting all the important crop data, the next contributor to the high cost of this fruit, especially during
step is pushing computing from the Cloud to the edge, harvesting and packaging stages [175]. As the strawberry
as John Deere [163] wants. In their proposed system, an ana- farms are grown mostly under greenhouse systems hence
lytics engine works locally on the farmer’s tractor rather than the harvesting robots are designed to move on defined paths
in the Cloud in order to adjust the local inputs. For this like rails where the translational motion is restricted and
purpose, they considered all the existing analytics and recom- robots can move backward and forward only. Robots devel-
mendations to modify the current data in real time depending oped by Agrobot are able to collect strawberries along the
on the field conditions. Based on this phenomenon, the man- side of strawberry plant rows in the field, further packed by
ufacturer is bringing their tractors to next level by connecting human operators [176]. For example, SW6010 by Agrobot is
their machine to the Internet and creating a method to display a specialized and semi-automatic robot towards the specific
the information wherever farmer wants to see it. task of strawberry harvesting [177]. Tektu T-100 is an all-
electric rechargeable strawberry harvester run silently with
C. HARVESTING ROBOTS zero emission inside the poly-tunnels [178]. The installed
Harvesting is the most critical stage during the production pickers are able to position over the crop rows and gather the
process, as this last phase dictates the crop’s output and, fruit quickly and efficiently, directly into punnets.

129562 VOLUME 7, 2019

M. Ayaz et al.: IoT-Based Smart Agriculture: Toward Making the Fields Talk

D. COMMUNICATION IN AGRICULTURE technology has its own worth but the FMS also plays a critical
Communication and reporting the information on a timely role which must be custom designed considering the specific
basis are considered the backbone of precision agriculture. application requirements. Generally, based on communica-
The real purpose cannot be achieved unless a firm, reliable, tion data rates and power consumption, wireless sensors for
and secure connection among various participating objects agriculture applications are divided in three broad categories
is provided. To achieve communication reliability, telecom as shown in table 2.
operators can play a crucial role in the agricultural sector.
If we truly want to implement IoT on a large scale in the TABLE 2. Data and power specifications of wireless sensors commonly
used for agriculture applications.
agriculture industry, we have to provide a suitably large
architecture. Here, the factors like cost, coverage, energy
consumption, and reliability are critical and have to be
considered before choosing the mean of communication.
Low-energy networks can provide connectivity only on one
site and mostly do not offer services in remote areas where
sensed data need to be transmitted to the farm management
system (FMS). Depending on availability, scalability and
application requirements, various communication modes and
technologies are being used for this purpose, most common
are discussed here,

1) CELLULAR COMMUNICATION
Cellular communication modes from 2G to 4G can be suit-
able, depending on the purpose and bandwidth requirement;
however, the reliability, and even availability, of a cellular
network in rural areas is a major concern. To tackle this,
data transmission via satellite is another option, but, here,
the cost of this communication mode is very high, which 2) ZIGBEE
makes it not suitable for small- and medium-sized farms. The Zigbee is primarily designed for a wide range of applications
choice of communication mode also depends on application especially to replace existing non-standard technologies.
requirements, such as some farms required sensors that can Depending on the application requirements, the devices based
operate with low data rate but need to work for long periods on this protocol can be one of three types including Coordi-
hence demand long battery life. For such scenarios, a new nator, Router and End User. Further, three different topolo-
range of Low Power Wide Area Network (LPWAN) is con- gies are supported by Zigbee networks named, Start, Cluster
sidered a better solution for cellular connectivity, not only Tree and Mesh [34]. Based on these characteristics, and
in terms of long battery life but also a larger connectivity further considering the agriculture application requirements,
range with affordable rates (2 to 15 USD per year) [179]. Zigbee can play vital role especially targeting the greenhouse
Currently, crop and pasture management are two of the main environment where usually short range communications are
applications where LPWAN networks are highly suitable, required. During monitoring the various parameters, the real
and, further considering its success, it can be utilized in many time data from the sensor node is transferred through Zigbee
other farming-related uses. to end server. For the applications like, irrigation and fertiliza-
Besides WAN connectivity option, many short range and tion, Zigbee modules are networked for communication, e.g.
medium level communications are being used in mesh net- in drip irrigation used to monitor soil contents like moisture.
works [180]. For example, a mesh-network of sensor nodes Further, SMS is forwarded to the farmer to update about the
collects data and transmits it to the gateway which is located field data where GSM is required at long distance or Blue-
somewhere in the same area. The gateway further sends this tooth module can help at the shorter distances.
data to the farm management system using the WAN net-
work. The communication technologies used within the mesh 3) BLUETOOTH
networks vary e.g. Bluetooth and Zigbee can be used to pro- Bluetooth is a wireless communication standard that connects
vide connectivity for peer-to-peer wireless communications. small-head devices together over shorter distances usually
From here, the sensed data forwarded to the FMS, which cooperating in a close proximity. Due to its advantages of low
gathers and analyses the information about all the activities power requirements, easy to use and low cost, this technology
happening at different parts even the historical data regarding is being utilized in many smart farming applications. Further,
the weather and climate updates, economic, products being Bluetooth making advancements in many IoT systems with
used and their specifications etc, in short making it decision the release of Bluetooth Low Energy (BLE) or commonly
farming. It is important to mention that, the communication known as Bluetooth Smart. The study conducted in [181]

VOLUME 7, 2019 129563

 M. Ayaz et al.: IoT-Based Smart Agriculture: Toward Making the Fields Talk

which tests Bluetooth and PLC (programmable logic con- need to transmit data at the same time, as experiments done
troller) with ICS (integrated control strategy), timer control in [188].
and soil moisture control approach for smart irrigation. The Figure 8 gives an idea that how end-to-end communica-
target of this study is to find an optimum utilization of tion possibilities can be divided in various layers to interact
water and energy consumption for various greenhouse or field with each other in order to provide the services for smart
applications. A moisture and temperature sensor based on agriculture.
BLE is developed in [182] especially focusing on the agri-
culture environments and weather conditions of crop fields.
Here the reason of choosing BLE for communication purpose
is due to its inherent support for smart phone accessibility.
Further, a similar effort is done in [183] where a new sensor
node is designed to monitor ambient light and temperature
employing BLE communication protocol preferable for IoT
based agriculture applications. Other than short range, WiFi
is utilized whenever LAN communications are required in
smart agriculture. Along short range connectivity, WiFi is uti-
lized whenever LAN communications are required in smart
agriculture. Study presented in [184] investigates a remote
monitoring system using WiFi, where the sensor nodes were
based on WSN802G modules. The deployed nodes commu-
nicate wirelessly with a central server, which is responsible
to collect and store the monitored data and further allow
displaying the information after required analysis.

FIGURE 8. End to end communication for smart farming.

4) LORA
LoRa wireless technology is a long-range, low-power plat-
form used extensively in IoT applications. Being low in
E. SMARTPHONES
power consumption, it offers LPWAN connectivity between
Despite its availability concerns for remote fields, cellular
the wireless sensors and the Cloud. It has proved itself much
communication is the major technology in rural areas; mobile
more effective and reliable than Bluetooth, Wi-Fi, etc. espe-
phones are a very common source and primary mode of
cially in restaurants or kitchen environments. Sensors based
communication whenever the need arises to contact or update
on LoRa can be installed in smaller devices for reliable mon-
most of the farming community. Recent advancements in the
itoring. Most importantly, LoRa signals can penetrate thick
smartphone industry have resulted in sharp price decreases,
and insulated objects, even buildings, and can, hence, cover a
making this industry more attractive, especially to the small-
larger network area.
holder growers in remote areas. The rapid spread of cellular
Overall, LoRa-based networks perform higher in terms of
networks in developing countries offers the opportunity to
lifespan and, at the same time, pose reduced maintenance
reach remote and dispersed farms with improved services.
and upkeep burden [185]. Considering the advantages, many
Mobile-phone-based agriculture services (m-services) are
researchers tested this communication method in kitchens,
far from their assumed potential; according to an analysis
storage rooms, and transport systems. In [186], a test was
done by GSM Association, only at 8 per cent [189].
conducted in a warehouse with a capacity to store 40 tons of
However, the flexibility and functionality, such as the camera,
apples, and results show that it provided full coverage, where
GPS, microphone accelerometer, proximity, and gyroscope,
the temperature and airflow readings transferred successfully
attract the IT experts, and they are developing more
with a packet rate of more the 96%. Similarly, [187] presents
appealing mobile apps that consider the farmer’s various
a system to achieve information traceability in the grain
needs [190], [191]. The conclusion of FAO’s 10-year inves-
transport system to ensure the food quality by monitoring the
tigations states that ‘‘solid information is needed regard-
temperature and humidity levels.
ing the impact of previous initiatives, including lessons
learned, in order to inform the design and approach of future
5) SIGFOX efforts’’ [192]. Interestingly, the smartphone is the first device
Sigfox is also used to provide network connectivity services that comes to mind when planning how to achieve this
to low-powered objects as ‘‘things’’ required. It is based on goal, especially considering the often dispersed and poorly
narrowband or ultra-narrowband technology; hence, it takes serviced areas.
very narrow chunks of spectrum and changes the phase of the Recent years have seen rapid expansion of research,
carrier radio wave to encode the data. Based on these charac- with researchers conducting a growing number of studies
teristics, it offers high level performance, even if 100 sensors and developing various models to highlight the scope of

129564 VOLUME 7, 2019

M. Ayaz et al.: IoT-Based Smart Agriculture: Toward Making the Fields Talk

smartphone crop applications. These researchers mostly TABLE 3. Smartphone based sensors that being used in various
agriculture applications.
belong to developing nations, as proposed systems are
based primarily in countries like Kenya [193], [194],
Ghana [195], [196], Nigeria [197], [198], Mali [199],
Uganda [200], and Zimbabwe [201], [202]. Although,
the scope of smartphone utilization in agriculture has
been more commonly observed in Africa, experiments
in countries like Cameroon [203], China [204], [205],
Turkey [206], [207], and India [208], [209] are also increas-
ing. Analysing the success of m-services depends on many
factors. One of the most comprehensive studies regarding
the use of mobile phones for various agriculture applications
was conducted to review all the important factors [210]. This
study concludes that the service will be of no effect if the
developer of the application does not truly understand the
farmer needs.
Obviously, the most important factor for such applications
is that the farmer should access and use them. In other words,
an easy-to-use, free or low-cost app that supports various
languages could attract the farmer’s attention. In addition,
developers should study and consider the relevant factors
before making their suggestions. For example, market prices
are of great interest to farmers, but would be of no use in
cases of bad roads and unavailability of proper transport vehi-
cles. The developers should target the problems of the wider
community instead of focusing only on farmers, considering
the transporters, brokers, and other agriculture experts as
well. Unfortunately, most of the applications are developed
on growers’ perceptions, instead of using independent and
verified market data. For this purpose, the developer should
not only focus on the data retrieved by independent investiga-
tors but also assess it under various usage patterns covering
longer durations.
Table 3 lists smartphone based sensors that are attracting
the researchers to utilize them for various agriculture pur-
poses. While, last column provide some of the references
where these sensors have been used. Further, Table 4 includes
some of the important mobile apps developed for various agri-
culture applications along their features and achievements. (for example, on farming and the processing of agricultural
products). To make it more effective, the scenario can be
F. CLOUD COMPUTING extended further to include access to consumer databases,
Precision agriculture is showing its potential and benefits by supply chains, and billing systems.
improving agricultural operations through better data-driven Surely, moving towards Cloud-based services offer oppor-
decision making. However, to continue this success, precision tunities to explore advancements, but it comes with new
agriculture not only requires better technology and tools to challenges, as well. First, a vast range of sensors are being
process data efficiently but also at a reasonable cost such developed and used in precision agriculture, each of which
that the received data can be used to make field decisions has its own data format and semantics. Secondly, most of
efficiently. For this purpose, farmers can use Cloud services the decision-support systems are application-specific while,
to access information from predictive analysis institutes so on the other hand, a farmer can be in the need of accessing var-
that they can choose the right product available according to ious systems for a specific application, e.g., soil monitoring.
their specific requirements. Cloud computing offers an edge Considering both of these cases, the Cloud-based decision-
to farmers to use knowledge-based repositories that contain support system not only needs to handle the diversity of data
a treasure of information and experiences related to farming and their formats but also must be able to configure these
practices as well as on equipment options available in the formats for different applications.
market with the necessary details. In most cases, all this An open Cloud-based system has been established by
comes along with expert advice from a wide range of sources AgJunction [243] which gathers and disseminates the data on

VOLUME 7, 2019 129565

 M. Ayaz et al.: IoT-Based Smart Agriculture: Toward Making the Fields Talk

TABLE 4. Selected important smartphone applications for various agriculture applications.

129566 VOLUME 7, 2019

M. Ayaz et al.: IoT-Based Smart Agriculture: Toward Making the Fields Talk

a form from different precise agriculture controllers, leading large areas in order to harvest data for further processing
to a decrease in costs and environmental impacts. Further- and analysis. Furthermore, UAVs, better known as drones,
more, ‘‘Akisai’’ Cloud [244], proposed by Fujitsu, focuses on fitted with high-resolution cameras and precise sensors, can
food and agricultural industries and incorporates information be flown over thousands of hectares of farms.
communication technology for increasing the food supply in The role of surveillance in all agriculture applications is
the coming years. highly critical, especially in forestry and crop monitoring due
Similarly, SourceTrace developed and offering Cloud- to the need to cover large areas [246]. Therefore, a fast, low-
based mobile applications to provide visibility and relations cost, real-time, and large-scale surveillance supported with an
between farms and markets, further tracking the value chain at accurate data acquisition and transmission facility is crucial
the source, e.g., ’eService Everywhere’ [245]. An important for agriculture production. Currently, mostly two options are
note about their applications is that, during the development, used to obtain aerial images of a field area: satellite and
they considered the farms’ remoteness and low bandwidth airplanes. Both of them are good for a macro view of a
environments. landscape, but they face serious issues in terms of quality
Figure 9 presents possible infrastructure and relationship when it comes to micro views. These macro-view images are
scenario of fluid computing including Edge, Mist and Fog not good in resolution and cannot offer the image quality
for smart agriculture. which is required during the analyses and decision mak-
ing. Secondly, not only the resolution but visiting frequency
also matters and, through both of these, it is not simple to
take and collect images frequently (on average, four times a
month [247]). Another serous issue is that these operate above
the cloud level where there is a strong possibility that both are
obstructed in bad weather.
When we talk about UAVs that provide an ‘‘eye in sky’’,
we can overcome— or even eliminate— the above mentioned
issues when we consider the micro views. The quality of
images taken through UAVs depends on the attached camera’s
resolution—normally dozens of times better than satellite
images—and, most importantly, we can adjust according to
application requirements. More specifically, UAVs supports
faster and better NDVI to assess crop conditions, like weed
mapping, leaf assessments, etc., and provide immediate feed-
back so that farmers can take timely actions. Similarly, UAVs
are better in terms of frequency, even if requires multiple
times in a single day, and are also the option least affected by
weather conditions, unless it is raining. Due to the mentioned
advantages, UAVs are considered the future of precision
agriculture, and this is the reason they are generating the
highest revenue amongst all agricultural robots developed for
precision agriculture. According to quoted figures by a report
published by FAO in 2018, it is estimated that the agriculture
FIGURE 9. Fluid computing infrastructure for smart farming. drone-related market to be worth USD 32.4 billion [248].
The current condition of the entire field is one of the most
valuable pieces of information to obtain in the precision pro-
V. UAVs IN AGRICULTURE gram. With the help of this collected data, a farmer can spot
Recently, the IoT has made remarkable progress in many problems early and rapidly; hence, appropriate interventions
industries, including farming sectors like poultry, fishing, can be applied. Agricultural drones represent a new way to
etc. but when we talk about agriculture, the communication collect field-level data; the results are on-demand whenever
facilities like base stations or Wi-Fi are very limited, which and wherever needed, as the drone can be easily and quickly
prevents the growth of the IoT in this sector. Such commu- deployed. Most importantly, it is not all about their hardware
nication infrastructure and related facilities are even worst in but the convenience, quality, and utility they are offering,
developing counties and rural areas, which is one of the major as the drone-enabled surveillance offers the real facility to
hurdles when introducing the IoT in the agriculture industry. have an idea of what is happening in the farm fields at that
The data acquired through the wireless sensors cannot be moment.
transmitted in the absence of reliable communication infras- The UAVs, used for agricultural applications usually fall
tructure. In such a scenario, UAVs offer an alternative, as they into two categories: fixed-wing and multi-rotor drones [249]
visit and communicate with the wireless sensors spread over (figure 10). Although both are available in various ranges

VOLUME 7, 2019 129567

 M. Ayaz et al.: IoT-Based Smart Agriculture: Toward Making the Fields Talk

FIGURE 11. NDVI based water stress map of 160 acre walnut orchard [258].

Due to their nature and flexibility, UAVs are being used

in a range of agricultural applications, including crop health
FIGURE 10. Types of agricultural drones.
monitoring, planting, plant counting, spraying, agriculture
photography, and many other variable rate applications.
in terms of cost, payload capacity and mostly distinguished After being equipped with automation and GPS capabilities,
based on hardware differences. For example, when it is they are ready to take the agriculture sector to a further-
required to cover a large area, fixed-wing drones are sug- modernized level. With every passing day, drones are becom-
gested due to their long-range flight capacity, most impor- ing more inexpensive and reliable, hence, making themselves
tantly they are crash tolerant e.g senseFly’s eBee SQ [250] an ideal choice for new farming applications. Focusing on
and DATAhawk [251]. On the other hand, multi-rotor drones the success of this technology, SAP (Systems, Applications,
are more common due to their easy and faster set up as and Products), one of the largest vendors of enterprise
can take off and land vertically. Multi-rotors actually have resource planning, has brought three of its major technolo-
many advantages over the fixed wings as they are easier gies together in order to make the information harvested
to operate, require no advance wind planning and have through UAVs more effective [257]. The technologies include
the ability to fly more precisely. Moreover, in scenarios the HANA, a Cloud database which supports speedy data
where low altitude flight is required in order to capture capture, retrieval, and analytics; the Leonardo IoT suite to
extremely detailed images, which is more common in agri- connect and exchange information over any protocol; and
culture applications then the multi-rotor are considered the the Connected Agriculture suite to provide a GUI-based,
better choice, some major examples belongs to this type are graphical-use interface to the farmers.
DJI Matrice 200 [252] and Introducing Scout by American While it depends on the situation, in most cases, few hours
Robotics [253] which considered a fully autonomous drone delays doesn’t leave serious impact in most of the agriculture
for daily agriculture scouting. applications but flight time is more important as need to cover
Generally, drones collect information through the light larger areas due field vastness.
reflected by the ground it is maneuvering, over, either from Not only vastness, but in some applications of pesticide
soil or plants. For agricultural applications, specific cameras and fertilizer, UAVs need to carry heavy payloads. In such
and sensors are used, depending on the grower’s interest— situations, optimum battery utilization becomes crucial to
most commonly mentioned are thermal and hyper-spectral. extend the flight time. For this purpose, many factors can
Thermal sensors can help to recognize the water quan- be considered to increase the drone efficiency. Firstly, when
tity, as leaves of plants with more water access appear flying, choose right conditions e.g. weather or air direction.
cooler in an image. The same phenomenon is used in near- Next, try to include optimum payload and place it appropri-
infrared (NIR) sensors; commonly used to note the difference ately. For this situation, it can be helpful to attach the payload
between the NIR reflectance and the visible reflectance, such near the field, better in smaller quantities and the refilling
as NDVI [254]. The resultant NDVI-based images help to again instead of putting heavy quantities. Further, depending
distinguish the water stress areas, as shown in figure 11. on area size and visiting frequency, optimum path selection
Hyper-spectral based sensors or cameras record the wave- plays a critical role. For this purpose, many routing schemes
lengths of both visible and invisible lights, are able to identify like [259], [260] are proposed especially for the UAVs so
the specific type of plant by measuring the color of reflected choosing and implementing the right scheme can provide
light. This reflected light is used to distinguish various plant clear difference. Considering the application of pesticide and
types, ultimately helping to detect the unwanted herbicide and UAV based irrigation where drone need to fly with heavy
weeds [255]. The idea is used in [256], where authors used the payloads then new procedures like tethering system can be
multispectral images to classify various weeds. helpful. In UAV tethering, a connection that provide power

129568 VOLUME 7, 2019

M. Ayaz et al.: IoT-Based Smart Agriculture: Toward Making the Fields Talk

through the long cable, is provided so that it can fly as long like its geometry and nutrients ultimately help to optimize
as you have power backup on the ground, most importantly it the crop management operations.
doesn’t require to lift heavy batteries.
Currently, agriculture is being considered one of the most D. IRRIGATION
favorable fields where UAVs can offer solutions to resolve Use of drones for irrigation applications is, again, two-fold.
many dominant and long-lasting issues. Some of key areas in On one side, equipping UAVs with a variety of sensors
which drones are already playing key roles to assist farmers and cameras can help to identify areas that are under water
throughout the crop cycle are highlighted below. stress and conclude what irrigation changes are required.
At the same time, they can be used for sprinkling water and
A. SOIL AND FIELD ANALYSIS pesticides on the crops precisely, especially in emergency
Drones are able to produce precise information to analyze cases, which would save both time and wastage. In [272],
the soil before sowing the crop, which helps to determine the multispectral images of citrus crops were acquired using the
most suitable crop for specific land; furthermore, it suggests fixed-wing UAV, where the retrieved data was used to assess
the seed type and its planting patterns. In [261] authors shared and detect structural and physiological changes in the targeted
their experimental results using Sirius I, a fixed-wing aircraft, crop. Further, [273], [274] are similar efforts in which UAVs
affixed with a Lumix GF1 digital camera by Panasonic to were used to estimate the crop water stress. Furthermore,
capture images from different sites to monitor the soil erosion UAVs are not only used to analyze the irrigation properties but
issues in Morocco. Similarly, authors in [262] targeted the also provide solutions by sprinkling water precisely over the
issues of soil analyses where they used Lumix DMC-LX 3 to water stress areas as in [275]. Due to this application of UAVs,
take the images and Pix4UAV for mapping the results. they are being considered the newest water-saving tool,
while their use is helping not only to increase watering effi-
ciency but also detect possible pooling or leaks in irrigation.
B. PLANTING
Examples like ’JT20L-606’ [276] and ’AGRASMG-1’ [277]
Millions of acres of land are currently under-utilized due are specialized drones that were developed and are being used
to being human inaccessible or lack of suitable workers. for this purpose.
Safety concerns of rough terrain are main reason not to utilize
these areas for forestry or agriculture purpose. For this pur- E. PLANT COUNTING AND GAP DETECTION
pose, drone based planting systems are being developed that Precision agriculture critically needs the spatial data on crop
decrease planting costs upto 85 percent [263]. Not only cost, density when making decisions during various applications.
but within shorter time as some recently developed drones can The quantity and plant numbering not only reflects the field
plant 100,000 trees in a single day [264]. These systems shoot emergence but allows better and more precise assessment
pods which include the seeds and necessary nutrients required of the yield production, in fact, determining the crop fate.
to grow the plant. This method is found very effective for Again, UAVs are offering flexible solutions for this pur-
rough terrain; most importantly the success rate is more than pose. In [278], authors performed digital counting of Maize
75% [265]. Due to the success and flexibility they offer, UAVs plants with the help of UAVs. Further, in [279], authors
are being considered the best candidate for plantation all proposed a method in which they used UAVs to estimate
over the world, from NASA engineer [266] to countries like the density of wheat plants at the emergence stage while a
Pakistan [267] and India [268]. Sony ILCE α5100L RGB camera was used to take the
images.
C. CROP MONITORING
Crop monitoring is one of tough jobs and facing low effi- F. SPRAYING THE PESTICIDES/HERBICIDES
ciency due to covering large area. Drones are offering the Similar to irrigation, UAVs can be used to spray herbicides/
solutions by allowing real-time monitoring of far farms, more pesticides on crops, but their use for these applications is
accurately and cost-effectively comparing with previously more critical. Spraying application would be highly efficient
used satellite imagery. The Microdrones +m [269] is an compared to current procedures; herbicides/pesticides are
accessory toolkit which provides aerial imaging facility to usually sprayed over the entire farm, which is not required in
observe the crop nutrients, moisture levels and monitoring most cases. If using an UAV to spray herbicide, it can spray
of other necessary parameters. A study conducted in [270] directly on the unwanted weeds or can target the affected
where authors used UAVs along digital camera to monitor areas only. Furthermore, as spraying using drones would be
the crop conditions. The purpose of the study was to find highly targeted, the drone would figure out and spray as
the relationship between the crop spectral characteristics and per requirements, helping to reduce the overall expenditures.
effect of fertilizer availability for plant health. Further, [271] Handling the sudden environment changes like wind direc-
presents an innovative procedure to compute and map the tion or speed is another issue for an UAV especially when
3-dimensional geometric characteristics of trees and tree- being used for spraying applications. For this purpose, [280]
rows. The generated maps can be helpful to understand the proposed a computer based system that autonomously adopt
relation between the trees’ growth and field related factors the UAV control rules to keep precise pesticide deposition.

VOLUME 7, 2019 129569

 M. Ayaz et al.: IoT-Based Smart Agriculture: Toward Making the Fields Talk

G. HEALTH ASSESSMENT existing available food safely and efficiently to more people
Scanning crops with visible and Infrared (IR) light sensors is the real subject of issue of the food industry [289].
fitted on drones can identify which plants may be infected A comprehensive report regarding the future food require-
by bacteria or fungus. Using UAVs, this can be done fre- ments published by World Resources Institutes (WRI)
quently and with flexibility. The early detection of any such in 2018, highlights that we need sustainable food industry to
issues helps to prevent the disease being spread to other feed 10 billion people by 2050. The report suggests a five-
plants or crop areas. Multispectral images can help to detect course menu of solutions to tackle the future food issues
the disease or sickness at early stages even before reaching where ‘‘reduction in food losses and waste’’ is declared as
the level in which it is possible to detect with the human the first and most important course [290]. Further it conclude
eye. Experiments done in [280] present the data collection that, reducing the food loss and waste only by 25% can
campaign performed over a sorghum crop which was severely help to reduce the food gap by 12%, the land gap by 27%
damaged by white grubs. Further, in [281], UAVs are used to and greenhouse gas mitigation gap by 15%. To have better
collect data from ground-based sensors, including a chloro- idea, detailed figures about the food loss are highlighted
phyll meter, water potential meter, and spectroradiometer, in figure 12 based on various geographical locations and
and the collected information was used to evaluate the plant considering the stages along food supply chain where these
health and crop condition, ultimately reflecting the ground losses are occurring.
truth.

H. DETECTION/RECOGNITION OF PLANT SPECIES

Recently, UAVs are being started to detect and recognize
the plant species, especially those are considered extinct or
remains only few on our Earth. An UAV is a best candidate to
perform this task as it can go in very remote locations those
are almost inaccessible to humans. As National Tropical
Botanical Garden (NTBG) stated that, Hibiscadelphus woodi,
a Hawaiian flower, which thought to be extinct in 2009, is
discovered on a vertical cliff face using the drone [282].
Similarly, using the special cameras that can detect the
forest biomass and fuel, forest fuel estimation is possible
through UAVs which is mostly being done from radars and
satellites [283], [284].
FIGURE 12. Food losses occurring along food supply chain [290].

VI. FOOD SAFETY AND TRANSPORTATION

With the increasing population, we have to feed around 25% Other than WRI, the figures released by United Nations’
more mouths until 2050, as compared to 2010, which appears FAO are also shocking as it is estimated that one-third of all
to be a monumental task when considering the vast hunger the global food produced for human use, valued at $1 trillion,
issues that the world is currently facing. The numbers are is lost or wasted each year [291]. The US food waste alone
shocking, as more than one-quarter of the planet’s population represents 1.3% of its total GDP [292]. Based on these scary
suffers from malnutrition [285] and nearly one billion people figures, it can be concluded that food waste is massive market
are chronically hungry [286]. Considering this current hunger inefficiency—the kind of which does not persist in other
scenario, it could be big ask to feed the new billions of industries [293]. Although, the loss of $1 trillion has its
mouths in the future. In order to handle the situation, some own worth, but, more crucially, environmental implications of
believe the solution lies simply in growing more food. This these losses are real, such as the water wasted to produce the
has also been reinforced by a latest research, published in the food that is never eaten being equal to the water that can fulfill
journal Bioscience, which concludes that from the present the needs of all of Africa [285]. Furthermore the CO2 emis-
time until 2050, the rise in total food production should be sions that resulted during the growing process of this food
in the bracket of 25-70%. [287]. However, what if someone are equivalent to removing every car off the road across the
tells you that there’s already enough food currently grown on world [290]. Moreover, according to another study conducted
farms to feed 10 billion people? [288]. by FAO, if you consider food waste as a source of green-
Yes, the good news is that we are already producing enough house gas emissions, it would be the third biggest emitter
food for that many people; however, it is a crucial matter to in the world behind only China and the United States [294].
figure out that how to distribute this food while maintaining In addition, when food waste goes to the landfill, where it
its quality. Considering this fact, one can say that increasing mostly ends up, it decomposes in the absence of oxygen and
the food production is also important, but transporting the produces methane, which is 23 times more potent than carbon

129570 VOLUME 7, 2019

M. Ayaz et al.: IoT-Based Smart Agriculture: Toward Making the Fields Talk

dioxide in terms of negative effects. In short, every which way To provide the recommended environment, a device with
you look at it; food waste is a major culprit in destroying our supported technology can be installed at the storage site, even
planet. in transport trucks. Further, it is linked to an online dashboard
Among all perishable food produced in the world today, that can be configured to send alerts in the event of abnormal
only 10% is preserved properly [295]. When we talk about temperature levels to trigger swift remedial action. Some of
most of the developed countries, a robust food cold chain key technologies available for this purpose and their use cases
is maintained where essential quality checks are followed, are mentioned here.
entailing temperature-regulated refrigerated warehouses to
refrigerated trucks to ensure that food gets from farm to A. COMPLIANCEMATE
market safely. On the other hand, many developing countries Compliance with hazard analysis and critical control points
lack such proper cold chain infrastructure, simply resulting (HACCP) offers a food safety and quality monitoring pro-
the majority of food spoiling when being transported to the gram which collects temperature data inside coolers and other
end-user. Considering this fact, there is a huge opportunity kitchen equipment continuously. For example, its integration
to cut food waste and improve food distribution by simply with Touchblock is used to capture temperatures in coolers
implementing a controlled-temperature transportation sys- and prep rooms at every minute [297].
tem. Based on the facts, one can conclude that increasing
food production is not sufficient to achieve food security, B. LAIRD’S SENTRIUS
but, rather, some practical actions are required to find skillful A battery-powered and long-range integrated sensor platform
ways for efficient distribution of the already available food. that leverages the benefits of LoRaWAN and Bluetooth Low
There are different ways to monitor and control food tem- Energy (BLE) connectivity. It provides LoRaWAN options
perature. The manual method of checking a thermometer at 868/915 MHz, based on the Semtech SX1272 and Nordic
and recording the temperature has many drawbacks, where nRF51 silicon. Further, it offers high RF performance in a pre-
someone must actually do it and, most importantly, take cise temperature and humidity. Two major series, including
the readings correctly. On other side, implementing an auto- RS1xx and RG1xx (multi-wireless gateways), work together
matic method that uses wireless sensors to electronically in order to provide Cloud-based services. Most importantly,
measure and record temperatures can substantially improve it requires an inexpensive endpoint radio and a more sophis-
food safety. This method allows for a continuous data stream ticated base station to manage the network. As compared to
of temperatures simply—24 hours a day, 7 days a week. LoRa, Sigfox communication tends to be better if it is headed
By doing so, temperatures can be recorded consistently up from the endpoint to the base station. Although it supports
and on time, leaving little room for interpretation; in short, the bidirectional functionality, its capacity going from the
the entire process is based on facts and nothing more. Further, base station back to the endpoint is constrained, as it provides
utilizing the recent technologies, the recorded data can be less link width going down than going up.
stored in the Cloud and accessed via any type of internet-
connected device. Notifications can be established that will C. CCP SMART TAG (RC4)
send real-time alerts if the temperature strays outside preset CCP claims to be a complete monitoring solution for the
limits, allowing for immediate action to remedy the situation. food service and food retail industry [298]. It is capable to
Further, IoT offers predictive maintenance and indicate when automate the temperature environment which meets the food
the monitoring equipment itself is going to end its useful safety regulations suggested for various food items.
life so it can be replaced before it fails and compromises Further, temperature and other data are interpreted and
product quality. These are only couple of scenarios, now if viewed on a service provider Cloud platform via web and
we consider the figures presented in figure 12, IoT has the mobile applications.
potential to monitor and keep the food quality at every stage
of the supply chain, from production to consumption. D. TEMPREPORTER
A research study conducted by Indian School of In compliance with HACCP (Hazard Analysis and Critical
Business in which students worked with a local grower to Control Points), it is used to monitor temperatures 24/7.
transport fruits and vegetables in refrigerated trucks from Further, it logs the readings automatically. Reports are auto-
Punjab to Bangalore, a distance of more than 2,500 km filled by considering HACCP & HPRA (Health Products
through rough roads under high temperatures. The results Regulatory Authority) recommendations regarding tempera-
were highly encouraging to implement the cold chain to ture monitoring.
transport the agriculture products. The out of conducted study According to Finistere Ventures report, as of 2018, around
brought benefits in three ways: (1) increased food shelf-life $2 billion has been invested globally in AgTech. Several
from one week to two months; (2) an up to 23% higher investments are expected to cross these figures in 2019.
profit for everyone linked in the supply chain was observed; Considering the future needs of IoT in agriculture applica-
(3) a 76% reduction of food wastage (post-harvest). Besides tions, almost all leading technological giants are supporting
all this, another critical factor is the emission of greenhouse this progress in their own way. Table 5 provides a list of sev-
gases was observed to be reduced by 16% [296]. eral of the leading global organizations who have proposed

VOLUME 7, 2019 129571

 M. Ayaz et al.: IoT-Based Smart Agriculture: Toward Making the Fields Talk

TABLE 5. Current status and future vision of major technological giants regarding the iot in agriculture industry.

129572 VOLUME 7, 2019

M. Ayaz et al.: IoT-Based Smart Agriculture: Toward Making the Fields Talk

TABLE 5. (Continued.) Current status and future vision of major technological giants regarding the iot in agriculture industry.

and are proposing initiatives in AgTech, especially in regard

to IoT-based agriculture solutions.

VII. CURRENT CHALLENGES AND FUTURE EXPECTATIONS

According to a plan announced in 2015 under ‘‘The
2030 Agenda for Sustainable Development’’, the UN and
international community set a target to end hunger by 2030.
However, recent figures released by WHO (World Health
Organization) do not look encouraging enough to support
the agenda, as more than 800 million people worldwide are
facing the food shortage—one out of every nine people [313].
Although these figures are quite alarming on their own, what
is more shocking is the quality of food. Other than availabil-
ity, the quality of food is becoming another serious issue and
even more critical. FIGURE 13. Major challenges for sustainable future agriculture.
According to a research supported by the Bill & Melinda
Gates Foundation published in ‘‘The Lancet’’, either short- conditions, and many more. As the world moves toward
age or poor diets are diverting 11 million people to an early urbanization, the rural populations are not only shrinking but
grave annually, making it more deadly than smoking [314]. are rapidly aging; hence, fewer and younger growers need to
The research, which reflects the effect of poor diet on health, step up to take the responsibility. Such population imbalance
was conducted in 195 countries from 1990 to 2017 and con- and generation shift can create serious implications, not only
cluded that one out of five deaths per year could be prevented for the remaining workforce but also for production patterns
by providing better diet. The report summarizes that, globally, and land tenure. Furthermore, on one side, arable land is
a diet lower in whole grains was the most common and lead- shrinking while, among the remaining regions, many are
ing risk factor for deaths. Other than the basic food needs, per only suitable for specific crops due to certain geographic and
capita incomes of most of the countries in 2050 are expected environmental limitations. Moreover, harsh climate changes
to be a multiple as compared to today’s levels [315]. Such are starting to affect almost every aspect of crop production.
an increase in income will result in a more health-conscious These changes are expected to enhance the intensity of many
population that expects quality food that is rich in fiber and of the existing long-term environmental issues, like droughts,
other minerals. Trends, like increased population where the floods, groundwater depletion, soil degradation, etc.
world needs to feed one third more mouths with increased During the 20th century, in most regions, growers kept
demand of quality food, show that food demands continue to following the traditional agricultural methods while trying
grow rapidly. to meet the food demands by greater utilizing fertilizers and
In response to all this, overall crop production needs to pesticides. Implementation of such chemicals is facing two
increase not only for food but cash-crops are also required to issues: these can help to increase the production to only a
grow in order to fulfill the demands of industry, like cotton certain level and, at the same time, their blind use is creating
and rubber, and, most importantly, increasing demands for irreversible implications to the environment. Furthermore,
bioenergy like ethanol. implementation of any resource, like water, seed, fertilizers,
Figure 13 presents a snapshot of major challenges that and pesticides, uniformly across an entire field is not going
future agriculture expected to face in 2050. This diagram, to solve the problem. Rather than dealing with every farm
basically presents three major issues: how to feed around and crop in same way, farmers need to use these resources
10 billion people; without using more land and; by reduc- according to the requirement of specific areas, even if they
ing the emission of greenhouse gasses by more than 60%. have to consider the requirement of every plant.
However, when we look closely then these three challenges Focusing on the above discussion, one can feel that the
lead to many new, including smaller rural labor, continu- farms and relevant crop operations need to be run differently
ously shrinking arable land, water scarcity, harsh weather than the past practices. One of the major reasons is the

VOLUME 7, 2019 129573

 M. Ayaz et al.: IoT-Based Smart Agriculture: Toward Making the Fields Talk

advancements in technology, including sensors, communica- their land is. Based on recent success, it is estimated that
tion methods, machines, and even robots. In fact, technology more than 75 million IoT-based devices will be operating in
has proved this already, as, in most developing countries; the agriculture industry by 2020. Further, the average farm is
more than 50% of the population is somehow engaged in expecting to generate 4.1 million data points on a daily bases
the agriculture industry yet are far behind in providing both by 2050 [316].
the quantity and quality when compared with the developed
countries, where less than 2% of the population is perform- B. COMMUNICATION
ing much better. The difference is clear, as countries like Real success of IoT in agriculture largely depends on
Australia, the US, and most of Europe are pioneers to employ advances in connectivity. From telecom’s perspective, pro-
the advanced tools and methods multiplying the crop yields viding mainly connectivity and other value-added services
during the last five decades. These comparisons show that has an immense potential and can influence the entire chain
recent technologies and advanced methods are making the greatly [317]. Most of the telecom operators around the globe
farms not only highly profitable but safe and environmentally offer connectivity services, but such services only represent a
friendly. tiny amount of the entire smart agriculture market. Consider-
Considering this scenario, future agriculture is expected ing its worth, especially in rural areas, the cellular operators
to evolve as a high-tech industry where interconnected sys- have to offer a new range of services targeting the growers’
tems will enjoy the luxury of artificial intelligence and demands. Accepting the fact that most of the community
Big Data facilities. The resultant systems will converge into belong to this industry are not highly educated and mostly
a single unit where farm machinery and management, start- unaware of new technologies, hence the operator should
ing from seeding to production forecasting, are combined. provide end-to-end solutions other than just providing the
By involving the advanced technologies like agricultural connectivity. If so, then it will certainly help to increase the
robots, Big Data, and cloud-computing artificial intelligence, market share of mobile and telecom operators. Further, these
agriculture can create a new era of superfusion. Following are operators need partnerships with the investors to provide end-
some of the key technologies and methods that need to apply; to-end solutions, which demands higher investment, even
focusing to achieve sustainable future agriculture. before advantages can be seen. The results of success when
inviting the investors depend on the nature of the partnership
A. WIRELESS SENSORS AND THE IOT and the involved bodies, like device manufacturers, solution
Wireless sensors placed strategically around fields are provid- providers, non-cellular connectivity service providers, sys-
ing farmers with up-to-date information in real time, allowing tem integrators, etc. On one side, the outcome of this partner-
them to adapt the care that the crops need, which results ship would help operators to enter deeper into the industry,
in higher food production with less waste. Wireless sensor ultimately boosting their market share. At the same time,
networks (WSNs) are also being used to inform farmers this opportunity can create strong relationships among the
about nearly all aspects of their crop growth as well as about organizations and farmers to help to educate them about the
the readiness state of the farm’s machinery, thus, helping in benefits of smart agriculture.
preventing loss of crop as well as enhancing the readiness of The success of cellular technology is only possible when
the equipment which cultivates it. WSNs with GPS capabil- service providers leverage its real benefits like portability,
ity are helping tractors in compensating for uneven terrain flexibility and luxury of two way communication to offer
and optimizing land preparation for growing crops. Recently, low cost but customized solutions. They must provide what
advances in image recognition and digital signal processing the farmer is in need, at the place they choose. Furthermore,
gave even more capabilities to WSN to accurately determine to provide fast penetration in agriculture industry, policy
crop quality and health. changes are required in order to provide access of reliable
In order to make agriculture sustainable, the use of the IoT and quality inputs. The research conducted in [172] which
will be at the center and forefront in agricultural operations. considers 23 studies where mostly belong to developing
This includes everything from water and power usage, crop countries, concludes that the cellular services and smartphone
transportation, farm machinery operation and maintenance technology carry a promising future for smallholder farmers
alerts, and market price updates. The IoT has the capability being capable to improve their yields.
to make these tasks streamlined and more predictable by Furthermore, licensed LPWA (low-power wide-area) tech-
recognizing the crop’s needs at every stage. It has already nology is expected to be a game changer for smart agri-
proved a breakthrough and is further going to change the culture. Due to its characteristics and supported services
way we look at various agriculture activities by providing the including low power consumption and efficient coverage are
farmer control over their land and assets in an unprecedented well suited to the geography and economics of agriculture
way, thus, maximizing their effectiveness and efficiency. hence expected to play a critical role in future smart farming.
Further, the future of the IoT can be shaped by the phenom- Consequently, narrowband IoT (NB-IoT) got strong industry
enal advances in WSNs and the fifth generation (5G) of cel- support and becoming an effective global standard for LPWA
lular mobile communication technologies to provide farmers connectivity. It has the potential to provide major connectivity
with real-time data and information anytime and everywhere changes in agriculture industry by changing the perceptions

129574 VOLUME 7, 2019

M. Ayaz et al.: IoT-Based Smart Agriculture: Toward Making the Fields Talk

about Internet capabilities. Believing in its future success, One of them is power issue as due to its nature; smart farming
it is expected that leading cellular operators with strong IoT requires wide use of energy. Among the main reasons of
ambitions can generate significant revenues by providing extensive power consumption some are including, long term
smart agriculture services when collaborating with LPWA sensor deployment, use of GPS repeatedly and transmission
technology providers. In order to achieve long term success of of sensed data via GPRS. Traditionally, farmers in remote
these short, mid and long range communication technologies, areas have bought and utilized renewable energy sources
necessary steps for infrastructure construction are required randomly and at a hefty price, which has limited their ability
towards attaining the technology-based agriculture. to use them in farming to a great extent. However to solve
the power issues in long term, deep analysis of power con-
C. UAVs AND OTHER ROBOTS
sumption sources like remote data transmission can help to
Drones are being widely used by farmers for crop growth tackle the problem at some extent. Further, smart grids and
monitoring and as a means to combat hunger and other microgrids, however, lend themselves to seamless integration
harmful environmental impacts. Furthermore, they are being of distributed energy sources (DERs), thus, making them
used to spray water and other pesticides efficiently, consid- appealing for adoption by farmers. The emergence of smart
ering the tough terrains, especially when the crops possess power meters has further given the farmers the confidence
different heights. Drones have proven their value, not only to invest in DER, especially since they have the option to
in terms of spraying speed but precision, as well, when sell the excess power to the grid. Recent advances in energy
compared with traditional machinery of same purpose. With storage devices, integrated electricity and heat systems will
recent advances in swarm technology and mission-based con- make DER even more attractive for farmers, as they will be
trol, groups of drones equipped with heterogeneous sensors, able to store energy and use the heat generated by cooling and
including 3D cameras, can work together to provide farmers heating when needed. However, healthy investment require-
with comprehensive capabilities to manage their land. With ments and public perceptions are two other barriers on the
the inclusion of UAVs in agriculture, farmers are able to put way to making these solutions successful.
their eye in sky, but many challenges need to be addressed
in order to enjoy the real advantages of this technology, F. HYDROPONICS AND VERTICAL FARMING (VF)
especially the integration of other technologies and how to Other than employing the advanced technologies, new agri-
use them in poor weather conditions. cultural practices can be very crucial to overcome the geo-
Beside drones, robotics within agriculture have improved graphic and resource limitation challenges. On one side,
productivity and resulted in higher and faster yields. Such arable land is shrinking, and, at the same time, it is estimated
robots, like spraying and weeding robots, are reducing agro- that three million people around the globe are migrating
chemical use. Robots equipped with laser and camera guid- to cities, resulting in more pressure on the existing limited
ance are being used for identifying and removing weeds urban resources [318]. Considering this rapid migration, it is
without human intervention. They navigate between rows estimated that by 2030, 60% of the world’s population is
of crops on their own, ultimately increasing the yield with going to depend on cities, and this number is further expected
reduced manpower. More recently, plant-transplanting and to rise to 68% until 2050 [319]. Considering both of these
fruit-picking robots are emerging to add a new level of issues, it could be disaster for food production in the near
efficiency to traditional methods. future with current agriculture practices. VF is an answer of
D. MACHINE LEARNING AND ANALYTICS these issues, as it meets the challenges of land and water
Machine learning and analytics are used to mine data for shortage and, at the same time, looks highly suitable to
trends. In farming, machine learning is used, for example, be adopted near the cities. VF is portrayed as the answer
to predict which genes are best suited for crop production. to the looming shortage of food and shrinking arable land,
This has been giving growers all over the world the best seed at least in some areas of the world. Further, hydroponics can
varieties, those which are highly suitable to respective loca- play a key role, as this method lowers the requirements of
tions and climate conditions. Machine-learning algorithms, water and space to a great extent. Rapid growths in computer
on the other hand, have indicated which products are of high power are propelling scientific discoveries in plant nutrition
demand and which products are currently unavailable in the and growth that would make VF even more appealing to
market. Thus, for the farmer, this has given valuable clues for growers.
future farming. Recent advances in machine learning and ana- Along VF and hydroponics, new and advanced solutions
lytics will make it possible for farmers to accurately classify are required to increase the arable land without disturb-
their products and weed out less desirable crops before they ing the forests and other natural animal habitats. For this,
arrive to customers. we have to focus on the deserts as these cover one third of
the Earth’s land surface. The solutions are started already as
E. POWER CONSUMPTION, RENEWABLE ENERGY, Norwegian and Chinese firms/experts are doing efforts in
MICROGRIDS, AND SMART GRIDS Dubai, Qatar, Jordan and Chinese deserts [320]–[322].
Despite its future opportunities, smart agriculture facing Agriculture is not just an industry; in fact, it provides the
some limitations that are holding back the growth of IoT. basis of human society, as the goal is not just to grow crops,

VOLUME 7, 2019 129575

 M. Ayaz et al.: IoT-Based Smart Agriculture: Toward Making the Fields Talk

but the target is the perfection of human being. A vibrant REFERENCES

and prosperous agriculture sector can provide the basis for a [1] World Population Projected to Reach 9.8 Billion in 2050, and 11.2 Billion
happy and healthy society, as recent decades witnessed this. in 2100. Accessed: Apr. 18, 2019. [Online]. Available: https://www.un.
org/development/desa/en/news/population/world-population-prospects-
The presence of advance technologies, especially the involve- 2017.html
ment of the IoT, matters a great deal in regard to reaching this [2] How is the Global Population Distributed Across the World?
goal. Environmental issues continue to cage the planet, which Accessed: Apr. 13, 2019. [Online]. Available: https://ourworldindata.
org/world-population-growth
increases the need for safe and clean agriculture. This is the [3] 68% of the World Population Projected to Live in Urban Areas by 2050,
reason humanity is witnessing a second green revolution, Says UN. Accessed: Mar. 15, 2019. [Online]. Available: https://www.un.
largely based on the IoT. The use of these technologies makes org/development/desa/en/news/population/2018-revision-of-world-
urbanization-prospects.html
the farming industry highly productive with reduced labor [4] Food Production Must Double by 2050 to Meet Demand From World’s
and other resource consumption; same time, minimizing the Growing Population. Accessed: Apr. 5, 2019. [Online]. Available:
impact on the environment. https://www.un.org/press/en/2009/gaef3242.doc.htm
[5] X. Zhang and E. A. Davidson, ‘‘Improving nitrogen and water manage-
Our planet has the resources, but we have to learn how ment in crop production on a national scale,’’ in Proc. AGU Fall Meeting
to utilize them wisely and precisely. Sensible use of tech- Abstr., Dec. 2018.
nology can lead us where we can utilize these resources [6] How to Feed the World in 2050 by FAO. Accessed: Sep. 6, 2019. [Online].
efficiently in order to ensure the food security of the current Available: https://www.fao.org/wsfs/forum2050/wsfs-forum/en/
[7] A. D. Tripathi, R. Mishra, K. K. Maurya, R. B. Singh, and D. W. Wilson,
and coming generations. For this purpose, we need collec- ‘‘Estimates for world population and global food availability for global
tive efforts to build such institutions that can shape long- health,’’ The Role of Functional Food Security in Global Health. 2019,
term decisions and polices to eliminate hunger effectively. pp. 3–24.
[8] M. Elder and S. Hayashi, ‘‘A regional perspective on biofuels in Asia,’’ in
On this route, the experience, tools, and support from those Biofuels and Sustainability (Science for Sustainable Societies). Springer,
nations that have succeeded in overcoming hunger should 2018.
provide to those regions that are fighting to feed its local [9] World Agriculture: Towards 2015/2030 by FAO. Accessed: Apr. 15, 2019.
[Online]. Available: https://www.fao.org/3/a-y4252e.pdf
mouths. Although growth in every industry matters, growth [10] L. Zhang, I. K. Dabipi, and W. L. Brown, ‘‘Internet of Things applications
in agriculture, particularly among small growers, can be for agriculture,’’ in Internet of Things A to Z: Technologies and Applica-
highly effective to control the undernourishment issues, tions, Q. Hassan, Ed., 2018.
[11] S. Navulur and M. N. Giri Prasad, ‘‘Agricultural management through
as more than 70% of the population of developing countries wireless sensors and Internet of Things,’’ Int. J. Elect. Comput. Eng.,
belongs to rural areas and somehow depends on agriculture vol. 7, no. 6, pp. 3492–3499, 2017.
sector. [12] E. Sisinni, A. Saifullah, S. Han, U. Jennehag, and M. Gidlund, ‘‘Industrial
Internet of Things: Challenges, opportunities, and directions,’’ IEEE
Trans. Ind. Informat., vol. 14, no. 11, pp. 4724–4734, Nov. 2018.
[13] M. Ayaz, M. Ammad-Uddin, I. Baig, and E.-H. M. Aggoune, ‘‘Wireless
VIII. CONCLUSION sensor’s civil applications, prototypes, and future integration possibilities:
A review,’’ IEEE Sensors J., vol. 18, no. 1, pp. 4–30, Jan. 2018.
The focus on smarter, better, and more efficient crop growing
[14] J. Lin, W. Yu, N. Zhang, X. Yang, H. Zhang, and W. Zhao, ‘‘A sur-
methodologies is required in order to meet the growing food vey on Internet of things: Architecture, enabling technologies, security
demand of the increasing world population in the face of the and privacy, and applications,’’ IEEE Internet Things J., vol. 4, no. 5,
ever-shrinking arable land. The development of new meth- pp. 1125–1142, Oct. 2017.
[15] X. Hi, X. An, Q. Zhao, H. Liu, L. Xia, X. Sun, and Y. Guo, ‘‘State-of-the-
ods of improving crop yield and handling, one can readily art Internet of Things in protected agriculture,’’ Sensors, vol. 19, no. 8,
see currently: technology-weaned, innovative younger people p. 1833, 2019.
adopting farming as a profession, agriculture as a means for [16] O. Elijah, T. A. Rahman, I. Orikumhi, C. Y. Leow, and M. N. Hindia,
‘‘An overview of Internet of Things (IoT) and data analytics in agricul-
independence from fossil fuels, tracking the crop growth, ture: Benefits and challenges,’’ IEEE Internet Things J., vol. 5, no. 5,
safety and nutrition labeling, partnerships between growers, pp. 3758–3773, Oct. 2018.
suppliers, and retailers and buyers. This paper considered all [17] Code of Conduct on Agricultural Data Sharing Signing.
Accessed: Apr. 13, 2019. [Online]. Available: https://www.ecpa.eu/
these aspects and highlighted the role of various technologies, news/code-conduct-agricultural-data-sharing-signing
especially IoT, in order to make the agriculture smarter and [18] Industry 4.0 in Agriculture: Focus on IoT Aspects. Accessed: Sep. 5,
more efficient to meet future expectations. For this purpose, 2019. [Online]. Available: https://ec.europa.eu/growth/tools-databases/
dem/monitor/content/industry-40-agriculture-focus-iot-aspects
wireless sensors, UAVs, Cloud-computing, communication [19] K. Thea, C. Martin, M. Jeffrey, E. Gerhard, Z. Dimitrios, M. Edward, and
technologies are discussed thoroughly. Furthermore, a deeper P. Jeremy, ‘‘Food safety for food security: Relationship between global
insight on recent research efforts is provided. In addition, megatrends and developments in food safety,’’ Trends Food Sci. Technol.,
vol. 68, pp. 160–175, Oct. 2017.
various IoT-based architectures and platforms are provided [20] How Blockchain and IoT Tech will Guarantee Food Safety. Accessed:
with respect to agriculture applications. A summary of cur- Sep. 6, 2019. [Online]. Available: https://www.dataversity.net/how-
rent challenges facing the industry and future expectations are blockchain-and-iot-tech-will-guarantee-food-safety/
listed to provide guidance to researchers and engineers. [21] A. Khanna and S. Kaur, ‘‘Evolution of Internet of Things (IoT) and its
significant impact in the field of precision agriculture,’’ Comput. Electron.
Based on all this, it can be concluded that every inch of Agricult., vol. 157, pp. 218–231, Feb. 2019.
farmland is vital to maximize crop production. However, [22] A. Tzounis, N. Katsoulas, T. Bartzanas, and C. Kittas, ‘‘Internet of Things
to deal with every inch accordingly, the use of sustainable in agriculture, recent advances and future challenges,’’ Biosyst. Eng.,
vol. 164, pp. 31–48, Dec. 2017.
IoT-based sensors and communication technologies is not [23] C. Dinkins and C. Jones, ‘‘Interpretation of soil test reports for agricul-
optional—it is necessary. ture,’’ Montana State Univ., Bozeman, MT, USA, Tech. Rep., 2013.

129576 VOLUME 7, 2019

M. Ayaz et al.: IoT-Based Smart Agriculture: Toward Making the Fields Talk

[24] Accessed: Apr. 15, 2019. [Online]. Available: https://www.agrocares. [46] A. F. Colaço and J. P. Molin, ‘‘Variable rate fertilization in citrus: A long
com/en/products/lab-in-the-box/ term study,’’ Precis. Agricult., vol. 18, no. 2, pp. 169–191, Apr. 2017.
[25] J. Martínez-Fernández, A. González-Zamora, N. Sánchez, A. Gumuzzio, [47] B. Basso, B. Dumont, D. Cammarano, A. Pezzuolo, F. Marinello, and
and C. M. Herrero-Jiménez, ‘‘Satellite soil moisture for agricultural L. Sartori, ‘‘Environmental and Economic benefits of variable rate
drought monitoring: Assessment of the SMOS derived soil water deficit nitrogen fertilization in a nitrate vulnerable zone,’’ Sci. Total Environ.,
index,’’ Remote Sens. Environ., vol. 177, pp. 277–286, May 2016. vols. 545–546, pp. 227–235, Mar. 2016.
[26] T.-G. Vågen, L. A. Winowiecki, J. E. Tondoh, L. T. Desta, and [48] N. Khan, G. Medlock, S. Graves, and S. Anwar, ‘‘GPS guided
T. Gumbricht, ‘‘Mapping of soil properties and land degradation risk in autonomous navigation of a small agricultural robot with automated
Africa using MODIS reflectance,’’ Geoderma, vol. 263, pp. 216–225, fertilizing system,’’ SAE Tech. Paper 2018-01-0031, 2018.
Feb. 2016. [49] R. Raut, H. Varma, C. Mulla, and V. R. Pawar, ‘‘Soil monitoring, ferti-
[27] P. V. Santhi, N. Kapileswar, V. K. R. Chenchela, and C. H. V. S. Prasad, gation, and irrigation system using IoT for agricultural application,’’ in
‘‘Sensor and vision based autonomous AGRIBOT for sowing seeds,’’ in Intelligent Communication and Computational Technologies. Singapore:
Proc. Int. Conf. Energy, Commun., Data Anal. Soft Comput., Chennai, Springer, Oct. 2017, pp. 67–73.
Tamil Nadu, Aug. 2017, pp. 242–245. [50] A. González-Briones, J. A. Castellanos-Garzón, Y. M. Martín, J. Prieto,
[28] H. Karimi, H. Navid, B. Besharati, H. Behfar, and I. Eskandari, and J. M. Corchado, ‘‘A framework for knowledge discovery from wire-
‘‘A practical approach to comparative design of non-contact sensing less sensor networks in rural environments: A crop irrigation systems
techniques for seed flow rate detection,’’ Comput. Electron. Agricult., case study,’’ Wireless Commun. Mobile Comput., vol. 2018, Jul. 2018,
vol. 142, pp. 165–172, Nov. 2017. Art. no. 6089280.
[51] G. Villarrubia, J. F. De Paz, D. H. De La Iglesia, and J. Bajo, ‘‘Combining
[29] Ice, Snow, and Glaciers and the Water Cycle. Accessed: Apr. 26, 2019.
multi-agent systems and wireless sensor networks for monitoring crop
[Online]. Available: https://water.usgs.gov/edu/watercycleice.html
irrigation,’’ Sensors, vol. 17, no. 8, p. 1775, Aug. 2017.
[30] What Percent of Earth is Water? Accessed: Apr. 15, 2019. [Online]. [52] A. J. S. Neto, S. Zolnier, and D. L. de Carvalho Lopes, ‘‘Development
Available: https://phys.org/news/2014-12-percent-earth.html and evaluation of an automated system for fertigation control in soilless
[31] Water Facts—Worldwide Water Supply. Accessed: Apr. 15, 2019. tomato production,’’ Comput. Electron. Agricult., vol. 103, pp. 17–25,
[Online]. Available: https://www.usbr.gov/mp/arwec/water-facts-ww- Apr. 2014.
water-sup.html [53] G. Palomino and J. Miguel, ‘‘Protected crops in SPAIN: Technology of
[32] Water for Sustainable Food and Agriculture by FAO. Accessed: Apr. 15, fertigation control,’’ in Proc. Agri-Leadership Summit, Haryana, India,
2019. [Online]. Available: https://www.fao.org/3/a-i7959e.pdf 2017.
[33] M. Hoogeveen, Y. Ono, S. Pfister, A.-M. Boulay, M. Berger, K. Nansai, [54] Great Famine, FAMINE, IRELAND [1845–1849]. Accessed: Sep.
K. Tahara, N. Itsubo, and A. Inaba, ‘‘Consistent characterisation factors at 6, 2019. [Online]. Available: https://www.britannica.com/event/Great-
midpoint and endpoint relevant to agricultural water scarcity arising from Famine-Irish-history
freshwater consumption,’’ Int. J Life Cycle Assessment, vol. 23, no. 12, [55] H. A. Bruns, ‘‘Southern corn leaf blight: A story worth retelling,’’ Rev.
pp. 2276–2287, 2018. doi: 10.1007/s11367-014-0811-5. Interpretation, vol. 109, no. 4, pp. 1218–1224, May 2017.
[34] K. V. de Oliveira, H. M. E. Castelli, S. J. Montebeller, and [56] Keeping Plant Pests and Diseases at Bay: Experts Focus on
T. G. P. Avancini, ‘‘Wireless sensor network for smart agriculture using Global Measures. Accessed: Apr. 13, 2019. [Online]. Available:
ZigBee protocol,’’ in Proc. IEEE 1st Summer School Smart Cities (S3C), http://www.fao.org/news/story/en/item/280489/icode/
Natal, Brazil, Aug. 2017, pp. 61–66. [57] R. P. Pohanish, Sittig’s Handbook of Pesticides and Agricultural Chemi-
[35] Irrigation & Water Use. Accessed: Sep. 3, 2019. [Online]. Available: cals, 2nd ed. 2015.
https://www.ers.usda.gov/topics/farm-practices-management/irrigation- [58] F. P. Carvalho, ‘‘Pesticides, environment, and food safety,’’ Food Energy
water-use/ Secur., vol. 6, no. 2, pp. 48–60, Jun. 2017.
[59] R. Waskom, T. Bauder, R. Pearson, ‘‘Best management practices for
[36] The UN Decade for Deserts and the Fight Against Desertification: Impact
agricultural pesticide use,’’ Bulletin #XCM-177, May 2017.
and Role of Dry Lands. [Online]. Available: https://www.unccd.int/un- [60] S. Kim, M. Lee, and C. Shin, ‘‘IoT-based strawberry disease prediction
decade-deserts-and-fight-against-desertification-impact-and-role- system for smart farming,’’ Sensors, vol. 18, no. 11, p. 4051, 2018.
drylands [61] R. Venkatesan, G. Kathrine, W. Jaspher, and K. Ramalakshmi, ‘‘Internet
[37] J. LaRue and C. Fredrick, ‘‘Decision process for the application of of Things based pest management using natural pesticides for small
variable rate irrigation,’’ Amer. Soc. Agricult. Biol. Eng., Dallas, TX, scale organic gardens,’’ J. Comput. Theor. Nanosci., vol. 15, nos. 9–10,
USA, Tech. Rep., Jul./Aug. 2012. pp. 2742–2747, Sep. 2018.
[38] H. Kiiski, H. Dittmar, M. Drach, R. Vosskamp, M. E. Trenkel, R. Gutser, [62] Semios Integrated Pest Management. Accessed: Jun. 23, 2019. [Online].
and G. Steffens, ‘‘Fertilizers, 2. types,’’ in Ullmann’s Encyclopedia of Available: http://semios.com/ipm/
Industrial Chemistry. 2009. [63] Spensa Z-Trap. Accessed: Jul. 13, 2018. [Online]. Available:
[39] FAO. Forests and Agriculture: Land-Use Challenges and Opportuni- http://spensatech.com/
ties. Accessed: Apr. 15, 2019. [Online]. Available: www.fao.org/3/a- [64] R. Oberti, M. Marchi, P. Tirelli, A. Calcante, M. Iriti, E. Tona,
i5588e.pdf M. Hočevar, J. Baur, J. Pfaff, C. Schütz, and H. Ulbrich, ‘‘Selective
[40] Why IOT is Reinventing Plant Fertilization. Accessed: Sep. 6, 2019. spraying of grapevines for disease control using a modular agricultural
[Online]. Available: https://www.iof2020.eu/latest/news/2017/09/why- robot,’’ Biosyst. Eng., vol. 146, pp. 203–215, Jun. 2016.
the-internet-of-things-is-reinventing-plant-fertilization [65] Potential Effects of Climate Change on Crop Pollination—FAO.
[41] G. Lavanya, C. Rani, and P. Ganeshkumar, ‘‘An automated low cost Accessed: Apr. 13, 2019. [Online]. Available: http://www.fao.org/
IoT based Fertilizer Intimation System for smart agriculture,’’ Sustain. fileadmin/templates/agphome/documents/Biodiversity-
Comput., Inform. Syst., to be published. pollination/Climate_Pollination_17_web___2_.pdf
[66] K. Stein, D. Coulibaly, K. Stenchly, D. Goetze, S. Porembski, A. Lindner,
[42] P. Benincasa, S. Antognelli, L. Brunetti, C. Fabbri, A. Natale, and
S. Konaté, and E. K. Linsenmair, ‘‘Bee pollination increases yield quan-
V. V. M. Sartoretti, ‘‘Reliability of NDVI derived by high resolution
tity and quality of cash crops in Burkina Faso, West Africa,’’ Sci. Rep.,
satellite and UAV compared To in-field methods for the evaluation of
vol. 7, Dec. 2017, Art. no. 17691.
early crop N status and grain yield in wheat,’’ Exp. Agricult., vol. 54,
[67] A. Wietzke, C. Westphal, P. Gras, M. Kraft, K. Pfohl, P. Karlovsky,
no. 4, pp. 604–622, 2018.
E. Pawelzik, T. Tscharntke, and I. Smit, ‘‘Insect pollination as a key
[43] H. Liu, X. Wang, and J. Bing-Kun, ‘‘Study on NDVI optimization of corn factor for strawberry physiology and marketable fruit quality,’’ Agricult.,
variable fertilizer applicator,’’ Agricult. Eng., vol. 56, no. 3, pp. 193–202, Ecosyst. Environ., vol. 258, pp. 197–204, Apr. 2018.
Sep./Dec. 2018. [68] S.-O. Chung, M.-C. Choi, K.-H. Lee, Y.-J. Kim, S.-J. Hong, and M. Li,
[44] J. Shi, X. Yuan, Y. Cai, and G. Wang, ‘‘GPS real-time precise point ‘‘Sensing technologies for grain crop yield monitoring systems: A
positioning for aerial triangulation,’’ GPS Solutions, vol. 21, pp. 405–414, review,’’ J. Biosyst. Eng., vol. 41, no. 4, pp. 408–417, 2016.
Apr. 2017. doi: 10.1007/s10291-016-0532-2. [69] S. Gholami, Z. K. Pishva, G. H. Talaei, and M. A. Dehaghi, ‘‘Effects of
[45] S. Suradhaniwar, S. Kar, R. Nandan, R. Raj, and A. Jagarlapudi, biological and chemical fertilizers nitrogen on yield and yield compo-
‘‘Geo-ICDTs: Principles and applications in agriculture,’’ in Geospatial nents in cumin (Cuminum cyminum L.),’’ J. Chem. Health Risks, vol. 4,
Technologies in Land Resources Mapping, Monitoring and Management, no. 2, pp. 55–64, 2014.
vol. 21, G. Reddy and S. Singh, Eds. Cham, Switzerland: Springer, 2018. [70] Accessed: Apr. 15, 2019. [Online]. Available: https://www.farmtrx.com/

VOLUME 7, 2019 129577

 M. Ayaz et al.: IoT-Based Smart Agriculture: Toward Making the Fields Talk

[71] L. Manfrini, E. Pierpaoli, M. Zibordi, B. Morandi, E. Muzzi, P. Losciale, [97] Global Smart Farming Market to Reach $23.14 Billion by 2022.
and L. C. Grappadelli, ‘‘Monitoring strategies for precise production of Accessed: Apr. 25, 2019. [Online]. Available: https://www.
high quality fruit and yield in Apple in Emilia–Romagna,’’ Chem. Eng. globenewswire.com/news-release/2018/08/02/1546021/0/en/Global-
Trans., vol. 44, pp. 301–306, Sep. 2015. Smart-Farming-Market-to-Reach-23-14-Billion-by-2022.html
[72] N. Torbick, D. Chowdhury, W. Salas, and J. Qi, ‘‘Monitoring rice agricul- [98] Q. Kong, H. Chen, Y. L. Mo, and G. Song, ‘‘Real-time monitoring of
ture across Myanmar using time series Sentinel-1 assisted by Landsat-8 water content in sandy soil using shear mode piezoceramic transducers
and PALSAR-2,’’ Remote Sens., vol. 9, no. 2, p. 119, 2017. and active sensing—A feasibility study,’’ Sensors, vol. 17, no. 10, p. 2395,
[73] Z. Wang, K. B. Walsh, and B. Verma, ‘‘On-tree mango fruit size estima- 2017.
tion using RGB-D images,’’ Sensors, vol. 17, no. 12, p. 2738, 2017. [99] N. Srivastava, G. Chopra, P. Jain, and B. Khatter, ‘‘Pest monitor and
[74] P. Udomkun, M. Nagle, D. Argyropoulos, B. Mahayothee, and J. Müller, control system using wireless sensor network (with special reference to
‘‘Multi-sensor approach to improve optical monitoring of papaya shrink- acoustic device wireless sensor),’’ in Proc. 27th Int. Conf. Elect. Electron.
age during drying,’’ J. Food Eng., vol. 189, pp. 82–89, Nov. 2016. Eng., Goa, India, Jan. 2013, pp. 1–7.
[75] C. Zhang, R. Wohlhueter, and H. Zhang, ‘‘Genetically modified foods: [100] V. Gasso-Tortajada, A. J. Ward, H. Mansur, T. Brøchner, C. G. Sørensen,
A critical review of their promise and problems,’’ Food Sci. Hum. Well- and O. Green, ‘‘A novel acoustic sensor approach to classify seeds based
ness, vol. 5, no. 3, pp. 116–123, 2016. on sound absorption spectra,’’ Sensors, vol. 10, no. 11, pp. 10027–10039,
[76] National Academies of Sciences, Engineering, and Medicine, Genetically 2010.
Engineered Crops: Experiences and Prospects. Washington, DC, USA: [101] J. R. Millan-Almaraz, R. de Jesus Romero-Troncoso,
National Academies Press, 2016. R. G. Guevara-Gonzalez, L. M. Contreras-Medina,
[77] M. Woods, Glass Houses: History of Greenhouses, Conservatories and R. V. Carrillo-Serrano, R. A. Osornio-Rios, C. Duarte-Galvan,
Orangeries. London, U.K.: Aurum Press, 1988. M. A. Rios-Alcaraz, and I. Torres-Pacheco, ‘‘FPGA-based fused
[78] R. R. Shamshiri, F. Kalantari, K. C. Ting, K. R. Thorp, I. A. Hameed, smart sensor for real-time plant-transpiration dynamic estimation,’’
C. Weltzien, D. Ahmad, and Z. M. Shad, ‘‘Advances in greenhouse Sensors, vol. 10, no. 9, pp. 8316–8331, 2010.
automation and controlled environment agriculture: A transition to plant [102] M. I. Husni, M. K. Hussein, M. S. B. Zainal, A. Hamzah, D. M. Nor,
factories and urban agriculture,’’ Int. J. Agricult. Biol. Eng., vol. 11, no. 1, and H. Poad, ‘‘Soil moisture monitoring using field programmable gate
pp. 1–22, 2018. array,’’ Indonesian J. Elect. Eng. Comput. Sci., vol. 11, no. 1, pp. 169–174,
[79] M. A. Akkaş and R. Sokullu, ‘‘An IoT-based greenhouse monitor- Jul. 2018.
ing system with Micaz motes,’’ Procedia Comput. Sci., vol. 113, [103] A. De la Piedra, A. Braeken, and A. Touhafi, ‘‘Sensor systems based
pp. 603–608, 2017. on FPGAs and their applications: A survey,’’ Sensors, vol. 12, no. 9,
[80] D. Cameron, C. Osborne, P. Horton, and M. Sinclair, ‘‘A sustain- pp. 12235–12264, 2012.
able model for intensive agriculture,’’ Univ. Sheffield, Sheffield, U.K., [104] S. C. Murray, ‘‘Optical sensors advancing precision in agricultural pro-
Tech. Rep., Dec. 2015. duction,’’ Photon. Spectra, vol. 51, no. 6, p. 48, Jun. 2018.
[81] [Online]. Available: https://www.sciencealert.com/the-world-has-lost-a- [105] F. P. Povh and W. de Paula Gusmão dos Anjos, ‘‘Optical sensors applied
third-of-its-farmable-land-in-the-last-40-years in agricultural crops,’’ Opt. Sensors-New Develop. Practical Appl., 2014.
[82] D. Pimentel and M. Burgess, ‘‘Soil erosion threatens food production,’’ [106] G. Pajares, ‘‘Advances in sensors applied to agriculture and forestry,’’
Agriculture, vol. 3, no. 3, pp. 443–463, 2013. Sensors, vol. 11, no. 9, pp. 8930–8932, 2011.
[83] K. Benke and B. Tomkins, ‘‘Future food-production systems: Vertical [107] I. Molina, C. Morillo, E. García-Meléndez, R Guadalupe, and
farming and controlled-environment agriculture,’’ Sustainability, Sci., M. I. Roman, ‘‘Characterizing olive grove canopies by means of ground-
Pract. Policy, vol. 13, no. 1, pp. 13–26, 2017. based hemispherical photography and spaceborne RADAR data,’’ Sen-
[84] Accessed: Apr. 25, 2019. [Online]. Available: https://aerofarms.com/
sors, vol. 11, no. 8, pp. 7476–7501, 2011.
[85] Maintaining Optimum Growth Conditions in Vertical Farms.
[108] J. S. Dvorak, M. L. Stone, and K. P. Self, ‘‘Object detection for agri-
Accessed: Apr. 15, 2019. [Online]. Available: https://www.azosensors.
cultural and construction environments using an ultrasonic sensor,’’
com/article.aspx?ArticleID=1173
J. Agricult. Saf. Health, vol. 22, no. 2, pp. 107–119, 2016.
[86] Vertical Farming and the IoT. Accessed: Apr. 15, 2019. [Online]. Avail-
[109] T. G. Álvarez-Arenas, E. Gil-Pelegrin, J. E. Cuello, M. D. Fariñas,
able: http://www.mintcontrols.com/vertical-farming-iot/
D. Sancho-Knapik, D. A. C. Burbano, and J. J. Peguero-Pina, ‘‘Ultrasonic
[87] Accessed: Jul. 27, 2019. [Online]. Available: https://www.veggitech.com/
sensing of plant water needs for agriculture,’’ Sensors, vol. 16, no. 7,
[88] Smart Wireless Sensor for Hydroponics Monitoring. Accessed:
p. 1089, Jul. 2016.
Jun. 18, 2019. [Online]. Available: https://www.ipi-singapore.org/
[110] G. Pajares, A. Peruzzi, and P. Gonzalez-de-Santos, ‘‘Sensors in agricul-
technology-offers/smart-wireless-sensor-hydroponics-monitoring
ture and forestry,’’ Sensors, vol. 13, no. 9, Sep. 2013, Art. no. 12132.
[89] Hydroponics Water Quality. Accessed: Apr. 25, 2019. [Online]. Avail-
[111] D. Andújar, R. Ribeiro, C. Fernández-Quintanilla, and J. Dorado, ‘‘Accu-
able: https://sensorex.com/hydroponics/
racy and feasibility of optoelectronic sensors for weed mapping in wide
[90] H. Ibayashi, Y. Kaneda, J. Imahara, N. Oishi, M. Kuroda, and H. Mineno, row crops,’’ Sensors, vol. 11, no. 3, pp. 2304–2318, 2011.
‘‘A reliable wireless control system for tomato hydroponics,’’ Sensors, [112] D. AndÞjar, A. Ribeiro, C. F. Quintanilla, J. Dorado, and J. Dorado,
vol. 16, no. 5, p. 644, 2016. ‘‘Assessment of a ground-based weed mapping system in maize,’’ Preci-
[91] A. Theopoulos, A. Boursianis, A. Koukounaras, and T. Samaras, ‘‘Pro- sion Agriculture. 2009.
totype wireless sensor network for real-time measurements in hydropon- [113] F. J. García-Ramos, M. Vidal, A. Boné, H. Malón, and J. Aguirre, ‘‘Anal-
ics cultivation,’’ in Proc. 7th Int. Conf. Modern Circuits Syst. Technol. ysis of the air flow generated by an air-assisted sprayer equipped with
(MOCAST), Thessaloniki, Greece, May 2018, pp. 1–4. two axial fans using a 3D sonic anemometer,’’ Sensors, vol. 12, no. 6,
[92] T. Nishimura, Y. Okuyama, A. Matsushita, H. Ikeda, and A. Satoh, pp. 7598–7613, 2012.
‘‘A compact hardware design of a sensor module for hydroponics,’’ in [114] T. K. Yew, Y. Yusoff, L. K. Sieng, H. C. Lah, H. Majid, and N. Shelida,
Proc. IEEE 6th Global Conf. Consumer Electron. (GCCE), Nagoya, ‘‘An electrochemical sensor ASIC for agriculture applications,’’ in
Japan, Oct. 2017, pp. 1–4. Proc. 37th Int. Conv. Inf. Commun. Technol., Electron. Microelectron.
[93] P. Tripodi, D. Massa, A. Venezia, and T. Cardi, ‘‘Sensing technologies (MIPRO), Opatija, Croatia, May 2014, pp. 85–90.
for precision phenotyping in vegetable crops: Current status and future [115] D. J. Cocovi-Solberg, M. Rosende, and M. Miró, ‘‘Automatic kinetic
challenges,’’ Agronomy, vol. 8, no. 4, p. 57, 2018. bioaccessibility assay of lead in soil environments using flow-through
[94] Y. Rouphael, L. Spíchal, K. Panzarová, R. Casa, and G. Colla, ‘‘High- microdialysis as a front end to electrothermal atomic absorption spec-
throughput plant phenotyping for developing novel biostimulants: From trometry,’’ Environ. Sci. Technol., vol. 48, pp. 6282–6290, May 2014.
lab to field or from field to lab?’’ Frontiers Plant Sci., vol. 9, p. 1197, [116] [Online]. Available: http://blog.agrivi.com/post/smart-sensors-for-
Aug. 2018. accurate-soil-measurements
[95] J. Zhou, D. Reynolds, D. Websdale, T. Le Cornu, O. Gonzalez-Navarro, [117] M. A. M. Yunus and S. C. Mukhopadhyay, ‘‘Novel planar electromag-
C. Lister, S. Orford, S. Laycock, G. Finlayson, T. Stitt, M. D. Clark, netic sensors for detection of nitrates and contamination in natural water
M. W. Bevan, and S. Griffiths, ‘‘Cropquant: An automated and scalable sources,’’ IEEE Sensors J., vol. 11, no. 6, pp. 1440–1447, Jun. 2011.
field phenotyping platform for crop monitoring and trait measurements [118] A. Hemmat, A. R. Binandeh, J. Ghaisari, and A. Khorsandi, ‘‘Devel-
to facilitate breeding and digital agriculture,’’ bioRxiv, Tech. Rep., 2017. opment and field testing of an integrated sensor for on-the-go measure-
[96] Accessed: Apr. 15, 2019. [Online]. Available: https://www.plant- ment of soil mechanical resistance,’’ Sens. Actuators A, Phys., vol. 198,
phenotyping-network.eu/ pp. 61–68, Aug. 2013.

129578 VOLUME 7, 2019

M. Ayaz et al.: IoT-Based Smart Agriculture: Toward Making the Fields Talk

[119] J. N. Schuster, M. J. Darr, and R. P. McNaull, ‘‘Performance benchmark [144] Accessed: Jun. 18, 2018. [Online]. Available: http://www.sol-chip.com/
of yield monitors for mechanical and environmental influences,’’ in Proc. ComSCC.asp
Agricult. Biosyst. Eng. Conf. Presentations, Jul. 2017, pp. 1–17. [145] Accessed: Jan. 18, 2019. [Online]. Available: https://www.ntmsensors.
[120] Managing Calibration Curves under John Deere Tractors in Yield Moni- com/hydrogen-sensors/
tor Systems, IOWA State University, Ames, IA, USA. [146] Accessed: Apr. 15, 2019. [Online]. Available: http://www.dynamax.com/
[121] C. Moureaux, E. Ceschia, N. Arriga, P. Béziat, W. Eugster, W. L. Kutsch, products/plant-growth-sensors/dex-fruit-stem-growth-dendrometer
and E. Pattey, ‘‘Eddy covariance measurements over crops,’’ in Eddy [147] Accessed: Apr. 15, 2019. [Online]. Available: https://gpsgate.com/
Covariance: A Practical Guide to Measurement and Data Analysis, devices/piccolo_atx
M. Aubinet, T. Vesala, and D. Papale, Eds. Dordrecht, The Netherlands: [148] Accessed: Apr. 15, 2019. [Online]. Accessed: Jun. 18, 2018. [Online].
Springer, 2012. Available: https://www.cid-inc.com/plant-science-tools/photosynthesis-
[122] A. Kumar, A. Bhatia, R. K. Fagodiya, S. K. Malyan, and B. L. Meena, measurement/ci-340-handheldphotosynthesis-system/
‘‘Eddy covariance flux tower: A promising technique for green- [149] Accessed: Jan. 18, 2019. [Online]. Available: https://www.campbellsci.
house gases measurement,’’ Adv. Plants Agricult. Res., vol. 7, no. 4, com/03002-wind-sentry
pp. 337–340, 2017. [150] Accessed: Jan. 18, 2019. [Online]. Available: https://www.aeroqual.com/
[123] A. Crabit, F. Colin, J. S. Bailly, H. Ayroles, and F. Garnier, ‘‘Soft water aqm-65-air-monitoring-station
level sensors for characterizing the hydrological behaviour of agricultural [151] Accessed: Jan. 18, 2019. [Online]. Available: http://www.stevenswater.
catchments,’’ Sensors, vol. 11, no. 5, pp. 4656–4673, 2011. com/products/sensors/soil/pogo/
[124] B. Howard, ‘‘LIDAR and its use in agriculture,’’ Agricult. Innov., to be [152] Accessed: Apr. 15, 2019. [Online]. Available: https://www.edaphic.
published. com.au/products/sap-flow-sensors/small-stems-petioles-flower-fruit-
[125] U. Weiss and P. Biber, ‘‘Plant detection and mapping for agricultural sap-flow-sensors/
[153] Accessed: Feb. 18, 2019. [Online]. Available: https://www.ecotech.
robots using a 3D LIDAR sensor,’’ Robot. Auton. Syst., vol. 59, no. 5,
com/product/meteorology-and-hvac/meteorological-sensors/metonemet-
pp. 265–273, 2011.
[126] P. Biber, U. Weiss, M. Dorna, and A. Albert. (2012). Navigation System station-one
[154] Accessed: Apr. 15, 2019. [Online]. Available: http://phyto-sensor.
of the Autonomous Agricultural Robot—‘BoniRob’. [Online]. Available:
com/SD-5P-SD-6P
http://www.cs.cmu.edu/~mbergerm/agrobotics2012/01Biber.pdf
[155] Accessed: Apr. 15, 2019. [Online]. Available: http://www.back2me.
[127] I. Del-Moral-Martínez, J. R. Rosell-Polo, J. Company, R. Sanz, A. Escolà,
com/index.cfm?action=product&pid=14
J. Masip, J. A. Martínez-Casasnovas, and J. Arnó, ‘‘Mapping vineyard
[156] Accessed: Apr. 15, 2019. [Online]. Available: https://www.
leaf area using mobile terrestrial laser scanners: Should rows be scanned
topconpositioning.com/insights/topcon-introduces-yieldtrakk-yield-
on-the-go or discontinuously sampled?’’ Sensors, vol. 16, no. 1, p. 119,
monitor-system
Jan. 2016. [157] Hello Tractor. The Uber for Tractors. Accessed: Apr. 25, 2019. [Online].
[128] A. Montagnoli, S. Fusco, M. Terzaghi, A. Kirschbaum, D. Pflugmacher, Available: http://impactjournalismday.com/story/hello-tractor/
W. B. Cohen, G. S. Scippa, and D. Chiatante, ‘‘Estimating forest above- [158] Agricultural Robots and Drones 2018–2038: Technologies, Markets
ground biomass by low density lidar data in mixed broad-leaved forests and Players. Accessed: Apr. 25, 2019. [Online]. Available:
in the Italian Pre-Alps,’’ Forest Ecosyst., vol. 2, no. 1, 2015, Art. no. 10. https://www.idtechex.com/research/reports/agricultural-robots-
[129] T. Mark and T. Griffin, ‘‘Defining the barriers to telematics for precision and-drones-2018-2038-technologies-markets-and-players-000578.asp
agriculture: Connectivity supply and demand,’’ presented at the SAEA [159] Agricultural Robots and Drones to Become a 45 Billion Dollar
Annu. Meeting, Austin, TX, USA, 2016. Industry by 2038. Accessed: Apr. 15, 2019. [Online]. Available:
[130] A. K. E. Mohamed, ‘‘Analysis of telematics systems in agriculture,’’ https://www.gim-international.com/content/news/agricultural-robots-
M.S. thesis, Dept. Mach., Utilization, CULS, Prague, Czech Republic, and-drones-to-become-a-45-billion-dollar-industry-by-2038
2013. [160] J. Oliver. (Aug. 2017). e-Agriculture-Internet of Things (LoT) for
[131] Digital Farming: What Does it Really Mean? And What is the Vision of Agriculture Webinar Series: IoT: The Internet of Tractors. [Online].
Europe’s Farm Machinery Industry for Digital Farming? in European Available: http://www.fao.org/e-agriculture/news/internet-things-lot-
Agricultural Machinery, CEMA, 2017. agriculture-webinar-series-iot-internet-tractors
[132] H. H. Jaafar and E. Woertz, ‘‘Agriculture as a funding source of ISIS: [161] Accessed: Apr. 21, 2019. [Online]. Available: https://www.caseih.com/
A GIS and remote sensing analysis,’’ Food Policy, vol. 64, pp. 14–25, apac/en-int/products/tractors/Magnum-Series
Oct. 2016. [162] IoT Tech Enables World’s First Connected Tractor. Accessed:
[133] S. G. Yalew, A. van Griensven, M. L. Mul, and P. van der Zaag, ‘‘Land Apr. 11, 2019. [Online]. Available: https://www.eenewseurope.com/
suitability analysis for agriculture in the Abbay basin using remote sens- news/iot-tech-enables-worlds-first-connected-tractor
ing, GIS and AHP techniques,’’ Model. Earth Syst. Environ. vol. 2, p. 101, [163] John Deere—Bringing AI to Agriculture. Accessed: Apr. 24, 2019.
Jun. 2016. [Online]. Available: https://rctom.hbs.org/submission/john-deere-
[134] I. R. Hegazy and M. R. Kaloop, ‘‘Monitoring urban growth and land use bringing-ai-to-agriculture/
change detection with GIS and remote sensing techniques in Daqahlia [164] S. G. Bronars, ‘‘A vanishing breed: How the decline in U.S. Farm laborers
governorate Egypt,’’ Int. J. Sustain. Built Environ., vol. 4, no. 1, over the last decade has hurt the U.S. economy and slowed production on
pp. 117–124, 2015, American farms,’’ New American Economy, Tech. Rep., Jul. 2015.
[135] I. Rose and M. Welsh, ‘‘Mapping the urban wireless landscape with [165] Farm Labor. Accessed: Apr. 20, 2019. [Online]. Available:
Argos,’’ in Proc. 8th ACM Conf. Embedded Netw. Sensor Syst. (SenSys), https://www.ers.usda.gov/topics/farm-economy/farm-labor/
New York, NY, USA, 2010, pp. 323–336. [166] Y. Zhao, L. Gong, Y. Huang, and C. Liu, ‘‘A review of key techniques of
[136] R. Patmasari, I. Wijayanto, R. S. Deanto, Y. P. Gautama, and vision-based control for harvesting robot,’’ Comput. Electron. Agricult.,
H. Vidyaningtyas, ‘‘Design and realization of automatic packet reporting vol. 127, pp. 311–323, Sep. 2016.
system (APRS) for sending telemetry data in Nano satellite communica- [167] A. Zujevs, V. Osadcuks, and P. Ahrendt, ‘‘Trends in robotic sensor tech-
tion system,’’ J. Meas., Electron., Commun., Syst., vol. 4, no. 1, pp. 1–7, nologies for fruit harvesting: 2010-2015,’’ Procedia Comput. Sci., vol. 77,
Jun. 2018. pp. 227–233, Jan. 2015.
[137] Accessed: Apr. 25, 2019. [Online]. Available: http://loupelectronics. [168] S. Bargoti and J. Underwood, ‘‘Deep fruit detection in orchards,’’ in Proc.
com/products/yield_monitor.html IEEE Int. Conf. Robot. Automat. (ICRA), Singapore, May/Jun. 2017,
[138] Accessed: Apr. 15, 2019. [Online]. Available: http://www.icstation.com/ pp. 3626–3633.
m214-soil-moisture-sensor-humidity-controller-module-99rh-auto [169] C. W. Bac, J. Hemming, B. A. J. van Tuijl, R. Barth, E. Wais, and
matic-control-irrigation-system-digital-display-controller-p-13099.html E. J. van Henten, ‘‘Performance evaluation of a harvesting robot for sweet
[139] Accessed: Apr. 25, 2019. [Online]. Available: https://agriculture.trimble. pepper,’’ J. Field Robot., vol. 34, no. 6, pp. 1123–1139, Sep. 2017.
com/product/ag-premium-weather/ [170] J. Feng, L. Zeng, and L. He, ‘‘Apple fruit recognition algorithm based on
[140] Accessed: Apr. 15, 2019. [Online]. Available: http://phyto-sensor.com/FI- multi-spectral dynamic image analysis,’’ Sensors, vol. 19, no. 4, p. 949,
LM-FI-MM-FI-SM Feb. 2019.
[141] Accessed: Jan. 17, 2019. [Online]. Available: https://www.pycno.co.uk/ [171] Accessed: Apr. 15, 2019. [Online]. Available: https://www.pepperl-
[142] Accessed: Jan. 18, 2019. [Online]. Available: http://www.ictinternational. fuchs.com/global/en/27566.htm
com/products/mp406/mp406-moisture-sensor/ [172] Accessed: Apr. 15, 2019. [Online]. Available: https://www.
[143] [Online]. Available: https://www.deere.com/en/technology-products/ fastcompany.com/40473583/this-strawberry-picking-robot-gently-picks-
precision-ag-technology/guidance/greenstar-3-2630/ the-ripest-berries-with-its-robo-hand

VOLUME 7, 2019 129579

 M. Ayaz et al.: IoT-Based Smart Agriculture: Toward Making the Fields Talk

[173] Accessed: Apr. 15, 2019. [Online]. Available: http://www.sweeper-robot. [198] M. J. Koyenikan and A. Ighoro, ‘‘Farmers’ use of mobile phone-based
eu/11-news/48-sweeper-demonstrated-its-harvesting-robot-for-the-first- services for accessing agriculture and rural development information
time in northern zone of Edo State, Nigeria,’’ Nigerian J. Rural Sociology,
[174] Accessed: Apr. 15, 2019. [Online]. Available: https://www.ffrobotics vol. 16, no. 2, pp. 23–28, 2015.
.com/ [199] F. Sousa, G. Nicolay, and R. Home, ‘‘Video on mobile phones as an
[175] S. G. Defterli, Y. Shi, Y. Xu, and R. Ehsani, ‘‘Review of robotic tech- effective way to promote sustainable practices by facilitating innovation
nology for strawberry production,’’ Appl. Eng. Agricult., vol. 32, no. 3, uptake in mali,’’ Int. J. Sustain. Develop. Res., vol. 5, no. 1, pp. 1–8,
pp. 301–318, 2016. Mar. 2019.
[176] M. Bolda. Robotic Strawberry Harvester on the Move. Accessed: [200] G. V. Nakato, F. Beed, H. Bouwmeester, I. Ramathani, S. Mpiira,
Sep. 6, 2019. [Online]. Available: http://ucanr.edu/blogs/ J. Kubiriba, and S. Nanavati, ‘‘Building agricultural networks of farmers
strawberries_caneberries/index.cfm?tagname=Agrobot and scientists via mobile phones: Case study of banana disease surveil-
[177] [Online]. Available: https://www.freshplaza.com/article/2104936/spain- lance in Uganda,’’ Can. J. Plant Pathol., vol. 38, no. 3, pp. 307–316, 2016.
sw-6010-advanced-technology-for-the-strawberry-harvest/ [201] B. Masuka, T. Matenda, J. Chipomho, N. Mapope, S. Mupeti, S. Tatsvarei,
[178] Tektu T100 Strawberry Harvester, Tech2reality. [Online]. Available: and W. Ngezimana, ‘‘Mobile phone use by small-scale farmers: A poten-
https://www.hortweek.com/technological-innovations-harvesting/fresh- tial to transform production and marketing in Zimbabwe,’’ South Afr. J.
produce/article/1001333 Agricult. Extension, vol. 44, no. 2, pp. 121–135, 2016.
[179] Beecham Research. (2016). An Introduction to LPWA Public Service [202] S. Musungwini, ‘‘Mobile phone use by zimbabwean smallholder farmers:
Categories: Matching Services to IoT Applications. [Online]. Available: A baseline study,’’ Afr. J. Inf. Commun., vol. 22, pp. 29–52, 2018.
http://www.beechamresearch.com/download.aspx?id=1049 [203] J. R. M. Nzie, J. C. Bidogeza, and N. A. Ngum, ‘‘Mobile phone use,
[180] C. Z. Zulkifli and N. N. Noor, ‘‘Wireless sensor network and Internet of transaction costs, and price: Evidence from rural vegetable farmers in
Things (IoT) solution in agriculture,’’ Pertanika J. Sci. Technol., vol. 25, Cameroon,’’ J. Afr. Bus., vol. 19, pp. 323–342, Nov. 2017.
no. 1, pp. 91–100, Jan. 2017. [204] J.-P. Qian, X.-T. Yang, X.-M. Wu, B. Xing, B.-G. Wu, and M. Li, ‘‘Farm
[181] G.-Z. Hong and C.-L. Hsieh, ‘‘Application of integrated control strat- and environment information bidirectional acquisition system with indi-
egy and Bluetooth for irrigating romaine lettuce in greenhouse,’’ IFAC- vidual tree identification using smartphones for orchard precision man-
PapersOnLine, vol. 49, no. 16, pp. 381–386, 2016. agement,’’ Comput. Electron. Agricult., vol. 116, pp. 101–108, Aug. 2015.
[182] B. Jonathan, ‘‘Evaluation of Bluetooth low energy in agriculture envi- [205] Q. Yu, Y. Shi, H. Tang, P. Yang, A. Xie, B. Liu, and W. Wu, ‘‘eFarm:
ronments: An empirical analysis of BLE in precision agriculture,’’ A tool for better observing agricultural land systems,’’ Sensors, vol. 17,
M.S. thesis, Malmö Högskola Univ., Malmö, Sweden, 2016. no. 3, p. 453, 2017.
[183] D. Taşkin, C. Taskin, and S. Yazar, ‘‘Developing a Bluetooth low energy [206] M. F. Işık, Y. Sönmez, C. Yılmaz, V. Özdemir, and E. N. Yılmaz, ‘‘Preci-
sensor node for greenhouse in precision agriculture as Internet of Things sion irrigation system (PIS) using sensor network technology integrated
application,’’ Adv. Sci. Technol. Res. J., vol. 12, no. 4, pp. 88–96, 2018. with IOS/Android application,’’ Appl. Sci., vol. 7, no. 9, p. 891, 2017.
[184] G. R. Mendez, M. A. M. Yunus, and S. C. Mukhopadhyay, ‘‘A WiFi based [207] E. Guler, T. Y. Sengel, Z. P. Gumus, M. Arslan, H. Coskunol, S. Timur,
smart wireless sensor network for an agricultural environment,’’ in Proc. and Y. Yagci, ‘‘Mobile phone sensing of Cocaine in a lateral flow assay
5th Int. Conf. Sens. Technol., Nov./Dec. 2011, pp. 405- -410. combined with a biomimetic material,’’ Anal. Chem., vol. 89, no. 18,
[185] J. Petäjäjärvi, K. Mikhaylov, M. Hämäläinen, and J. Iinatti, ‘‘Evaluation pp. 9629–9632, 2017.
of LoRa LPWAN technology for remote health and wellbeing monitor- [208] X. Fu and S. Akter, ‘‘The impact of mobile phone technology on agri-
ing,’’ in Proc. 10th Int. Symp. Med. Inf. Commun. Technol. (ISMICT), cultural extension services delivery: Evidence from India,’’ J. Develop.
Worcester, MA, USA, Mar. 2016, pp. 1–5. Stud., vol. 52, no. 11, pp. 1561–1576, 2016.
[186] R. Jedermann, M. Borysov, N. Hartgenbusch, S. Jaeger, M. Sellwig, and [209] A. Das, D. Basu, and R. Goswami, ‘‘Accessing agricultural information
W. Lang, ‘‘Testing Lora for food applications—Example application for through mobile phone: Lessons of IKSL services in West Bengal,’’ Indian
airflow measurements inside cooled warehouses with apples,’’ Procedia Res. J. Extension Educ., vol. 12, no. 3, pp. 102–107, 2012.
Manuf., vol. 24, Jan. 2018, pp. 284–289. [210] H. Baumüller, ‘‘The little we know: An exploratory literature review on
[187] N. Zhu, Y. Xia, Y. Liu, C. Zang, H. Deng, and Z. Ma, ‘‘Temperature the utility of mobile phone-enabled services for smallholder farmers,’’
and humidity monitoring system for bulk grain container based on Lora J. Int. Develop., vol. 30, no. 1, pp. 134–154, 2018.
wireless technology,’’ in Proc. ICCCS, 2018, pp. 102–110. [211] S. Chung, L. E. Breshears, and J.-Y. Yoon, ‘‘Smartphone near infrared
[188] A. Lavric, A. I. Petrariu, and V. Popa, ‘‘Long range SigFox communica- monitoring of plant stress,’’ Comput. Electron. Agricult., vol. 154,
tion protocol scalability analysis under large-scale, high-density condi- pp. 93–98, Nov. 2018.
tions,’’ IEEE Access, vol. 7, pp. 35816–35825, 2019. [212] A. J. S. McGonigle, T. C. Wilkes, T. D. Pering, J. R. Willmott, J. M. Cook,
[189] H. Metcalfe, Mobile for Development Impact Products and Services F. M. Mims, and A. V. Parisi, ‘‘Smartphone spectrometers,’’ Sensors,
Landscape Annual Review. London, U.K.: GSM Association, 2015. vol. 18, no. 1, p. 223, 2018.
[190] G. Alfian, M. Syafrudin, and J. Rhee, ‘‘Real-time monitoring system [213] N. Moonrungsee, S. Pencharee, and J. Jakmunee, ‘‘Colorimetric analyzer
using smartphone-based sensors and nosql database for perishable supply based on mobile phone camera for determination of available phosphorus
chain,’’ Sustainability, vol. 9, no. 11, p. 2073, Nov. 2017. in soil,’’ Talanta, vol. 136, pp. 204–209, May 2015.
[191] S. Pongnumkul, P. Chaovalit, and N. Surasvadi, ‘‘Applications of [214] A. Camacho and H. Arguello, ‘‘Smartphone-based application for agri-
smartphone-based sensors in agriculture: A systematic review of cultural remote technical assistance and estimation of visible vegeta-
research,’’ J. Sensors, vol. 2015, Jul. 2015, Art. no. 195308. tion index to farmer in Colombia: AgroTIC,’’ Proc. SPIE, vol. 10783,
[192] e-Agriculture 10 Year Review Report, FAO, Rome, Italy, 2015. Oct. 2018, Art. no. 107830K.
[193] S. Wyche and C. Steinfield, ‘‘Why don’t farmers use cell phones to [215] M. Prosdocimi, M. Burguet, S. Di Prima, G. Sofia, E. Tero, J. R. Comino,
access market prices? Technology affordances and barriers to market A. Cerdà, and P. Tarolli, ‘‘Rainfall simulation and Structure-from-Motion
information services adoption in rural Kenya,’’ Inf. Technol. Develop., photogrammetry for the analysis of soil water erosion in Mediterranean
vol. 22, no. 2, pp. 320–333, 2016. vineyards,’’ Sci. Total Environ., vol. 574, pp. 204–215, Jan. 2017.
[194] H. Baumüller, ‘‘Agricultural innovation and service delivery through [216] P. Han, D. Dong, X. Zhao, L. Jiao, and Y. Lang, ‘‘A smartphone-based soil
mobile phones analyses in kenya,’’ Ph.D. dissertation, Fac. Agricult., color sensor: For soil type classification,’’ Comput. Electron. Agricult.,
Univ. Bonn, Bonn, Germany, 2015. vol. 123, pp. 232–241, Apr. 2016.
[195] M. Hidrobo and D. Gilligan, ‘‘Using quantitative methods to evaluate [217] X. Wan, J. Cui, X. Jiang, J. Zhang, Y. Yang, and T. Zheng, ‘‘Smartphone
mobile phone technology based nutrition and agriculture advisory ser- based hemispherical photography for canopy structure measurement,’’
vices in Ghana,’’ Brighton, U.K., Tech. Rep., 2017. Proc. SPIE, vol. 10621, Jan. 2018, Art. no. 106210Q.
[196] I. Barnett, B. Faith, J. Gordon, and C. Sefa-Nyarko, ‘‘External evaluation [218] R. Stiglitz, E. Mikhailova, C. Post, M. Schlautman, J. Sharp,
of mobile phone technology-based nutrition and agriculture advisory R. Pargas, B. Glover, and J. Mooney, ‘‘Soil color sensor data collec-
services in Africa and South Asia,’’ Mobile Phones, Agricult., Nutrition tion using a GPS-enabled smartphone application,’’ Geoderma, vol. 296,
Ghana, Qualitative Midline Study Rep. Brighton, Tech. Rep., 2019. pp. 108–114, Jun. 2017.
[197] E. O. Adio, Y. Abu, S. K. YUsuf, and S. Nansoh, ‘‘Use of agricultural [219] X. Xie, X. Zhang, B. He, D. Liang, D. Zhang, and L. Huang, ‘‘A system
information sources and services by farmers for improve productivity in for diagnosis of wheat leaf diseases based on Android smartphone,’’ Proc.
Kwara state,’’ Library Phil. Pract. (e-J.), 2016. SPIE, vol. 10155, Oct. 2016, Art. no. 1015526.

129580 VOLUME 7, 2019

M. Ayaz et al.: IoT-Based Smart Agriculture: Toward Making the Fields Talk

[220] Z. Kou and C. Wu, ‘‘Smartphone based operating behaviour mod- [239] M. A. Carmona, F. J. Sautua, O. Pérez-Hernández, and J. I. Mandolesi,
elling of agricultural machinery,’’ IFAC-PapersOnLine, vol. 51, no. 17, ‘‘AgroDecisor EFC: First Android app decision support tool for timing
pp. 521–525, 2018. fungicide applications for management of late-season soybean diseases,’’
[221] L. Frommberger, F. Schmid, and C. Cai, ‘‘Micro-mapping with smart- Comput. Electron. Agricult., vol. 144, pp. 310–313, Jan. 2018.
phones for monitoring agricultural development,’’ in Proc. 3rd ACM [240] B. B. Machado, J. P. M. Orue, M. S. Arruda, C. V. Santos,
Symp. Comput. Develop., 2013, Art. no. 46. D. S. Sarath, W. N. Goncalves, G. G. Silva, H. Pistori, A. R. Roel,
[222] O. Debauche, S. Mahmoudi, A. L. H. Andriamandroso, P. Manneback, and J. F. Rodrigues, Jr., ‘‘BioLeaf: A professional mobile application to
J. Bindelle, and F. Lebeau, ‘‘Cloud services integration for farm animals’ measure foliar damage caused by insect herbivory,’’ Comput. Electron.
behavior studies based on smartphones as activity sensors,’’ J. Ambient Agricult., vol. 129, pp. 44–55, Nov. 2016.
Intell. Humaniz. Comput., pp. 1–12, May 2018. [241] A. Pérez-Castro, J. A. Sánchez-Molina, M. Castilla, J. Sánchez-Moreno,
[223] F. Orlando, E. Movedi, D. Coduto, S. Parisi, L. Brancadoro, V. Pagani, J. C. Moreno-Úbeda, and J. J. Magán, ‘‘cFertigUAL: A fertigation man-
T. Guarneri, and R. Confalonieri, ‘‘Estimating leaf area index (LAI) in agement app for greenhouse vegetable crops,’’ Agricult. Water Manage.,
vineyards using the PocketLAI smart-app,’’ Sensors, vol. 16, no. 12, vol. 183, pp. 186–193, Mar. 2017.
p. 2004, 2016. [242] F. Ceballos, B. Kramer, and L. M. Robles, ‘‘The feasibility of picture-
[224] A. Camacho and H. Arguello, ‘‘Smartphone-based application for agri- based insurance (PBI): Smartphone pictures for affordable crop insur-
cultural remote technical assistance and estimation of visible vegeta- ance,’’ SSRN, IFPRI Discuss. Paper 1788, Dec. 2018. [Online]. Available:
tion index to farmer in Colombia: AgroTIC,’’ Proc. SPIE, vol. 10783, https://ssrn.com/abstract=3324536
Oct. 2018, Art. no. 107830K. [243] Cloud Computing Helps Agriculture Industry Grow. Accessed: Sep. 6,
[225] A. L. H. Andriamandroso, F. Lebeau, Y. Beckers, E. Froidmont, 2019. [Online]. Available: https://www.ibm.com/blogs/cloud-computing/
I. Dufrasne, B. Heinesch, P. Dumortier, G. Blanchy, Y. Blaise, and 2015/01/23/cloud-computing-helps-agriculture-industry-grow/
J. Bindelle, ‘‘Development of an open-source algorithm based on iner- [244] Accessed: Apr. 15, 2019. [Online]. Available: https://www.fujitsu.com/
tial measurement units (IMU) of a smartphone to detect cattle grass global/about/resources/news/press-releases/2012/0718-01.html
[245] Accessed: Apr. 25, 2019. [Online]. Available: http://www.sourcetrace.
intake and ruminating behaviors,’’ Comput. Electron. Agricult., vol. 139,
com/apps/
pp. 126–137, Jun. 2017. [246] M. A. Uddin, A. Mansour, D. Le Jeune, M. Ayaz, and E.-H. Aggoune,
[226] M. F. M. Azam, S. H Rosman, M. Mustaffa, S. M. S. Mullisi, H. Wahy,
‘‘UAV-assisted dynamic clustering of wireless sensor networks for crop
M. H. Jusoh, and M. I. Ali, ‘‘Hybrid water pump system for hilly
health monitoring,’’ Sensors, vol. 18, no. 2, p. 555, 2018.
agricultural site,’’ in Proc. 7th IEEE Control Syst. Graduate Res. [247] M. Ruwaimana, B. Satyanarayana, V. Otero, A. M. Muslim, M. S. A,
Colloq. (ICSGRC), Aug. 2016, pp. 109–114. S. Ibrahim, D. Raymaekers, N. Koedam, and F. Dahdouh-Guebas,
[227] J. E. Herrick, A. Beh, E. Barrios, I. Bouvier, M. Coetzee, D. Dent, E. Elias,
‘‘The advantages of using drones over space-borne imagery in the
T. Hengl, J. W. Karl, H. Liniger, J. Matuszak, J. C. Neff, L. W. Ndungu,
mapping of mangrove forests,’’ PLoS ONE, vol. 13, no. 7, 2018,
M. Obersteiner, K. D. Shepherd, K. C. Urama, R. van den Bosch, and
Art. no. e0200288.
N. P. Webb, ‘‘The land-potential knowledge system (LandPKS): Mobile [248] Drones for Agriculture by FAO. Accessed: Apr. 15, 2019. [Online].
apps and collaboration for optimizing climate change investments,’’ Available: www.fao.org/3/i8494en/i8494en.pdf
Ecosyst. Health Sustainability, vol. 2, no. 3, p. e01209, 2016. [249] L. Tang and G. Shao, ‘‘Drone remote sensing for forestry research and
[228] W. Palomino, G. Morales, S. Huamán, and J. Telles, ‘‘PETEFA: Geo- practices,’’ J. Forestry Res., vol. 26, no. 4, pp. 791–797, Dec. 2015.
graphic information system for precision agriculture,’’ in Proc. IEEE [250] Best Drones for Agriculture 2019: The Ultimate Buyer’s Guide.
25th Int. Conf. Electron., Elect. Eng. Comput. (INTERCON), Aug. 2018, Accessed: Apr. 16, 2019. [Online]. Available: https://bestdroneforthejob.
pp. 1–4. com/drone-buying-guides/agriculture-drone-buyers-guide/
[229] A. Sopegno, A. Calvo, R. Berruto, P. Busato, and D. Bocthis, ‘‘A Web [251] Accessed: Mar. 14, 2019. [Online]. Available: https://www.questuav.com/
mobile application for agricultural machinery cost analysis,’’ Comput. drones/datahawk-agriculture/
Electron. Agricult., vol. 130, pp. 158–168, Nov. 2016. [252] Accessed: Apr. 12, 2019. [Online]. Available: https://www.
[230] M. V. Bueno-Delgado, J. M. Molina-Martínez, R. Correoso-Campillo, precisionhawk.com/drones
and P. Pavón-Mariño, ‘‘Ecofert: An Android application for the optimiza- [253] A Fully Autonomous Drone for Daily Scouting. Accessed:
tion of fertilizer cost in fertigation,’’ Comput. Electron. Agricult., vol. 121, Apr. 14, 2019. [Online]. Available: https://www.agweb.com/
pp. 32–42, Feb. 2016. article/a-fully-autonomous-drone-for-daily-scouting/
[231] R. Jordan, G. Eudoxie, K. Maharaj, R. Belfon, and M. Bernard, [254] D. Cozzolino, K. Porker, and M. Laws, ‘‘An overview on the use of
‘‘AgriMaps: Improving site-specific land management through mobile infrared sensors for in field, proximal and at harvest monitoring of cereal
maps,’’ Comput. Electron. Agricult., vol. 123, pp. 292–296, Apr. 2016. crops,’’ Agricult.-Basel, vol. 5, no. 3, pp. 713–722, 2015.
[232] J. C. Ferguson, R. G. Chechetto, C. C. O’Donnell, B. K. Fritz, [255] T. Adão, J. Hruška, L. Pádua, J. Bessa, E. Peres, R. Morais, and
W. C. Hoffmann, C. E. Coleman, B. S. Chauhan, S. W. Adkins, J. J. H. I. Sousa, ‘‘A review on UAV-based sensors, data processing and
G. R. Kruger, and A. J. Hewitt, ‘‘Assessing a novel smartphone applications for agriculture and forestry,’’ Remote Sens., vol. 9, p. 1110,
application—SnapCard, compared to five imaging systems to quantify 2017.
droplet deposition on artificial collectors,’’ Comput. Electron. Agricult., [256] I. Sa, Z. Chen, M. Popovic, R. Khanna, F. Liebisch, J. Nieto, and
vol. 128, pp. 193–198, Oct. 2016. R. Siegwart, ‘‘weedNet: Dense semantic weed classification using multi-
[233] D. Freebairn, B. Robinson, D. McClymont, S. Raine, E. Schmidt, spectral images and MAV for smart farming,’’ IEEE Robot. Autom. Lett.,
V. Skowronski, and J. Eberhard, ‘‘SoilWaterApp—Monitoring soil vol. 3, no. 1, pp. 588–595, Jan. 2018.
[257] (Mar. 2018). E-Agriculture in Action: Drones for Agriculture.
water made easy,’’ in Proc. 18th Aust. Soc. Agron. Conf., Sep. 2017,
[Online]. Available: http://www.fao.org/in-action/e-agriculture-strategy-
pp. 2015–2018.
[234] M. Scholz, ‘‘Enhancing adoption of integrated weed management—An guide/documents/detail/en/c/1114182/
[258] How to Be Smarter Than Your Investors—Continuous Customer Discov-
australian farmer’s perspective,’’ Outlooks Pest Manage., vol. 29, no. 2, ery. Accessed: Apr. 25, 2019. [Online]. Available: https://steveblank.com/
pp. 66–69, Apr. 2018. 2014/02/19/how-to-be-smarter-than-your-investors-continuous-
[235] R. C. L. Suen, K. T. T. Chang, M. P.-H. Wan, Y. C. Ng, and B. C. Y. Tan,
customer-discovery/
‘‘Interactive experiences designed for agricultural communities,’’ in Proc. [259] Q. Yang and S.-J. Yoo, ‘‘Optimal UAV path planning: Sensing data
32nd Annu. ACM Conf. Hum. Factors Comput. Syst. (CHIEA), May 2014, acquisition over IoT sensor networks using multi-objective bio-inspired
pp. 551–554. algorithms,’’ IEEE Access, vol. 6, pp. 13671–13684, 2018.
[236] A. C. Bartlett, A. A. Andales, M. Arabi, and T. A. Bauder, ‘‘A smartphone [260] J. Dai, Y. Wang, C. Wang, J. Ying, and J. Zhai, ‘‘Research on hierarchical
app to extend use of a cloud-based irrigation scheduling tool,’’ Comput. potential field method of path planning for UAVs,’’ in Proc. IEEE 2nd
Electron. Agricult., vol. 111, pp. 127–130, Feb. 2015. Adv. Inf. Manage., Communicates, Electron. Automat. Control Conf.
[237] K. De Sousa, G. Detlefsen, O. Rivera, E. De Melo, D. Tobar, and (IMCEC), Xi’an, China, May 2018, pp. 529–535.
F. Casanoves, ‘‘Using a smartphone app to support participatory agro- [261] S. D’Oleire-Oltmanns, I. Marzolff, K. D. Peter, and J. B. Ries,
forestry planning in Central America,’’ in Proc. 14th Ann. World Congr., ‘‘Unmanned aerial vehicle (UAV) for monitoring soil erosion in
Sep. 2015, pp. 1–9. morocco,’’ Remote Sens., vol. 4, no. 11, pp. 3390–3416, 2012.
[238] W. Maldonado, T. T. B. Valeriano, and G. de Souza Rolim, ‘‘EVAPO: [262] A. Eltner, C. Mulsow, and H. Maas, ‘‘Quantitative measurement
A smartphone application to estimate potential evapotranspiration using of soil erosion from Tls and Uav data,’’ Int. Arch. Photogramm.,
cloud gridded meteorological data from NASA-POWER system,’’ Remote Sens., Spatial Inf. Sci., vol. XL-1, no. W2, pp. 119–124,
Comput. Electron. Agricult., vol. 156, pp. 187–192, Jan. 2019. 2013.

VOLUME 7, 2019 129581

 M. Ayaz et al.: IoT-Based Smart Agriculture: Toward Making the Fields Talk

[263] Harvesting the Agricultural Potential of Drones. Accessed: [288] We Already Grow Enough Food for 10 Billion People—And Still
Sep. 6, 2019. [Online]. Available: https://www.sanfordjournal.org/sjpp/ Can’t End Hunger. Accessed: Apr. 15, 2019. [Online]. Available:
2017/harvesting-the-agricultural-potential-of-drones https://www.huffpost.com/entry/world-hunger_n_1463429
[264] Accessed: Apr. 15, 2019. [Online]. Available: https://www.droneseed.co/ [289] J. M. Mandyck and E. B. Schultz, Food Foolish: The Hidden Connection
[265] These Tree-Planting Drones are About to Start an Entire Forest Between Food Waste, Hunger and Climate Change. 2015.
From the Sky. Accessed: Apr. 15, 2019. [Online]. Available: [290] How to Sustainably Feed 10 Billion People by 2050, in 21 Charts.
https://www.fastcompany.com/40450262/these-tree-planting- Accessed: Jul. 22, 2019. [Online]. Available: https://www.wri.org/
drones-are-about-to-fire-a-million-seeds-to-re-grow-a-forest blog/2018/12/how-sustainably-feed-10-billion-people-2050-21-charts
[266] Former NASA Engineer Plans to Plant 1 Billion Trees a Year Using [291] Key Facts on Food Loss and Waste You Should Know! Accessed:
Drones. Accessed: Apr. 13, 2019. [Online]. Available: https://www. Apr. 15, 2019. [Online]. Available: http://www.fao.org/save-food/
iflscience.com/environment/drones-take-reforestation-new-heights/ resources/keyfindings/en/
[267] Drones to Help Pakistan Plant Millions of Trees. Accessed:
[292] The Roadmap to Reduce U.S. Food Waste by 20 Percent.
Apr. 16, 2019. [Online]. Available: https://gulfnews.com/world/asia/
Accessed: Sep. 8, 2019. [Online]. Available: https://www.refed.com/
pakistan/drones-to-help-pakistan-plant-millions-of-trees-1.61811858
[268] Planting Trees Using Drones is Being Tested in India. Accessed: downloads/ReFED_Report_2016.pdf
Apr. 14, 2019. [Online]. Available: https://www.thecivilengineer. [293] Food Waste Around the World. Accessed: Apr. 15, 2019. [Online]. Avail-
org/news-center/latest-news/item/1388-planting-trees-using-drones-is- able: http://www.mysales-labs.com/food-waste/
being-tested-in-india [294] FAO. Food Wastage Footprint & Climate Change. Accessed: Sep. 8, 2019.
[269] Accessed: Apr. 18, 2019. [Online]. Available: https://www.microdrones. [Online]. Available: www.fao.org/3/a-bb144e.pdf
com/en/integrated-systems/mdaccessory/mdmapper-accessories/ [295] Global Summit, Global Impact. Accessed: Apr. 15, 2019. [Online]. Avail-
[270] R. Szewczyk, C. Zieliński, M. Kaliczyńska, Automation 2018: Advances able: https://foodforthoughtfulaction.com/2016/12/30/global-summit-
in Automation, Robotics and Measurement Techniques. 2019. global-impact/
[271] J. Torres-Sánchez, F. López-Granados, N. Serrano, O. Arquero, and [296] India Pilot Study Shows How the Cold Chain Can Help Reduce Food
J. M. Peña, ‘‘High-throughput 3-D monitoring of agricultural-tree plan- Loss and Carbon Emissions. Accessed: Apr. 15, 2019. [Online].
tations with unmanned aerial vehicle (UAV) technology,’’ PLoS ONE, Available: https://www.carrier.com/carrier/en/us/news/news-article/
vol. 10, no. 6, 2015, Art. no. e0130479. india_pilot_study_shows_how_the_cold_chain_can_help_reduce_food_
[272] C. Romero-Trigueros, P. A. Nortes, J. J. Alarcón, J. E. Hunink, loss_and_carbon_emissions.aspx
M. Parra, S. Contreras, P. Droogers, and E. Nicolás, ‘‘Effects of saline [297] D. Booth, ‘‘Internet of Things builds capacity for automatic temperature
reclaimed waters and deficit irrigation on Citrus physiology assessed by logging,’’ J. Environ. Health, vol. 77, no. 10, p. 34, Jun. 2015.
UAV remote sensing,’’ Agricult. Water Manage., vol. 183, pp. 60–69, [298] CCP Technologies Continues to Make an Impact With the Internet of
Mar. 2017. Things. Accessed: Apr. 15, 2019. [Online]. Available: https://smallcaps.
[273] H. Hoffmann, R. Jensen, A. Thomsen, H. Nieto, J. Rasmussen, and com.au/ccp-technologies-continues-make-impact-internet-of-things/
T. Friborg, ‘‘Crop water stress maps for an entire growing season from [299] AI for Earth. Accessed: Apr. 15, 2019. [Online]. Available: https://www.
visible and thermal UAV imagery,’’ Biogeosciences, vol. 13, no. 24, microsoft.com/en-us/ai/ai-for-earth?activetab=pivot1%3aprimaryr6
pp. 6545–6563, 2016.
[300] FarmBeats: AI, Edge & IoT for Agriculture. Accessed: Apr. 15,
[274] S. Park, D. Ryu, S. Fuentes, H. Chung, E. Hernández-Montes, and
2019. [Online]. Available: https://www.microsoft.com/en-us/research/
M. O’Connell, ‘‘Adaptive estimation of crop water stress in nectarine and
project/farmbeats-iot-agriculture/
peach orchards using high-resolution imagery from an unmanned aerial
vehicle (UAV),’’ Remote Sens., vol. 9, no. 8, p. 828, 2017. [301] Open Agriculture Foundation: Creating an Open-Source Ecosystem
[275] Tasmanian Farmers Use Drones to Make Irrigation More Effi- to Revolutionize the Future of Food. Accessed: Apr. 15, 2019.
cient. Accessed: Apr. 15, 2019. [Online]. Available: https://www.abc. [Online]. Available: https://cloud.google.com/data-solutions-for-change/
net.au/news/2017-07-27/drone-technology-on-farms/8746272 open-agriculture/
[276] Accessed: Apr. 15, 2019. [Online]. Available: http://www.wecanie.com/ [302] Making Informed Decisions With AI. Accessed: Mar. 19, 2019.
html/sprayer/products/20L-sprinkler-drone-in-agriculture.html [Online]. Available: https://www.ibm.com/watson/whitepaper/informed-
[277] Accessed: Apr. 15, 2019. [Online]. Available: https://www.dji.com/mg-1 decisions-ai/
[278] F. Gnädinger and U. Schmidhalter, ‘‘Digital counts of maize plants by [303] Infiswift IoT Platform for Agriculture. Accessed: Apr. 5, 2019. [Online].
unmanned aerial vehicles (UAVs),’’ Remote Sens., vol. 9, no. 6, p. 544, Available: https://www.intel.com/content/www/us/en/internet-of-things/
2017. infiswift-enterprise-iot-platform-for-agricultural-solution-brief.html
[279] X. Jin, S. Liu, F. Baret, M. Hemerlé, and A. Comar, ‘‘Estimates of [304] Agriculture. Accessed: Apr. 15, 2019. [Online]. Available:
plant density of wheat crops at emergence from very low altitude UAV https://www.jasper.com/customers/agriculture
imagery,’’ Remote Sens. Environ., vol. 198, pp. 105–114, Sep. 2017. [305] Accessed: Apr. 15, 2019. [Online]. Available: https://aerofarms.com/
[280] B. S. Faiçal, H. Freitas, P. H. Gomes, L. Y. Mano, G. Pessin, [306] Intelligent IoT Powers Purdue’s Digital Agriculture Initiative for Food
A. C. P. L. F. de Carvalho, B. Krishnamachari, and J. Ueyama, ‘‘An adap- Security Worldwide. Accessed: Apr. 15, 2019. [Online]. Available: https://
tive approach for UAV-based pesticide spraying in dynamic environ- www.hpe.com/us/en/newsroom/blog-post/2017/09/intelligent-iot-
ments,’’ Comput. Electron. Agricult., vol. 138, pp. 210–223, Jun. 2017. powers-purdues-digital-agriculture-initiative-for-food-security-
[281] E. Puig, F. Gonzalez, G. Hamilton, and P. Grundy, ‘‘Assessment
of crop insect damage using unmanned aerial systems: A machine worldwide.html
learning approach,’’ in Proc. 21st Int. Congr. Modelling Simulation, [307] IBM Africa and Hello Tractor Pilot AI/Blockchain Agtech Platform.
Nov./Dec. 2015. Accessed: Mar. 20, 2019. [Online]. Available: https://techcrunch.com/
[282] D. Do, F. Pham, A. Raheja, and S. Bhandari, ‘‘Machine learning tech- 2018/12/20/ibm-africa-and-hello-tractor-pilot-ai-blockchain-agtech-
niques for the assessment of citrus plant health using UAV-based digital platform/
images,’’ Proc. SPIE, vol. 10664, May 2018, Art. no. 106640O. [308] Why Qualcomm Ventures Sees Farm Connectivity Challenges as Opportu-
[283] Hawaiian Flower, Thought Extinct, Rediscovered by a Drone. Accessed: nity. Accessed: Apr. 15, 2019. [Online]. Available: https://agfundernews.
Apr. 15, 2019. [Online]. Available: https://edition.cnn.com/2019/04/ com/qualcomm-ventures.html
18/us/hawaiian-flower-thought-extinct-rediscovered-trnd/index.html [309] Accessed: Apr. 15, 2019. [Online]. Available: http://ninjacart.in/
[284] R. Clark, ‘‘Mapping and estimating forest fuel with radar remote sens- [310] Farmeasy. Accessed: Apr. 7, 2019. [Online]. Available: https://www.
ing,’’ Fire Sci. Brief., no. 57, Jul. 2009, p. 1. qualcommventures.com/companies/internet-things/farmeasy
[285] How do we Feed the Planet in 2050? Accessed: Apr. 15, 2019. [Online].
Available: https://www.theguardian.com/preparing-for-9-billion/2017/ [311] Qualcomm Announces Leadership Change in Latin America.
sep/13/population-feed-planet-2050-cold-chain-environment Accessed: Mar. 25, 2019. [Online]. Available: https://www.qualcomm.
[286] 2018 World Hunger and Poverty Facts and Statistics. Accessed: com/news/releases/2011/09/27/qualcomm-announces-leadership-
Apr. 15, 2019. [Online]. Available: https://www.worldhunger.org/ change-latin-america
world-hunger-and-poverty-facts-and-statistics/ [312] Accessed: Apr. 15, 2019. [Online]. Available: http://www.farm2050.com/
[287] M. C. Hunter, R. G. Smith, M. E. Schipanski, L. W. Atwood, and [313] Global Hunger Continues to Rise, New UN Report Says. Accessed:
D. A. Mortensen, ‘‘Agriculture in 2050: Recalibrating targets for sustain- Apr. 15, 2019. [Online]. Available: https://www.who.int/news-room/
able intensification,’’ Bioscience, vol. 67, no. 4, pp. 386–391, Feb. 2017. detail/11-09-2018-global-hunger-continues-to-rise—new-un-report-says

129582 VOLUME 7, 2019

M. Ayaz et al.: IoT-Based Smart Agriculture: Toward Making the Fields Talk

[314] Health Effects of Dietary Risks in 195 Countries. Accessed: ZUBAIR SHARIF received the B.S. degree from
Apr. 19, 2019. [Online]. Available: https://www.thelancet. COMSATS University, Vehari, Pakistan, in 2017,
com/journals/lancet/article/PIIS0140-6736(19)30041-8/fulltext and the M.S. degree from COMSATS University,
[315] FAO. Global Agriculture Towards 2050: High-Level Expert Forum on Sahiwal, in 2019, both in computer science. He is
How to Feed the World in 2050. Accessed: Oct. 2009. [Online]. Avail- currently a part-time Faculty Member with COM-
able: www.fao.org/wsfs/forum2050/wsfs-background-documents/wsfs- SATS University, Vehari. His research interests
expert-papers/en/ include wireless sensors and ad hoc networks.
[316] Why IoT, Big Data & Smart Farming are the Future of Agri-
culture. Accessed: Apr. 17, 2019. [Online]. Available: https://www.
businessinsider.com/internet-of-things-smart-agriculture-2016-10
[317] Huawei. The Connected Farm—A Smart Agriculture Market Assess-
ment. Accessed: Sep. 9, 2019. [Online]. Available: https://huaweihub.
com.au/the-connected-farm-a-smart-agriculture-market-assessment/
[318] World Migration Report, Int. Org. Migration, Grand-Saconnex,
Switzerland, 2015.
[319] 68% of the World Population Projected to Live in Urban Areas by 2050.
Accessed: Apr. 18, 2019. [Online]. Available: https://www.un.org/
development/desa/en/news/population/2018-revision-of-world- ALI MANSOUR received the M.Sc. and Ph.D.
urbanization-prospects.html degrees from INPG, France, and the HDR degree
[320] Chinese Scientists Successfully Grow Rice in Dubai Desert. (Habilitation a Diriger des Recherches, the high-
Accessed: Apr. 15, 2019. [Online]. Available: https://gbtimes.com/ est degree in France) from UBO, France. He
chinese-scientists-successfully-grow-rice-in-dubai-desert held many positions, including the postdoctoral
[321] How One Norwegian Firm Hopes to Turn the Deserts Green. position at LTIRF, France, Researcher at RIKEN,
Accessed: Mar. 10, 2019. [Online]. Available: https://www. Japan, Teacher and Researcher at ENSIETA,
arabianbusiness.com/technology/409003-making-the-deserts-go-green France, Senior Lecturer at Curtin University,
[322] Abu Dhabi Partners With Chinese Firm to Convert Desert Into Farmland. Australia, Invited Professor at ULCO, France, and
Accessed: Apr. 15, 2019. [Online]. Available: https://gulfbusiness.com/ Professor at Tabuk University, Saudi Arabia. Since
abu-dhabi-partners-with-chinese-firm-to-convert-desert-into-farmland/ September 2012, he has been a Professor with École Nationale Supérieure
de Techniques Avancées Bretagne (ENSTA Bretagne), Brest, France. He
has published more than 170 refereed publications. His H-index is 27. He
MUHAMMAD AYAZ received the M.S. degree is the author of a book and the coauthor of three other books and four
in computer science from SZABIST Islamabad book chapters. During his career, he had successfully supervised several
Pakistan and the Ph.D. degree in information tech- postdoctoral and Ph.D. candidates and M.Sc. students. His research interests
nology from University Teknologi PETRONAS, include source separation, high-order statistics, signal processing, robotics,
Malaysia, in 2007 and 2011, respectively. He is telecommunication, biomedical engineering, electronic warfare, and cogni-
currently an Assistant Professor and a Researcher tive radio. He was the Vice President of the IEEE Signal Processing Society
with the Sensor Networks and Cellular Systems in Western Australia for two years. He has also been the Lead Guest Editor
(SNCS) Research Center, University of Tabuk, of the EURASIP Journal on Advances in Signal Processing.
Saudi Arabia. He led various research projects
funded by the University of Tabuk and the Ministry
of Higher Education, Saudi Arabia, especially related to water quality
monitoring. He is the author of many research articles published by the
IEEE, Elsevier, Springer, Wiley, and other well-known journals. His research
interests include mobile and sensor networks, routing protocols, network
security, and underwater acoustic sensor networks.
EL-HADI M. AGGOUNE received the M.S. and
Ph.D. degrees in electrical engineering from the
University of Washington (UW), Seattle, USA.
MOHAMMAD AMMAD-UDDIN received the He has taught graduate and undergraduate courses
M.S. degree in computer networks from the in electrical engineering at many universities in
COMSATS Institute of Information Technology, USA and abroad. He has served at many academic
Islamabad, Pakistan, the M.Sc. degree in com- ranks, including the Endowed Chair Professor,
puter science in software engineering from Bahria the Vice President, and the Provost. His research
University, Islamabad, and the Ph.D. degree in work is referred to in many patents, including the
wireless sensors network form ENSTA, Bretagne, patents assigned to ABB, Switzerland, and EPRI,
France. He is CCNA and CCAI Certified. He USA. He is currently a Professor and the Director of the Sensor Networks
is currently a Senior Researcher with the Sensor and Cellular Systems (SNCS) Research Center, University of Tabuk, Tabuk,
Networks and Cellular Systems Research Centre, Saudi Arabia. He is also a Registered Professional Engineer in the State of
University of Tabuk, Saudi Arabia. He taught many graduate and under- Washington. He has authored many articles in the IEEE and other journals
graduate computer courses in a number of universities in Saudi Arabia and and conferences. He has been serving on many technical committees. His
abroad. He led many research and development projects in the areas of wire- research interests include power systems, wireless sensor networks, scientific
less sensor networks, underwater sensor networks, and smart agriculture. visualization, and neural networks. One of the laboratories, he directed won
He is the author of many research articles published in the IEEE and other the Boeing Supplier Excellence Award. He was also the winner of the IEEE
journals. He is listed as an Inventor in a patent registered in USA. His research Professor of the Year Award at the UW Branch. He is listed as an Inventor in
interests include routing, clustering, and localization of sensors nodes in a major patent assigned to Boeing Company.
wireless sensor networks.

VOLUME 7, 2019 129583