See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/340326857

Internet of Things Empowered Smart Greenhouse

Article in IEEE Journal of Radio Frequency Identification · March 2020

DOI: 10.1109/JRFID.2020.2984391

CITATIONS READS
146 22,150

3 authors, including:

George Xiao Zheng Liu

National Research Council Canada University of British Columbia - Okanagan
205 PUBLICATIONS 3,175 CITATIONS 376 PUBLICATIONS 10,245 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Rakiba Rayhana on 21 November 2020.

The user has requested enhancement of the downloaded file.

 IEEE JOURNAL OF RADIO FREQUENCY IDENTIFICATION, VOL. 4, NO. 3, SEPTEMBER 2020 195

Internet of Things Empowered Smart

Greenhouse Farming
Rakiba Rayhana, Student Member, IEEE, Gaozhi Xiao , Fellow, IEEE, and Zheng Liu , Senior Member, IEEE

Abstract—The rapid change of climate, population explosion, century [5]. Since then, greenhouse farming has become one
and reduction of arable lands are calling for new approaches to of the fastest-growing industries in the world [6]. Greenhouse
ensure sustainable agriculture and food supply for the future. farming is a house-like structure, constructed with glass or
Greenhouse agriculture is considered to be a viable alternative
and sustainable solution, which can combat the future food cri- plastic materials to produce various types of crops in any
sis by controlling the local environment and growing crops all season [7]. The roof is generally covered with transpicuous
year round, even in harsh outdoor conditions. However, green- material to preserve the required climatic conditions for the
house farms persist many challenges for efficient operation and plant growth (such as, temperature, humidity, lighting, and so
management. The evolving Internet of Things (IoT) technologies, on) and also to safeguard the plants from the pests, diseases
which encompass the smart sensors, devices, network topolo-
gies, big data analytics, and intelligent decision is believed to be and adverse environmental conditions [8].
the solution in addressing the key challenges facing the green- Greenhouse farms can be employed in different environ-
house farming, such as greenhouse local climate control, crop mental conditions. In an extreme cold weather condition,
growth monitoring, crop harvesting and etc. This paper reviews “cold frame;” a type of small-scale greenhouse farms can be
the current greenhouse cultivation technologies as well as the installed. A cold frame can trap the heat from the sunlight and
state-of-the-art of IoT technologies for smart greenhouse farms.
The paper also highlights the major challenges that need to be can preserve the temperature for the plants. On the other hand,
addressed. for dry and hot weather, “shade house” greenhouses are used
to provide shades and preserve the moisture for the plants [9].
Index Terms—IoT, RFID, sensors, smart farming, greenhouse
farming, agriculture. The advantages of these greenhouse farms are that it can
help the farmers to produce different types of crops by chang-
ing the local environmental conditions according to the plant’s
I. I NTRODUCTION requirement (temperature, light, moisture, nutrients) [6]. It
can prolong the growing season for the cultivation and can
ITH the fast growth of population, industrialization,
W climate change, the spreading of environmental pollu-
tion, the arable land around the world are decreasing year by
provide high quality crops with efficient use of pesticides,
manures, water and labors [10], [11]. Greenhouse farms can
prevent plants from harsh environmental conditions, such as
year [1]. This poses significant challenges for future food secu-
heavy rainfall or high solar radiation. It prevents the plants
rity. According to a recent survey conducted by United Nation
to be infected by aerial borne diseases as it provides a pro-
Food and Agriculture Organization (FAO), the population of
tected environment for plant growth [12]. Depending upon the
the world would increase to 9.73 billion by 2050 [2] and 11.2
greenhouse facilities, the yield of the crops can increase up to
billion by the end of 2100 [2]. This upsurges a global concern
10-12% than the traditional farming [9]. The greenhouse can
for the farmers to meet future food demands. Moreover, the
also be used to grow crops which are not able to survive the
abrupt changes in the weather and increase of salinity in the
outdoor conditions. However, the maintenance of the green-
freshwater spiral the pressure for the farmers [3]. Therefore,
house farms is quite strenuous as it needs more resources than
the traditional agriculture industry requires a radical change
the traditional farms [9]. The yield of the crops are greatly
to ensure ecological and sustainable food supply. Greenhouse
dependent upon the appropriate parameter control, such as
agriculture or farming technology is considered to be one of
cooling, heating, lighting, water flow, ventilation, level of CO2
the viable and alternative solutions to provide food security
and etc. inside the farm [6]. This can influence the economic
and to ensure socio-ecological sustainability in the future [4].
profits and can increase the complexities of the decision mak-
Greenhouse farming technology was first introduced com-
ing process for the farmers. Integrating the Internet of Things
mercially in The Netherlands and France in the nineteenth
(IoT) technology with the greenhouse to make it “smart,” i.e.,
Manuscript received February 18, 2020; accepted March 21, 2020. Date the smart greenhouse, is believed to be the solution. Hence,
of publication March 31, 2020; date of current version August 26, 2020. smart greenhouse farming can leverage these problems and aid
(Corresponding author: Gaozhi Xiao.) the farmers to increase the yield of the crops [13].
Rakiba Rayhana and Zheng Liu are with the Department of Electrical
Engineering, University of British Columbia, Kelowna, BC V1V 1V7, Canada The term “Internet of Things (IoT)” refers to a mass-
(e-mail: [email protected]; [email protected]). system which is connected to numerous sensors, embedded
Gaozhi Xiao is with the Advanced Electronics and Photonics Research controllers, decision-making platforms, Internet, and a cloud
Center, National Research Council of Canada, Ottawa, ON K1A 0R6, Canada
(e-mail: [email protected]). server. The sensors collect the data and automatically feed
Digital Object Identifier 10.1109/JRFID.2020.2984391 them to the cloud server. The cloud servers store and allow
2469-7281 
c 2020 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.
See https://www.ieee.org/publications/rights/index.html for more information.

Authorized licensed use limited to: The University of British Columbia Library. Downloaded on September 23,2020 at 18:46:50 UTC from IEEE Xplore. Restrictions apply.
 196 IEEE JOURNAL OF RADIO FREQUENCY IDENTIFICATION, VOL. 4, NO. 3, SEPTEMBER 2020

TABLE I
G REENHOUSES BASED ON C LIMATIC C ONDITIONS

TABLE II
T YPES OF G REENHOUSE C ULTIVATION T ECHNOLOGIES

the data to be accessed remotely [14]. Therefore, IoT systems lowered the cost of construction. However, the construction of
are implemented mainly to monitor and take intelligent actions the greenhouse farms does not only depend upon the mate-
without human intervention, such as cooling, heating, lighting, rial (plastic, glass). It is highly dependent upon the local
irrigation, on and off of the motors and actuators by analyzing climatic conditions [16]. The common requirement of con-
the sensor data [15], making the farm smart. Smart greenhouse structing a greenhouse farm is to select a flat land site by
farms can also aid the farmers to gain knowledge about the ensuring adequate (all year round) solar radiation [16].
season or the most suitable time for harvesting, quality of The primary principle and the history of the greenhouse
the soil, amount of nutrients needed for healthy plant growth, farms depict that it was used to grow crops or vegetables which
quality of water and many other important factors. In other requires heat for the growth in the cold climatic regions. This is
words, smart greenhouse farming can make farming reliable, due to the greenhouse effect and wind-break effects which aid
cost-effective, and maximize the yield of the crops with the in plant growth. However, with the advancement of new tech-
minimum number of labors [13]. This can help the farmers nologies and smart systems, greenhouses have become more
to manage the farms and optimize the resources efficiently. It versatile and now they can be constructed in almost any cli-
is envisioned that in the future the growing of crops inside a matic condition. Generally speaking, climatic conditions can
smart greenhouse (or at least part of the process) can be fully be classified into three main categories: 1) cold and temper-
automated and remotely controlled. ate winter 2) tropical and sub-tropical 3) arid or semi-arid
This paper aims to assist researchers and practitioners in the regions [16]. Depending upon the weather and the invest-
related fields, to gain the knowledge of the smart greenhouse ment of the farmers, the selection of the shape, construction
farming, the state-of-the-art and the current challenges. The material, roof covering and ventilation of the greenhouse is
paper summarizes the existing greenhouse cultivation tech- varied [17]. Table I summarizes the types of greenhouse farms,
nologies, analyzes the current status of IoT technologies and their construction materials, and the crops being grown within.
their suitability for applications on smart greenhouse farming, In terms of cultivation technologies, greenhouses can be
and highlights the shortcomings of the different IoT systems. mainly divided into three types [25] as presented in Table II.
This paper aims to serve as a baseline for the researchers to A brief description of each type of greenhouse is stated in the
resolve the current challenges. following.
The paper is organized as follows. Section II discusses the 1) Low Technology Greenhouse Farms (LTG): Low-
different types of greenhouse cultivation technologies that are Technology greenhouse farms are made up of a simple house
currently used. Section III reviews the current IoT architectures like a wooden structure which has a roof covering with plas-
being used in smart greenhouse farmings. Section IV discusses tics. It is normally built in colder or temperate zones [12]. This
the sensor technologies that are used in monitoring the cli- type of farm has no/low-grade heating system, no ventilation
mate, soil condition, plant growth, and disease development. and often has poor control of the internal climatic conditions.
Section V describes the data transmission and processing. LTG farms are usually used to produce vegetables and flowers
Section VI concludes the paper by highlighting the major and the installation cost is less than $25-$30 per square meter
challenges for potential future research directions. of the total field area [11], [25]. This type of greenhouse farms
is out of the scope of this paper.
2) Medium Technology Greenhouse Farms (MTG): Medium
II. G REENHOUSE C ULTIVATION T ECHNOLOGIES Technology Greenhouse Farms are constructed with metal
Since 1991 [11], there has been a substantial increase in frames. Both plastic and glass panes are used as roof covering.
the practicing of greenhouse agriculture around the world. Unlike LTG farms, MTG has better efficiency in control-
This is mainly due to the advent of plastic technology which ling the climatic conditions, making the farm independent

Authorized licensed use limited to: The University of British Columbia Library. Downloaded on September 23,2020 at 18:46:50 UTC from IEEE Xplore. Restrictions apply.
 RAYHANA et al.: INTERNET OF THINGS EMPOWERED SMART GREENHOUSE FARMING 197

from the external conditions. This makes it easier to construct can grow all year round crops for sustainable food supply [30].
in the tropical or subtropical climate zones [12]. Advanced The environmental parameters can be monitored by installing
farming technologies, such as hydroponics, aeroponic and IoT enabled sensors and controlled by IoT enabled infrastruc-
nutrient film farming employs MTG. MTGs are often semi tures. To eradicate the manual monitoring problem for the
to fully automated farms and can be used to cultivate high- farmers, a cloud server is connected through the Internet with
end flowers, such as roses, decorative pot plants and out of the sensors [30]. The installed cloud server makes it possi-
season vegetables. The installation cost of the farm is around ble for the remote data processing and the control of IoT
$30-$100 per square meter [11], [25]. enabled infrastructures, such as the window opener and irriga-
3) High Technology Greenhouse Farms (HTG): High tion valves. The IoT empowered smart greenhouse provides an
Technology Greenhouse Farms (HTG) are usually constructed optimum and capital-intensive solution and reduces the need
with galvanized iron frames and glass as the roof covering. of most manual interventions [30], thus solving the challenges
HTG farms have all the advanced climate control technolo- discussed above.
gies, making it suitable for any climatic conditions [26]. It
can control the heating by sensing the internal climatic condi-
tions and provide forced cooling and ventilation. Also, it can III. OVERVIEW OF I OT A RCHITECTURES U SED IN S MART
control the humidity, luminous level and CO2 levels inside G REENHOUSE FARMING
the greenhouse. HTGs are used in vertical farming and it can Nowadays, the IoT enabled devices are widely used in
minimize the labor cost for the farms. It is often employed in the control and diagnostic systems of the agriculture indus-
colder climatic regions to produce different types of vegeta- try [31], [32]. The IoT technologies comprise of sensors,
bles and decorative plants. The installation cost of HTG farms actuators, cloud computing based data amenities, drones, nav-
are about $100-$200 per square meter [11], [25]. igation and analytical system, which allows the architecture
to make intelligent decisions to increase the crop yield [33].
A. Challenges in Greenhouse Farming IoT devices can provide information about the environmental
Greenhouse farming faces many issues. Some of them have variables including, humidity, temperature and climatic condi-
been tackled and many of them are still lingering around. For tions, and also about the field variables such as soil and plant
example, the fluctuation in temperature, and humidity, and bio-masses [34]. It can be employed to predict and monitor
the inadequate lighting. The temperature variation which is the quality of the crops for the consumers [35]. Additionally,
a common problem in the greenhouses changes the physical IoT can be used to collect data and store them in cloud com-
parameters of the plant. Depending on the variation extent, it puting devices to create alert and send short messages services
might reduce the productivity of the crops, delays the harvest- (SMS) to the farmers. The data stored in the cloud can also
ing time and make it non-viable for all year round production. be used to develop predictive models which can prognosti-
On the other hand, the fluctuation in humidity might result in cate the variables that affect the crops [35]. An illustration of
nutrient deficiency and pest attacks of the plants. Furthermore, the envisioned IoT empowered smart greenhouse farming is
inadequate lightning can decrease CO2 level and the rate of presented in Fig. 1. This section overviews the current state
photosynthesis for the plants. This directly affects the yield of the IoT architectures that are used for smart greenhouses,
and the quality of the crops [11]. in particular, the three typical greenhouse growing techniques,
The main challenges for the greenhouse farms are that it i.e., traditional, hydroponics and vertical farming.
requires extensive supervision to control and maintain the
required atmosphere for the plant growth in order to achieve A. IoT Implementation for Traditional Greenhouse Farming
expected yield and crop quality [27]. This introduces the need
Traditional greenhouse farming technique is mainly adopted
for skilled/ knowledgeable labors for the monitoring of the
for the growing of fruits and vegetables. A general IoT archi-
environmental conditions (temperature, humidity, luminosity)
tecture for this type of crop growing is presented in Fig. 2 One
and taking an efficient decision to protect the crops from pests
of the main goals of employing IoT in greenhouse farming is
or other problems [28].
to ensure a long term sustainable solution for the farmers. In
2012, Li et al. [36] developed an IoT based environmental
B. Potential Solutions With Smart Greenhouse Integrated monitoring system for the greenhouse by integrating wire-
With IoT less networks, mobile network, and the Internet to perform
To resolve the aforementioned challenges integrating IoT remote monitoring of the plants in real-time. The architec-
into the greenhouse to make it smart is believed to be the ture comprises of CC2420 (ZigBee) Radio frequency modules,
most suitable solution. IoT can not only automate the moni- temperature and humidity sensors, single-chip microcomputer
toring and controlling system of the farm but it can also aid the (STC9051 SCM) with wireless transmission module. This
farmers in decision making and the efficient utilization of the architecture could help to warn the farmers about the plant
resources (water, pesticides, heating and etc.) [29]. The inte- conditions through short messages and also could provide
gration of smart technologies, which now become part of IoT real-time information of the plants [36].
with greenhouse farming is also known as “Smart Greenhouse In 2014, Kaewmard and Saiyod [37] designed a portable
Farming”. By definition, smart greenhouse farming refers to measurement system which included the sensors to measure
high technology-based cost-effective greenhouse farms which air humidity, temperature, and moisture of the soil to control

Authorized licensed use limited to: The University of British Columbia Library. Downloaded on September 23,2020 at 18:46:50 UTC from IEEE Xplore. Restrictions apply.
 198 IEEE JOURNAL OF RADIO FREQUENCY IDENTIFICATION, VOL. 4, NO. 3, SEPTEMBER 2020

Fig. 1. Envisioned IoT empowered Smart greenhouse.

the water flow/ irrigation system of the vegetables via smart- the developed network was very efficient for monitoring the
phones and wireless sensor network. An underground water plant growth and detecting disease of the tomato plants,
flow sensor was developed and installed in the root of the and was 1.71 times faster than other established detection
vegetable crops to monitor soil moisture and instruct the Xbee methods [41]. Aleotti et al. [42] presented a mobile based
module to take action (supplying water) [37]. Minh et al. [38] platform with embedded IoT devices combined with an intel-
developed a light-weight prototype by combining sensor data ligent decision making system (Irriframe) to control the water
and communication modules to automate the environment of and monitor the growth of tomato plants in real time. An
the greenhouse farm and named it as “Smart IoT Kit”. The unmanned aerial vehicle, UAV (VirtualRobotix SPARK quad-
study was conducted on the growth of maize and oyster mush- copter) which is connected with Raspberry Pi 3 computer and
room and the prototype could help the farmers to control the a Raspberry Pi NoIR camera was used to collect images.
farms from any place and time [38]. The images are then processed with Orthophoto Processing
Mittal et al. [39] developed a system which can measure up Module providing the condition of the tomato plants to the
to fourteen ambient parameters under the surface of the soil farmers [42].
with the aid of different IoT enabled devices and cloud. The Another study regarding the distribution of water was con-
system is also capable of monitoring the growth of cabbage ducted by [6], where they designed an embedded IoT based
and capsicum. This system includes ZigBee WSN, sensors to system which has intelligent sensors and actuators. It uses
measure the temperature and moisture of the soil, and the rel- Arduino as an integration platform, ZigBee and cloud services
ative air humidity. It is integrated with mKRISHI platform for networking and smart interactions. The system also incor-
and HTTPS is used as transport protocol between the edge porates IoT hardware-encryption security (TM4C129E) to
and cloud computing. This monitoring system was reported ensure data security. The work assisted the farmer to improve
improving the production of the crops by up to 10% and the harvest of the vegetables [6]. Medela et al. [32] developed
reducing the cost by 20% [39]. a new IoT architecture by incorporating sensors, wireless
Park et al. [40] introduced a layer-based data analysis networks and weather data to monitor the growth of grapes.
approach to monitor the growth of tomatoes. The data was Although significant progress has been made so far for the
collected from their own developed IoT architecture which development of IoT architectures targeting the efficient growth
comprised of humidity, temperature (indoor and outdoor), of fruit and vegetables, such as the monitoring of global envi-
soil moisture, CO2 concentration sensors with Raspberry ronmental conditions inside the greenhouse and the detection
Pi 3 and DART framework for IoT gateways. The results of diseases, many obstacles are still laying ahead. For example,
from this study showed that the data generated from the the parameters obtained from the sensors (temperature, humid-
system could help the growers predict the yield of toma- ity, luminosity) are not enough for the building of an efficient
toes and the layered analysis could ensure the privacy of the decision making platform. More environmental parameters
data [40]. Another study on tomato fields was conducted by such as field water level, plant nutrients (nitrogen, calcium),
Rupanagudi et al. [41] to monitor plant growth and the dis- chlorophyll are required. Another example is that the auto-
eases. The authors used robotic platforms to collect videos mated watering systems in the current IoT based greenhouse
of the tomato plants from the field and the live video feed farms are not crop specific and it often ends up over-watering
is processed with the help of Java, cloud computing devices the plants [36]. This pushes for the optimization of the irri-
and machine learning algorithms. The study concluded that gation/water supply [42]. Furthermore, a broad knowledge of

Authorized licensed use limited to: The University of British Columbia Library. Downloaded on September 23,2020 at 18:46:50 UTC from IEEE Xplore. Restrictions apply.
 RAYHANA et al.: INTERNET OF THINGS EMPOWERED SMART GREENHOUSE FARMING 199

Fig. 3. A typical block diagram of hydroponics in smart greenhouse.

pH, light intensity and globe environmental humidity sen-

Fig. 2. A general IoT architecture for fruits and vegetables. sors. Mehra et al. [46] proposed a prototype for tomato
plants which integrates sensors with the Raspberry Pi 3 and
Arduino board to monitor the environmental parameters in
the environment and how to grow different types of crops is real-time. The prototype also uses Artificial Neural Network
required in order to respond efficiently from the collected data. (ANN) to process multiple parameters (such as light intensity,
Even though the current IoT devices aid in the automation of humidity, temperature and so on) and store the data on the
the actuators or pumps, the decision making system is not cloud in real-time. The developed model showed a satisfac-
there or not intelligent enough to be crop specific [37]. For tory performance (88% accuracy) and it can be improved by
instance, when the lights are automatically turned on with the considering other important parameters [46].
aid of sensor data, the current system is not capable of indicat- Paulchamy et al. [47] developed an “Intelligent hydroponic
ing that how long the light should be on and what should be the box” with IoT which can automatically turn on and off the
light intensity for the specific crop that is being planted [38]. actuators by sensing the change in the sensor data. This archi-
This leads to energy waste. tecture is capable of lowering the CO2 level and can control
the water flow through the intelligent plant care hydroponic
(IPCH) box [47]. Pitakphongmetha et al. [48] proposed to use
B. IoT Implementation for Hydroponics a wireless sensor network (WSN) to transmit the sensor data to
The word “hydroponic” refers to the condition where plants the cloud. It is interfaced through an android application called
are grown in the water instead of soil. Hydroponics are partic- “Blynk” was claimed to be able to successfully monitor the
ularly suitable for greenhouse farms. One of the most common plant growth of hydroponics [48].
hydroponic greenhouse vegetables is lettuce [43]. Nowadays, The advantage of IoT based hydroponic greenhouse farming
broccoli and tomatoes are also planted through hydropon- is that it does not require unnecessary pesticides for the soil.
ics [44]. However, growing those plants is very difficult as The water requirement is less than the traditional greenhouse
they are very sensitive to the change of pH in the water and farms. Also, the plant yield is unaffected by any external envi-
the level of proper nutrients. Therefore, IoT architectures were ronmental changes [49]. Despite of the IoT enabled devices,
explored to automate the monitoring of the plants inside the the main challenge of hydroponic farms are that it requires
greenhouse [44]. Saraswathi et al. [44] developed an IoT based extensive manual monitoring to sprinkle waters and to provide
mechanism to control the pH of the water through a smart- the required nutrients [46]. The scalability of the hydro-
phone application which could help to monitor the current con- ponic farm is also a big issue as the IoT based hydroponic
dition of the plants. A Raspberry Pi 3 board is interfaced with architectures have high initial investment to set up. The com-
the humidity, temperature, pH and pressure sensor. An android munication network that is usually used in the hydroponic
smartphone application is developed by using Python program- farms is not suitable for long distance communication [49].
ming language which can help the farmer (users) to know Therefore, an efficient and reliable communication network
about the current condition of the plants. Fig. 3 shows a typical should be developed to improve the communication system to
IoT architecture for hydroponic plants in the greenhouse. easily access the data. In addition, how to correlate the plant
Ferentinos and Albright [45] developed a GUI control growth data with the climatic conditions and nutrient levels is
system with the aid of the Bayesian predictive network. The still in its infancy. This poses great challenges for the devel-
system actuates the control based on the feedback from the opment of decision making algorithm. In majority cases, the
sensors which include the plant growth data, water level, decision making process is still experience based.

Authorized licensed use limited to: The University of British Columbia Library. Downloaded on September 23,2020 at 18:46:50 UTC from IEEE Xplore. Restrictions apply.
 200 IEEE JOURNAL OF RADIO FREQUENCY IDENTIFICATION, VOL. 4, NO. 3, SEPTEMBER 2020

place [51]. Smart vertical farms can have an automatic irri-

gation system and can reduce the labor expenses. However,
vertical farming requires adequate lighting/ luminosity lev-
els which increases energy consumption. This also adds to
the investment capital. Vertical farming is relatively new in
studies. Therefore, different types of alert systems can be
integrated to warn the farmers. Also, no studies related to the
fully automated and optimized models of vertical farming are
conducted yet.

IV. S ENSOR T ECHNOLOGIES FOR S MART G REENHOUSE

One of the most indispensable parts of the IoT architectures
is the sensor systems which provides information about the
environment and plants in the greenhouse. With the emergence
of new materials, new theories and new fabrication process,
Fig. 4. An illustration of a vertical farm. agriculture technologies have also matured a lot [53] through
the efforts of some leading manufacturers such as ENDEVCO,
C. IoT Implementation for Vertical Farming Bell & Howell, HoneyWell, Solartron and Foxboro [54]. The
most common sensors that are deployed in smart greenhouses
Vertical farming is a vertical assembled farm which is often
are humidity, temperature, soil moisture, light intensity, heat
built-in indoor greenhouse farms (controlled environment).
and gas sensors.
Vertical farming is an eco-friendly and an insecticide free farm.
The sensor technologies that are used in greenhouses can be
However, this type of farm requires high level of maintenance
mainly divided into two types, i.e., bio/chemical sensors, and
to grow the plants. In vertical farming, LED lights are used to
physical parameter sensors [54]. Biosensors [55] are the type
compensate for sunlight. Fig. 4 shows a typical vertical farm-
of sensors which monitor the information directly related to
ing layout. Similar to other greenhouse farms, the temperature,
the plant/organism including microbial sensors, adaptive sen-
humidity, soil moisture also needs to be controlled [50].
sor [56], and enzyme sensor [57]. This type of sensor is often
Recently, Bhowmick et al. [50] proposed to use IoT based
used to identify the residue of insecticides, antibiotic and toxic
wireless monitoring for vertical farming. The IoT architec-
elements (gas) [58], [59]. Physical sensors are mainly used to
ture incorporates sensors to obtain the data of temperature,
collect the data related to the environments such as tempera-
light intensity, humidity, and soil moisture, a wireless micro-
ture, humidity and heat sensors [54]. Micro electro-mechanical
controller module for data transmission and a Web-based
system (MEMS) has been employed to make low power, com-
application to observe the sensor data. The sensor data are
pact, cheap and reliable sensors [60]. The following section
analyzed through a wireless module (Intel Edison) with the
discusses different types of sensors that are being widely used
Arduino board, which is connected with Linux OS, Eclipse
for monitoring purposes in smart greenhouse farms.
and Intel XDK. The Web server that is used to interface
the data to the cloud is ThingSpeak and for mobile devices,
the Virtuino android app was used to send SMS [50]. A. Monitoring of Climate
Bin Ismail and Thamrin [51] developed an IoT based plant One of the most importan components of smart greenhouse
monitoring system for “ Typhonium Flagelliforme” herbs. farming is to monitor the climatic conditions or the environment
The architecture incorporates automatic on/ off system for inside farms. The real-time monitoring can help the farmers
the water pump, wireless sensors, (one water level sensor to take preventive measures for the adverse conditions and
and three soil moisture level sensors) and LED lights for protect the plants from diseases. Marques and Pitarma [61]
photosynthesis. The system is interfaced with the Ethernet developed an automatic environment monitoring system with
shield which helps for communication purposes. The system Sun Spot, an affordable wireless sensor network which is pro-
shows promising results but it could be further enhanced by grammed by Java. The system uses Java virtual machine (VM)
using efficient LED lights and by integrating alarm systems with the Arduino board to monitor the temperature of the
for the farms [51]. Belista et al. [52] developed an Internet greenhouse and data are accumulated through Facebook [61].
based modular network which could effectively transmit the Wu et al. [62] developed a WSN based modular environment
sensor data to the clouds to monitor the plant growth for let- monitoring structure for smart greenhouse. The structure com-
tuce farms. The developed server was named “Samba” and prises of three layers, i.e., sensor layer, transmission layer
through Raspberry Pi 3 it could be connected to any operating and application layer. The WSN can obtain the environmen-
system (Windows or Linux). The designed network performed tal data of the tomato farms in real-time and transmit the
well and it can be further optimized by considering more data to the cloud server. The results of this study show that
environmental parameters [52]. WSN is a stable network for monitoring the plants growing
Smart vertical greenhouse farms have many advantages. One conditions [62]. Saraswati et al. [63] proposed and imple-
of them is that the productivity of the crops remains the same mented an android based climate monitoring system for spinach
even if any natural calamities (typhoon, storm, hurricanes) take plants by integrating sensors to gather the data of humidity,

Authorized licensed use limited to: The University of British Columbia Library. Downloaded on September 23,2020 at 18:46:50 UTC from IEEE Xplore. Restrictions apply.
 RAYHANA et al.: INTERNET OF THINGS EMPOWERED SMART GREENHOUSE FARMING 201

TABLE III
D IFFERENT T YPES OF S ENSORS T HAT ARE U SED IN G REENHOUSE FARMS

temperature, light intensity and level of carbon dioxide inside the

greenhouses.
Some particular plants such as onions and shallots require
very cold climatic conditions to grow whereas plants like mel-
ons and cucumber require warm climatic conditions. Hence,
Danita et al. [64] proposed an IoT based embedded technology
which could integrate sensor parameters (moisture, temper-
ature, humidity) with a Raspberry Pi 3 and ThingSpeak
cloud interface to provide the ideal conditions for the plants
inside the greenhouses. Pahuja et al. [65] proposed an inte-
grated WSN to monitor the micro-climatic conditions inside Fig. 5. A block diagram of soil moisture monitoring system.
a greenhouse farm by using VPD-based multiple inputs and
multiple-output (MIMO) controller with a battery operated
RS-485 actuator and embedded temperature and humidity sen-
sors. A model based on pre-programmed TinyOS was used
to automate the environmental conditions inside a greenhouse are all dependent upon the moisture level of the roots in the
farm. Their study proved that the climatic analysis of the soil. The information about soil moisture can help the farm-
greenhouse could aid the farmer to determine effective con- ers to comprehend the conditions and make decisions. Soil
ditions for greenhouse vegetables, leading to a higher yield moisture evaluates the water content of the soil [75]. Studies
of crop production [65]. Nayyar and Puri [66] developed a showed that almost 80% of the crop is damaged due to over
low-cost Smart Stick which could provide live monitoring irrigation which helps to rot the plant roots and makes the root
information of the environmental sensors (temperature and soil more fungal prone [76]. Also, if the water level is insufficient,
moisture). The stick is placed at the farms to obtain live feeds plants may die out due to the insufficient nutrients and oxygen
and synchronized with smart devices (phones and/or tablets). levels. Hence, monitoring the soil moisture level is a crucial
The stick comprises of Arduino Mega 2560, BreadBoard part of the agriculture industry.
BB400, ESP8266 Wi-Fi Module, 11.2 V Li-ion battery, solar For illustration, a block diagram of a soil moisture mon-
plate, soil moisture sensor and DS18B20 temperature sen- itoring system is presented in Fig. 5 [75] developed a low
sor. The results from this device showed 99% accuracy [66]. cost soil moisture monitoring system which can inform the
Table III summarizes the sensors that were used to monitor farmers via email and short messages. The architecture is
the climatic conditions inside the greenhouse farms. designed with NodeMCU V1.0 modules consisting of ESP826
he studies conducted for the monitoring of climate mainly micro-controller. The Losant platform (MQTT) is used in this
based on real-time monitoring of the environmental parame- architecture to visualize all the data from the sensors in real-
ters, employing the low cost WSN solutions and the integrated time and from any place [75]. Adamo et al. [77] proposed an
embedded technology. Nevertheless, no predictive analysis acoustic algorithm to measure the soil moisture in real-time.
models have been proposed yet for the analyses of the envi- Similar research was carried by [78]. The main aim of the
ronmental parameters. Thus, a subsequent decision making study was to develop an efficient and cost-effective soil mois-
model for predictive analysis models should also be proposed ture monitoring system so that small-scale and large scale
to understand what are the actions that are supposed to take by farmers can use it. The system is developed in an Arduino
the farmers and at what time they should take the appropriate board. A neural network is used for automatic correction of
actions. the system so that the required soil moisture is maintained.
The preliminary results from the system showed a promis-
ing performance [78]. Zhang et al. [72] proposed a combined
B. Monitoring of Soil Conditions system for monitoring soil moisture, soil temperature, air tem-
Soil moisture is an essential element in all types of farms perature and humidity by interfacing with JN5139 node and
(traditional or greenhouse) [75]. This is because plants or crops IEEE802.15.4 or ZigBee protocol for citrus growing. HA2001

Authorized licensed use limited to: The University of British Columbia Library. Downloaded on September 23,2020 at 18:46:50 UTC from IEEE Xplore. Restrictions apply.
 202 IEEE JOURNAL OF RADIO FREQUENCY IDENTIFICATION, VOL. 4, NO. 3, SEPTEMBER 2020

and HA2002 sensors were used for soil moisture and tem-
perature measurement respectively, while the SHT17 sensor
was employed to monitor air temperature and humidity [72].
Apart from those sensors discussed above, YL69 is one of the
popular soil moisture sensors that is used in research [73].
Monitoring the soil moisture is a challenging task as
high accuracy is required. It was found that even a slight
deviation from the target moisture value could cause either
over-watering or under-watering. For monitoring soil mois-
ture level, precise irrigation models need to be developed
as soon as possible. A programmable user interface devices
which can measure against a threshold can be designed and
interfaced with the microcontroller to receive the signals Fig. 6. Plant growth monitoring with Kinetic sensor.
through multiplexing [76]. Also, in-depth knowledge of crop
features, such as the rooting period, indeterminate growth [78],
harvesting period and plant cutting period should be known Hadabas et al. [81] proposed a similar study by using
by the farmers so that the soil moisture can be controlled Raspberry Pi 3 interfaced with IoT-cloud gateway to develop
accordingly. a low cost plant growth monitoring system. The architecture
also uses RGB camera sensors and they are placed at the
C. Monitoring of Plant Growth top of the plants. The designed platform exhibited satisfactory
results in determining the plant conditions (growth, size and
In order to improve the crop yield and production, color) [81]. Slamet et al. [82] proposed a cost-effective opto-
monitoring of the plant growth is essential. Recently, electronic sensor to monitor the growth of guava fruits. The
Khan and Hussain [79] proposed a novel sensing architecture optoelectronic sensor is built on a reflective tape like struc-
which could monitor the climatic conditions inside the green- ture which is enclosed around the fruit. The sensor collects
house farms. The novel idea was that instead of placing the the real-time data of the fruit growth and transmits it to the
sensors (temperature, humidity, soil moisture sensors) close to cloud. The estimated error for the radial growth of the fruit
the root of the plants, the sensors should be placed on the leaf was less than 2 mm, proving that it can be an effective sensor
of the plant. In this way, the exact values or the measurement to monitor the plant/ fruit growth [82].
for the plants can be obtained. Therefore, an ultra lightweight There have been substantial studies on the plant growth, par-
sensory platform is developed by Khan and Hussain [79] ticularly the accurate height measurement of the plants. Low
which is transparent and integrated with light, humidity and cost sensors have been explored for the real-time plant/fruit
temperature sensors, batteries and electronics. The sensory growth monitoring. Nevertheless, the major challenge of the
platform weighs about 0.44g which makes it suitable to place monitoring system of the plant growth is that the system
on the leaf of the plants. Besides, their research developed requires extensive calibration and the calibration varies from
another flexible type of strain sensor which can monitor the one hardware to another [73]. Therefore, a complete versa-
growth of the plants at the micrometer level. tile platform with effective and efficient sensors should be
The strain sensors are equipped with batteries enveloped developed by considering all the potential situations of the
inside a biodegradable paper and deployed through drones in plant growth (for example, direction or angle of plant growth).
larger greenhouse farms. The drones itself can take readings
at regular intervals from the wrapped strain sensors which can
provide the humidity, temperature, and moisture of the farms. D. Monitoring of Plant Diseases
The sensor was tested for plant growth of bamboo shoots and Although greenhouse crops or plants are grown in the con-
the results were very satisfactory [79]. Okayasu et al. [80] also trolled environment, the plants are still prone to pests and
presented a cost-effective smart sensing architecture which can diseases. Therefore, to maintain the high production yield
monitor the field conditions as well as can monitor the plant of the crops, farmers should monitor the plant diseases and
growth. The field monitoring system incorporates humidity, air make decisions to prevent the diseases as soon as possible.
temperature, light-intensity, CO2 and soil moisture level sen- Patil and Kale [83] developed an integrated system which
sors with a micro controller and it is powered with solar panels. takes all the necessary data from the temperature, humid-
However, for monitoring the plant growth, a typical RGB-D ity, light and pH sensors and aids in predicting the pest and
sensor (Kinetic sensor, a product of Microsoft, Technologies, disease of the plants with Ubi-Sense Mote (M) and commu-
as illustrated in Fig. 6) which integrates both the RGB cam- nicates via an IEEE 802.15.4 to the farmers with SMS. The
era and an IR camera, is used. The distance recordings are disease identification is also done from the images that are col-
taken from Time of Flight (ToF) technology. The test was lected from camera modules for the tomato fields. The study
conducted for a leafy type of plant, named as “Komatsuna” shows that farmers were benefited from this type of monitoring
and the sensors measured 2.5D image of the plants. The plant system [83].
features such as color of the leaf, height, shape and size were Materne and Inoue [84] developed an early disease and pest
all measured through the Kinetic sensors [80]. detection system by using the eight environmental parameters

Authorized licensed use limited to: The University of British Columbia Library. Downloaded on September 23,2020 at 18:46:50 UTC from IEEE Xplore. Restrictions apply.
 RAYHANA et al.: INTERNET OF THINGS EMPOWERED SMART GREENHOUSE FARMING 203

TABLE IV
S UMMARY OF C OST FOR THE D EVELOPED S YSTEM FOR
S MART G REENHOUSE

Fig. 7. Monitoring plant diseases with cameras. The image has been used
from [85] and reproduced with permission.

from the sensor data. The system uses Linear Regression,

Logistic Regression, KNN cluster algorithm and Random
Forest to predict the break out of the diseases inside the
farm [84]. Jumat et al. [85] also developed a low cost smart harsh, such as extremely dry, extremely humid, extremely
disease detection system which can identify the early break- high solar radiation and etc. The harsh condition might
outs for the greenhouse farms by using Raspberry Pi 3 and RF damage/shorten the life of sensory nodes including the battery.
as a classifier (example is shown on Fig. 7) [86] developed a Very few studies have been conducted in this area.
smart disease detection system for rice fields. Drones and cam-
eras are used to capture the image of the plants and images V. DATA T RANSMISSION AND P ROCESSING
are directly transferred through GPS sensors via the Wi-Fi
The data transmission system is one of the fundamental
communication protocol. Images are processed with machine
components of the IoT technologies used in greenhouse farm-
learning and segmentation algorithms to predict the possi-
ing. Like many other industries, greenhouse farming is heavily
bilities of disease development. The results from this study
dependent upon the real-time information of the farms as
showed that location and types of the disease that the plant
delays in data transmission might rupture the entire farm.
is going to develop can be pre-identified and an alert can be
In recent years, wireless sensor networks (WSN) have
sent to the farmers to take necessary actions [86].
gained popularity over wired transmission technologies, such
For monitoring the plant diseases, early detection is very
as field bus [54]. This is because WSN has efficient protocols,
crucial. Many studies regarding low cost early detection of
is easy to maintain and relatively cost-effective than cable
plant disease system has been done with the help of predictive
based technologies. The following section provides detail
analytics or neural network algorithms. Image processing has
reviews on state-of-the-art data transmission technologies that
also been used in some of the studies. However, none of the
are used in the agriculture industries.
studies included the meteorological information [84], which is
a key factor of the plant diseases or pest attack. Also, no stud-
ies regarding the disease classification [86] of the plants were A. Wireless Sensor Network (WSN)
conducted. These studies could help the farmers in decision WSN comprises of the multiple number of devices which
making process and also make the decision making tool to be is often known as the sensor nodes. The sensor node consists
robust. of microcontroller, batteries, sensors (temperature, humidity
In summary, sensors have found wide applications inside and so on), data transmission unit, and small-scale compu-
the smart greenhouse. One of the main issues facing all tational devices [92]. Fig. 9 shows a general block diagram
types of sensors is the cost. The current technology only of a sensor node. These sensor nodes are then dispersed into
allows the implementation of sensors at a global level. It different areas to accumulate the data and perform simple post-
is impossible to install sensors for the monitoring of each processing in accordance with the integrated decision making
crop, or every part of the big plant. Several kinds of research framework. The post-processed data are then sent to the access
have been undertaken to make the IoT sensor based system point (AP) which aids the remote monitoring of the region of
efficient and cost-effective. Minh et al. [38] proposed a interest [92]. Fig. 8 shows a typical deployment of WSN in
cost effective architecture with lightweight communication greenhouse farms.
system, which is believed to be affordable for both small- WSN can be divided into two main categories: Wireless
scale and large-scale farmers. Heble et al. [70] proposed a underground sensor networks (WUSN) and terrestrial WSN
low cost system with low powered network with high battery (TWSN) [93]. In smart greenhouse farms, WUSN is placed
life. Reference [74] reports an affordable large-scale WSN inside the soil and the underground sensor node relays the
sensor network for smart greenhouse farms. For the plant information through a gateway node. This information can be
growth and environment monitoring, [80] and [87] designed transferred to the Internet and stored in remote databases to
a low cost monitoring device for plants inside the green- inform the farmers through smartphones [94]. Nevertheless,
houses. For predicting the yield of the crops, [67] proposed an the downside of the WSUN is that it can only be used for
affordable real-time integration of machine learning with sen- small distance communication and the large quantity of nodes
sors data summarizes several low cost solutions published in are necessary [93]. On the other hand, TWSN is the sensor
Table IV. nodes that are planted on the ground surface to collect the
Hardware reliability is another big issue facing the imple- information regarding temperature, soil moisture, pH, humid-
mentation of the sensing system inside greenhouses farming. ity and CO2 level of the farms and etc. TWSN usually are
The environment conditions inside a greenhouse could be very a type of MEMS sensors and the nodes communicate with

Authorized licensed use limited to: The University of British Columbia Library. Downloaded on September 23,2020 at 18:46:50 UTC from IEEE Xplore. Restrictions apply.
 204 IEEE JOURNAL OF RADIO FREQUENCY IDENTIFICATION, VOL. 4, NO. 3, SEPTEMBER 2020

Fig. 8. Deployment of Wireless Sensor Nodes (WSNs) inside a greenhouse farm.

three main topologies for the network such as, Star, Mesh and
Cluster Tree. A ZigBee network is created when a coordinator
chooses an available network and a channel [98]. Once the
channel is created, the router and the end device start to com-
municate and transfer the data to all the connected devices
from the same address. However, the end devices are always
battery powered but they are designed to operate in low power
modes [99].
Fig. 9. A typical block diagram of a sensor node [96].
3) BLE: Bluetooth Low Energy (BLE) is used in the IoT
integrated devices as it can operate in short range (10 m) and
lower bandwidth [100]. The benefits of using BLE is that it
themselves to make intelligent decisions on the basis of the consumes less power and has a lower initialization time with
sensed data [93]. extensive support for star topology sensor nodes. BLE is often
In the past few years, the IoT technologies have evolved used to collect sensor data (for example, soil temperature and
a lot and so does the WSN protocols. The protocols are moisture) via mobile phones [101]. However, the downside
mainly developed in terms of cost, battery life, bandwidth and of BLE is that the connectivity and data transmission secu-
data rate [95]. The following section briefly describes some rity is low. Also, currently, it can communicate only with two
commonly used WSN protocols (summarized in Table V). devices [100].
1) IEEE 802.15.4: The IEEE 802.15.4 standard WSN pro- 4) 6LoWPAN: This is an IPv6 protocol which can sup-
tocol is one of the most used communication protocol in smart port star and mesh network topologies with IEEE 802.15.4
greenhouse farms. It uses the Medium Access Control (MAC) and has lower bandwidth. 6LoWPAN also consumes lower
layers and the physical (PHY) layers of the Low-Rate Wireless power [95]. It has been used to monitor the soil vari-
Personal Network (LR-WPAN). LR-WPAN is a cost-effective ables and climatic (moisture, temperature) condition of the
network which consumes very low power and has low pro- greenhouse plants [102]. Suryady et al. [103] presented a
cessing requisite. Additionally, LR-WPAN is easy to install performance based evaluation study on the greenhouse plants
and it can transfer data very efficiently within a short range with 6LowPAN. The study shows that there are some packet
of operation. It has a good battery span and it is relatively transmission delays between the sensor node to the gate-
cheaper than all other communication protocols [97]. In smart way and this is because the packets were routed by the
greenhouse farms, IEEE 802.15.4 protocol is often integrated gateway [103].
with radio modules, such as TelosB (TPR2400 which oper- 5) Cellular: Cellular protocol is suitable for those devices
ates with 2AA alkaline battery), ATMega256RFR2 (built-in which require a higher transmission rate of data. It uses 3G, 4G
temperature and RGB sensor with a rechargeable batteries) and GSM modules to ensure the high speed reliable transmis-
and OpenMote-CC2538 (ARM Cortex-M3 core and XBEE sion of data through the Internet. This technology is good to
boards) [92]. be deployed for an underground wireless network. However, it
2) ZigBee: Similar to IEEE 802.15.4, ZigBee uses LR- requires more battery power and the operation cost is slightly
WPAN wireless system which ensures reliable and secured higher [104].
data transmission and communication between the networks. 6) LoRa WAN: LoRa WAN is based on IEEE 802.11 ah
In smart greenhouse farms, ZigBee is now considered as a standard and it uses LoRa network, which allows to operate
global standard for WSN communication protocol. ZigBee for long distance communications. It is based on Peer-to-Peer
devices consist of a coordinator, a router/ modem which has (P2P) and star topologies. LoRa WAN also consumes very

Authorized licensed use limited to: The University of British Columbia Library. Downloaded on September 23,2020 at 18:46:50 UTC from IEEE Xplore. Restrictions apply.
 RAYHANA et al.: INTERNET OF THINGS EMPOWERED SMART GREENHOUSE FARMING 205

TABLE V
T YPICAL R EPORTED WSN C OMMUNICATION P ROTOCOLS T HAT W ERE U SED IN S MART G REENHOUSES

low power and it can cover a wide range of area when it is retain the video information by using edge nodes to cache
deployed in the greenhouse farms [105], [106]. the content [116]. Edge computing is one extension of this
7) RFID: Radio Frequency Identification (RFID) architec- paradigm which supports the cloud computing system. It
ture aids the machine to affirm and record all the sensor supports data storage and has considerable integration of com-
information by using radio waves [107]. The architecture com- puting technology which is named cloudlets [117] and fog
prises of host, tags and readers [108]. The reader collects all nodes (microdata center) [118]. Fog nodes are often deployed
the information through radio waves and communicates with close to the sensors or cell phone devices [119].
each other [109]. Three types of tags are used, i.e., active, Chen et al. [120] proposed fog computing “ThriftyEdge”,
passive and battery assisted passive. The active tag consumes also known as mobile edge computing network, as an alter-
more power and is battery powered, which can be detected native to the traditional cloud computing network for smart
far away. The passive tag only consumes the power harvested greenhouse farms. The network uses sorting topologies which
from the reader and can only be read in a short distance. RFID employ the edge resources with minimum occupancy and
protocol has been explored for the monitoring of the crop fulfills the requirements to maintain the quality of standard
conditions as well as the environmental conditions [108]. QoS for the data transmission. The results from the defined
8) Wi-Fi: Wi-Fi is the most commonly used WSN proto- network were superior to all other proposed networks and can
col and it is based on IEEE 802.11, 802.11a, 802.11b, 802.11g help the end user (farmers) to obtain all information very
and 802.11n standards [69]. Wi-Fi uses Wireless Local Area efficiently [120]. Zamora-Izquierdo et al. [121] proposed a
Network (WLAN) connectivity and it has been used to moni- versatile three-tier platform that mitigates the software and
tor plants, climatic conditions [110], and irrigation control in hardware compatibility issues through “Network Function
smart greenhouse farms. Visualization” (NFV) to substantiate decision-based operations
The smart greenhouse means that there must be uninter- in the edge layer of the platform. Message Queue Telemetry
rupted and efficient connection between the devices under Transport (MQTT) was used as a communication medium to
adverse environmental conditions and the Internet. However, communicate with other subsystems and FIWARE is used as a
cases were reported that the WSN was damaged due to cloud layer to store all the data and aid the farmers to monitor
the environmental conditions [111], [112], [113] within the and control the farms in real-time [121].
greenhouse farms. Moreover, the plant growth can sometimes Fan and Gao [122] proposed an intelligent IoT based mobile
develop multi-path distribution problem which introduces edge computing (MEC) water monitoring system by using
noise in the background and decreases the data transmission the real-time sensor data from the crops (as illustrated in
speed [114]. Fig. 10). It can speed up the data transmission process by
Recently reported in [54], ad hoc wireless network can be establishing an efficient feedback mechanism between the sen-
used to the situations where the nodes are exposed to harsh sors and the devices. Generally, a traditional water monitoring
conditions. However, ad hoc wireless network uses LPWAN system consists of 1) sensor unit 2) data transfer unit 3) router
which has exorbitant deployment and gateway cost [54]. Smart unit 4) network management unit 5) open-ability unit 6) edge
greenhouses also have interoperability issues as there are server unit 7) network server unit [122].
too many standards of communication protocols and data 2) Cloud Computing: Cloud computing is an Internet cloud
modeling interface. This generate problems in the connectivity reservoir which uses centralized or distributed computing
(poor coverage of 3G/4G) of the devices which is another key technology. Cloud computing can apply parallel and dis-
issue to solve in smart farming [33], [69]. Although, SigFox tributed computing simultaneously or individually and it can
and LoRa network technologies have leveraged this issue for be developed on the data center virtually or physically [130].
small dataset but there is no solution yet to deal with large Cloud computing provides different types of services such as
dataset [115]. infrastructure, hardware and software services to IoT enabled
devices. The advantage of cloud computing is that it can facil-
B. Edge and Cloud Computing itate dynamic monitoring of the virtualized computing system
1) Edge Computing: In the year of 1990, Akamai and store the data at a very low expense [131], [132]. This
proposed “Content delivery networks” (CDN) to improve the aids the farmers to use the video, image and short mes-
performance of the Web which can use less bandwidth to sage services to obtain the information about the greenhouse

Authorized licensed use limited to: The University of British Columbia Library. Downloaded on September 23,2020 at 18:46:50 UTC from IEEE Xplore. Restrictions apply.
 206 IEEE JOURNAL OF RADIO FREQUENCY IDENTIFICATION, VOL. 4, NO. 3, SEPTEMBER 2020

entire system. As there is the lack of homogeneity in the

network coverages, verification and privacy protection also
varies. Two of the most commonly data security problems
arose due to RFID tags and WSN [54]. RFID tags can be
rewritten from the Internet through the readers this leads to the
possibility of invasion in the data protection system. For WSN,
the sensor nodes are open as it is deployed for unsupervised
monitoring purposes. This makes it susceptible to the ille-
gal users and external intrusion, such as, node capture attack,
information tapping, replay attack and so on [139], [140]. Data
processing capacity is another challenge: There are several
devices available in the market which claim to offer superior
data processing power. However, how to do this with a rea-
sonable investment. This affects the medium and small-scale
smart farmers as they can not afford expensive powerful data
processing devices to their farms [33].

C. Decision Making
Fig. 10. MEC based water monitoring system. In tradition farming, decision making is generally experi-
ence based. Those decisions include the crop planting and har-
vesting time, fertilizer applying, irrigation, pesticide applying
farms at a minimum cost [133]. Also, cloud computing can and etc. As all those decisions are subject to the environmen-
handle large scale intelligent computing systems at the same tal conditions, particularly the weather and soil conditions,
time [41]. it is difficult to develop an expert system to facilitate the
In smart greenhouse farms, cloud computing uses decision making for field farming. Nevertheless, in a smart
three different model layers: Infrastructure, Platform and greenhouse, the soil condition and local environment can
Software [134]. For instance, Rupanagudi et al. [41] used be precisely monitored and controlled, also the plant con-
Intel core-i3 32 bit processor as the Infrastructure (hardware), dition can be monitored in its whole growing cycle. This
Microsoft Azure as the platform and Java application as the makes it possible to develop an expert system to grow crops
software to identify diseases and unripe tomatoes by using inside a smart greenhouse, just like setting up a cooking
real-time video footage of the Web camera. Hsu et al. [135] program.
developed a cloud integration platform to perform data anal- Machine learning or artificial intelligence [141] is the
ysis on large data set and to automate the farm monitoring approach being taken to process the sensing data collection
process even if the information of the network is limited. The and facilitate the decision making process. Some examples
developed network could also increase the data processing include Naive Bayes, Gaussian Mixture Model (GMM), artifi-
speed which can be used in the pest management of the smart cial neural networks, decision trees, k-nearest neighbor, deep
greenhouse farms [135]. learning [142], fuzzy logic’s. Machine learning can make
Gayatri et al. [136] proposed a real-time smart monitoring predictions and can find the correlation between variables
platform to improve the yield of the crops by using a cloud- to provide intelligent decisions or solutions for the smart
based wireless network. The embedded wireless network could greenhouse farmings. This technology has been already used
operate very efficiently by transferring all the sensor data to the to plan irrigation systems, identify and predict plant dis-
cloud which aided the farmers to use energy efficiently [136]. eases [143], [144]. In [145], yellow sticky papers were used
Abdullah et al. [137] developed software called “Agri-System” to capture the pests. The pests were then counted with the
with the integration of cloud computing, IoT and sensor data. aid of machine learning and image processing algorithms.
The system aims to maintain the sufficiency of the smart green The machine learning algorithm could also identify a variable
house farms by ensuring all the environmental parameters are (light intensity) which had a direct impact of the pest growth
under control. A fuzzy controller is used to minimize the com- on the greenhouse cabbage farms [145]. Jumat et al. [85]
plexity [137]. Channe et al. [138] proposed a mobile based proposed a machine learning based prototype to identify and
application, “Agrocloud” which is an integration of big data, classify “Septoria” plant diseases in smart greenhouse farms.
IoT, sensor and cloud technology. The Beagle Black Bone of The proposed prototype uses Raspberry Pi 3 module and
IoT helps to obtain the environmental and soil parameters and Random Forest as a classifier to detect the plant diseases
transfer it to the Agrocloud to store the data. Agrocloud also efficiently [85].
provides agro-vendors, marketing and government schemes to Machine learning was also used by Khirade and A. Patil to
support farmers during and after the harvest time [138]. detect plant decease [146] image data was fed to the machine
Data and system security have become a vital challenge and artificial neural networks were used to classify the disease
in recent years. Network or cyber attacks make all the data of the plants. In addition, a machine learning IoT based com-
very vulnerable and it can completely be wiped out from the mercial system is developed by Palli et al. [147] to automate

Authorized licensed use limited to: The University of British Columbia Library. Downloaded on September 23,2020 at 18:46:50 UTC from IEEE Xplore. Restrictions apply.
 RAYHANA et al.: INTERNET OF THINGS EMPOWERED SMART GREENHOUSE FARMING 207

the farm monitoring, fertilizer spraying and notify the farmers smart greenhouses provides a viable solution for
when pests are attacking the plants. this. Reference [154] reported research in this area.
In [41], the proposed cloud technologies and the designed • The development of an “expert system” to facilitate farm-
algorithms in Java determined the amount of insecticides that ers in the decision making process, this needs to integrate
needed to be sprayed on the plants. This can avert the excessive the environment data and the crop growing data using
usage of insecticides and can protect the plants, the farmer artificial intelligence.
and consumer’s health and the environment. Apart from this,
several fuzzy logic based models were developed to control ACKNOWLEDGMENT
the microclimate inside the greenhouse farms [148], [149]. In National Research Council of Canada (NRC) is acknowl-
addition, models have been developed to utilize the greenhouse edged for the funding assistance and support to conduct
infrastructures (water pumps, actuators, lighting and etc.) in a this research. This research was conducted by the Advanced
most efficient way [42], [130], [150] as well as estimate the Electronics and Photonics (AEP) research group, NRC,
crop yields, for example TOMGRO [136], TOMSIM [151] Ottawa, Canada.
and TOMPOUSSE [152], [153] have been developed in recent
years to estimate the tomato yield from the smart greenhouse
R EFERENCES
farming.
The decision making part (also called an expert system) of [1] T. Folnovic, Loss of Arable Land Threaten World Food Supplies, Agrivi,
London, U.K., 2011. [Online]. Available: https://blog.agrivi.com
the IoT architecture for smart greenhouse farming is currently [2] The Future of Food and Agriculture–Trends and Challenges, Food Agr.
in its early development stage. Significant efforts are required Org., Rome, Italy, 2017.
to make them applicable or become a “menu” type system. [3] N. Panwar, S. Kaushik, and S. Kothari, “Solar greenhouse an option for
renewable and sustainable farming,” Renewable Sustain. Energy Rev.,
Several factors greatly contribute to this, including, vol. 15, no. 8, pp. 3934–3945, 2011.
• Lacking of proper IoT technology to monitor crop [4] G. Tiwari, Greenhouse Technology for Controlled Environment.
growth, such as height, growth rate, maturity and quality. Harrow, U.K.: Alpha Sci. Int. Ltd., 2003.
[5] Historical Background of Greenhouses, Emerald Agr. Technol.,
• Lacking of local environment condition parameters sur- Kolhapur, India. Accessed: Dec. 19, 2019. [Online]. Available:
rounding the crops. Current sensing technology only http://www.emerald-agri.com
monitors the global environment inside the greenhouse [6] S. Vatari, A. Bakshi, and T. Thakur, “Green house by using IoT and
cloud computing,” in Proc. IEEE Int. Conf. Recent Trends Electron. Inf.
which can be very different from the local ones. Commun. Technol. (RTEICT), Bangalore, India, 2016, pp. 246–250.
• Lacking of the correlation between the crop growth data [7] J. Xu, Y. Li, R. Wang, W. Liu, and P. Zhou, “Experimental performance
and the environment data. of evaporative cooling pad systems in greenhouses in humid subtropical
climates,” Appl. Energy, vol. 138, pp. 291–301, Jan. 2015.
[8] S. Pallavi, J. D. Mallapur, and K. Y. Bendigeri, “Remote sensing and
controlling of greenhouse agriculture parameters based on IoT,” in
VI. C ONCLUSION AND F UTURE P ERSPECTIVES Proc. IEEE Int. Conf. Big Data IoT Data Sci. (BID), Pune, India,
2017, pp. 44–48.
As discussed in this review paper, although IoT empow- [9] M. Tapiwa, “The greenhouse concept as used in agriculture,
ered smart greenhouse farming offers countless current and its advantages and disadvantages,” ACADEMIA, San Francisco,
potential benefits, the technology has also advanced remark- CA, USA. Accessed: Dec. 19, 2019. [Online]. Available:
https://www.academia.edu
ably in recent years. It is believed that major efforts from [10] L. Yang, B. Huang, W. Hu, Y. Chen, M. Mao, and L. Yao, “The
the researchers, engineers and farmers are needed in order to impact of greenhouse vegetable farming duration and soil types on phy-
improve the crop yield at lower cost, particularly if we envision toavailability of heavy metals and their health risk in eastern China,”
Chemosphere, vol. 103, pp. 121–130, May 2014.
the future smart greenhouse farming as a “menu selection” [11] A. Pardossi, F. Tognoni, and L. Incrocci, “Mediterranean greenhouse
process similar to the cooking. Below are the summary of the technology,” Chronica Horticulturae, vol. 44, no. 2, pp. 28–34, 2004.
major challenges facing the IoT empowered smart greenhouse [12] C. A. van der Spijk. Greenhouse Technology Adoption Among Small
and Medium-Scale Tomato Farmers in Kenya. Accessed: Dec. 20, 2019.
farming: [Online]. Available: http://www.3r-kenya.org
• Low cost and reliable solution for the monitoring of plant [13] R. Dagar, S. Som, and S. K. Khatri, “Smart farming—IoT in agricul-
growth, including all the plant traits (height, leaf decease, ture,” in Proc. IEEE Int. Conf. Inventive Res. Comput. Appl. (ICIRCA),
Coimbatore, India, 2018, pp. 1052–1056.
fruit/grain/leaf dimension and etc.) and quality (protein, [14] T. Wiangtong and P. Sirisuk, “IoT-based versatile platform for precision
sugar and etc.). farming,” in Proc. IEEE 18th Int. Symp. Commun. Inf. Technol. (ISCIT),
• Low cost and reliable sensor solution for the monitoring Bangkok, Thailand, 2018, pp. 438–441.
[15] T. J. Jeaunita, V. Sarasvathi, M. Harsha, B. Bhavani, and T. Kavyashree,
of local climate surrounding individual crop or most part “An automated greenhouse system using agricultural Internet of Things
of a large plant. for better crop yield,” in Proc. Smart Cities Symp., 2018, pp. 129–134.
• Low cost and reliable sensor solution for the monitoring [16] Good Agricultural Practices for Greenhouse Vegetable Crops-
Principles for Mediterranean Climate Areas, FAO, Rome, Italy, 2013.
of soil conditions (chemical compositions). [Online]. Available: http://www.fao.org
• A unified data transmission protocol. The inter-operability [17] Greenhouse Farming in India (Beginner Guide 2019),
and generalized standards for data annotation, visu- AgricultureGuruji, Donawade, India, Mar. 2018. [Online]. Available:
https://agricultureguruji.com
alization and decision should be considered. This [18] Grower’s Guide: What are Cold Frame Greenhouses?
would require the coordination between major industry ShelterLogic, Watertown, CT, USA, 2018. [Online]. Available:
associations. https://www.shelterlogic.ca.
[19] U. Çakır and E. Şahin, “Using solar greenhouses in cold climates and
• Secure data transmission and storage. Integrating evaluating optimum type according to sizing, position and location: A
blockchain technology with the IoT empowered case study,” Comput. Electron. Agr., vol. 117, pp. 245–257, Sep. 2015.

Authorized licensed use limited to: The University of British Columbia Library. Downloaded on September 23,2020 at 18:46:50 UTC from IEEE Xplore. Restrictions apply.
 208 IEEE JOURNAL OF RADIO FREQUENCY IDENTIFICATION, VOL. 4, NO. 3, SEPTEMBER 2020

[20] 10 of the Best Diy Greenhouses and Cold Frames for Your Backyard. [43] D. S. Domingues, H. W. Takahashi, C. A. Camara, and S. L. Nixdorf,
Gardener’s Path, Feb. 2017. Accessed: Dec. 20, 2019. [Online]. “Automated system developed to control PH and concentration of
Available: https://gardenerspath.com nutrient solution evaluated in hydroponic lettuce production,” Comput.
[21] Cold Frame Gardening, Vegetable Gardener, Aug. 1996. Accessed: Electron. Agr., vol. 84, pp. 53–61, Jun. 2012.
Dec. 20, 2019. [Online]. Available: http://www.vegetablegardener.com [44] D. Saraswathi, P. Manibharathy, R. Gokulnath, E. Sureshkumar, and
[22] Difference Between Shade and Green Houses? Clear Up the Confusion K. Karthikeyan, “Automation of hydroponics green house farming
Here, C-MAC, Girraween NSW, Australia, Jul. 2019. [Online]. using IoT,” in Proc. IEEE Int. Conf. Syst. Comput. Autom. Netw.
Available: https://www.cmac.com.au (ICSCA), Pondicherry, India, 2018, pp. 1–4.
[23] Shade Houses & Material, Grower’s Supply, South Windsor, [45] K. Ferentinos and L. Albright, “Predictive neural network modeling
CT, USA. Accessed: Dec. 23, 2019. [Online]. Available: of ph and electrical conductivity in deep–trough hydroponics,” Trans.
https://www.growerssupply.com ASAE, vol. 45, no. 6, pp. 2007–2015, 2002.
[24] J. Xie et al., “Gobi agriculture: An innovative farming system that [46] M. Mehra, S. Saxena, S. Sankaranarayanan, R. J. Tom, and
increases energy and water use efficiencies. A review,” Agronomy M. Veeramanikandan, “IoT based hydroponics system using deep
Sustain. Develop., vol. 38, no. 6, p. 62, 2018. neural networks,” Comput. Electron. Agr., vol. 155, pp. 473–486,
[25] F. Tognoni, A. Pardossi, and G. Serra, “Strategies to match greenhouses Dec. 2018.
to crop production,” in Proc. Int. Symp. Growing Media Hydroponics [47] B. Paulchamy, N. Balaji, S. D. Pravatha, P. H. Kumar, and
481, 1997, pp. 451–462. T. J. Frederick, “A novel approach for automating & analyzing hydro-
[26] J. Kipp, “Optimal climate regions in mexico for greenhouse ponic farms using Internet of Things,” Int. J. Sci. Res. Comput. Sci.
crop production,” Wageningen UR Greenhouse Horticul., Bleiswijk, Eng. Inf. Technol., vol. 3, no. 3, pp. 1230–1234, 2018.
The Netherlands, Rep. GTB-1024, 2010. [48] J. Pitakphongmetha, N. Boonnam, S. Wongkoon, T. Horanont,
[27] S. Prathibha, A. Hongal, and M. Jyothi, “IoT based monitoring system D. Somkiadcharoen, and J. Prapakornpilai, “Internet of Things for
in smart agriculture,” in Proc. IEEE Int. Conf. Recent Adv. Electron. planting in smart farm hydroponics style,” in Proc. IEEE Int. Comput.
Commun. Technol. (ICRAECT), Bangalore, India, 2017, pp. 81–84. Sci. Eng. Conf. (ICSEC), Chiang Mai, Thailand, 2016, pp. 1–5.
[28] Y. Tavakoli and S. Khoshkam, “The impact of organic fertilizers on [49] R. Vidhya and K. Valarmathi, “Survey on automatic monitoring of
production of organic greenhouse cucumber,” Mediterr. J. Soc. Sci., hydroponics farms using IoT,” in Proc. IEEE 3rd Int. Conf. Commun.
vol. 4, no. 14, p. 249, 2013. Electron. Syst. (ICCES), Coimbatore, India, 2018, pp. 125–128.
[29] P. T. Kidd, “The role of the Internet of Things in enabling sustainable [50] S. Bhowmick, B. Biswas, M. Biswas, A. Dey, S. Roy, and S. K. Sarkar,
agriculture in Europe,” Int. J. RF Technol., vol. 3, no. 1, pp. 67–83, “Application of IoT-enabled smart agriculture in vertical farming,”
2012. in Advances in Communication, Devices and Networking. Singapore:
[30] IoT Applications in Agriculture, IoT for all, Jan. 2020. Accessed: Springer, 2019, pp. 521–528.
Dec. 23, 2019. [Online]. Available: https://www.iotforall.com [51] M. I. H. Bin Ismail, and N. M. Thamrin, “IoT implementation
[31] J. Muangprathub, N. Boonnam, S. Kajornkasirat, N. Lekbangpong, for indoor vertical farming watering system,” in Proc. IEEE Int.
A. Wanichsombat, and P. Nillaor, “IoT and agriculture data analysis for Conf. Elect. Electron. Syst. Eng. (ICEESE), Kanazawa, Japan, 2017,
smart farm,” Comput. Electron. Agr., vol. 156, pp. 467–474, Jan. 2019. pp. 89–94.
[32] A. Medela, B. Cendón, L. González, R. Crespo, and I. Nevares, “IoT [52] F. C. L. Belista, M. P. C. Go, L. L. Luceñara, C. J. G. Policarpio,
multiplatform networking to monitor and control wineries and vine- X. J. M. Tan, and R. G. Baldovino, “A smart aeroponic tailored for IoT
yards,” in Proc. Future Netw. Mobile Summit, Lisboa, Portugal, 2013, vertical agriculture using network connected modular environmental
pp. 1–10. chambers,” in Proc. IEEE 10th Int. Conf. Humanoid Nanotechnol. Inf.
[33] C. Brewster, I. Roussaki, N. Kalatzis, K. Doolin, and K. Ellis, “IoT in Technol. Commun. Control Environ. Manag. (HNICEM), Baguio City,
agriculture: Designing a europe-wide large-scale pilot,” IEEE Commun. Philippines, 2018, pp. 1–4.
Mag., vol. 55, no. 9, pp. 26–33, Sep. 2017. [53] A. Antonacci, F. Arduini, D. Moscone, G. Palleschi, and
[34] Z. Pang, Q. Chen, W. Han, and L. Zheng, “Value-centric design of V. Scognamiglio, “Nanostructured (bio)sensors for smart agriculture,”
the Internet-of-Things solution for food supply chain: Value creation, TrAC Trends Anal. Chem., vol. 98, pp. 95–103, Jan. 2018.
sensor portfolio and information fusion,” Inf. Syst. Front., vol. 17, no. 2, [54] X. Shi, X. An, Q. Zhao, H. Liu, L. Xia, X. Sun, and Y. Guo, “State-
pp. 289–319, 2015. of-the-art Internet of Things in protected agriculture,” Sensors, vol. 19,
[35] J. M. Talavera et al., “Review of IoT applications in agro-industrial and no. 8, p. 1833, 2019.
environmental fields,” Comput. Electron. Agr., vol. 142, pp. 283–297, [55] F. Li, X. Wang, X. Sun, Y. Guo, and W. Zhao, “A dual-signal
Nov. 2017. amplification strategy for kanamycin based on ordered mesoporous
[36] S. L. Li, Y. Han, G. Li, M. Zhang, L. Zhang, and Q. Ma, “Design carbon-chitosan/gold nanoparticles-streptavidin and ferrocene labelled
and implementation of agricultral greenhouse environmental monitor- DNA,” Analytica Chimica Acta, vol. 1033, pp. 185–192, Nov. 2018.
ing system based on Internet of Things,” in Proc. Appl. Mech. Mater., [56] Y. Jiao et al., “Homogeneous electrochemical aptasensor based on
vol. 121, 2012, pp. 2624–2629. a dual amplification strategy for sensitive detection of profenofos
[37] N. Kaewmard and S. Saiyod, “Sensor data collection and irrigation con- residues,” New J. Chem., vol. 42, no. 17, pp. 14642–14647, 2018.
trol on vegetable crop using smart phone and wireless sensor networks [57] Y. Zheng, Z. Liu, Y. Jing, J. Li, and H. Zhan, “An acetyl-
for smart farm,” in Proc. IEEE Conf. Wireless Sens. (ICWiSE), Subang, cholinesterase biosensor based on ionic liquid functionalized graphene–
Malaysia, 2014, pp. 106–112. gelatin-modified electrode for sensitive detection of pesticides,” Sens.
[38] Q. T. Minh et al., “A cost-effective smart farming system with knowl- Actuators B Chem., vol. 210, pp. 389–397, Apr. 2015.
edge base,” in Proc. 8th Int. Symp. Inf. Commun. Technol., 2017, [58] J. Bao et al., “3D graphene/copper oxide nano-flowers based acetyl-
pp. 309–316. cholinesterase biosensor for sensitive detection of organophosphate
[39] A. Mittal, S. Sarangi, S. Ramanath, P. V. Bhatt, R. Sharma, and pesticides,” Sens. Actuators B Chem., vol. 279, pp. 95–101, Jan. 2019.
P. Srinivasu, “IoT-based precision monitoring of horticultural crops— [59] F. Li, Y. Guo, X. Wang, and X. Sun, “Multiplexed aptasensor based
A case-study on cabbage and capsicum,” in Proc. IEEE Global on metal ions labels for simultaneous detection of multiple antibiotic
Humanitarian Technol. Conf. (GHTC), San Jose, CA, USA, 2018, residues in milk,” Biosens. Bioelectron., vol. 115, pp. 7–13, Sep. 2018.
pp. 1–7. [60] C. Negara, K.-U. Vieth, M. Lafontaine, and M. Freund, “Automatic
[40] J. Park, J.-H. Choi, Y.-J. Lee, and O. Min, “A layered features analysis fruit sorting by non-destructive determination of quality parameters
in smart farm environments,” in Proc. Int. Conf. Big Data Internet using visible-near infrared to improve wine quality: II. Regression
Thing, 2017, pp. 169–173. analysis,” NIR News, vol. 25, no. 1, pp. 4–6, 2014.
[41] S. R. Rupanagudi, B. Ranjani, P. Nagaraj, V. G. Bhat, and [61] G. Marques and R. Pitarma, “Agricultural environment monitoring
G. Thippeswamy, “A novel cloud computing based smart farming system using wireless sensor networks and IoT,” in Proc. IEEE 13th
system for early detection of borer insects in tomatoes,” in Proc. IEEE Iber. Conf. Inf. Syst. Technol. (CISTI), Caceres, Spain, 2018, pp. 1–6.
Int. Conf. Commun. Inf. Comput. Technol. (ICCICT), Mumbai, India, [62] G. H. Wu, F. Liu, J. X. Li, and W. Wang, “Environmental monitoring
2015, pp. 1–6. system designing: A Internet of Things approach,” in Proc. Appl. Mech.
[42] J. Aleotti, M. Amoretti, A. Nicoli, and S. Caselli, “A smart precision- Mater., vols. 644–650, 2014, pp. 3342–3345.
agriculture platform for linear irrigation systems,” in Proc. IEEE [63] I. Saraswati et al., “Applications of temperature and humidity moni-
26th Int. Conf. Softw. Telecommun. Comput. Netw. (SoftCOM), Split, toring system at aerophonic plants based on IoT,” in Proc. Matec Web
Croatia, 2018, pp. 1–6. Conf., vol. 218, 2018, Art. no. 03017.

Authorized licensed use limited to: The University of British Columbia Library. Downloaded on September 23,2020 at 18:46:50 UTC from IEEE Xplore. Restrictions apply.
 RAYHANA et al.: INTERNET OF THINGS EMPOWERED SMART GREENHOUSE FARMING 209

[64] M. Danita, B. Mathew, N. Shereen, N. Sharon, and J. J. Paul, “IoT [85] M. H. Jumat, M. S. Nazmudeen, and A. T. Wan, “Smart farm pro-
based automated greenhouse monitoring system,” in Proc. IEEE 2nd totype for plant disease detection, diagnosis & treatment using IoT
Int. Conf. Intell. Comput. Control Syst. (ICICCS), Madurai, India, 2018, device in a greenhouse,” in Proc. 7th Brunei Int. Conf. Eng. Technol.,
pp. 1933–1937. Bandar Seri Begawan, Brunei, 2018, p. 48.
[65] R. Pahuja, H. Verma, and M. Uddin, “A wireless sensor network for [86] N. Kitpo and M. Inoue, “Early rice disease detection and position
greenhouse climate control,” IEEE Pervasive Comput., vol. 12, no. 2, mapping system using drone and IoT architecture,” in Proc. IEEE
pp. 49–58, Apr.–Jun. 2013. 12th South East Asian Techn. Univ. Consortium (SEATUC), vol. 1.
[66] A. Nayyar and E. V. Puri, “Smart farming: IoT based smart Yogyakarta, Indonesia, 2018, pp. 1–5.
sensors agriculture stick for live temprature and moisture mon- [87] S. Tenzin, S. Siyang, T. Pobkrut, and T. Kerdcharoen, “Low cost
itoring using arduino cloud computing & solar technology,” in weather station for climate-smart agriculture,” in Proc. IEEE 9th
Proc. Int. Conf. Commun. Comput. Syst. (ICCCS-2016), 2016, Int. Conf. Knowl. Smart Technol. (KST), Chonburi, Thailand, 2017,
Art. no. 9781315364094-121. pp. 172–177.
[67] R. Varghese and S. Sharma, “Affordable smart farming using IoT and
[88] D. P. Rubanga, K. Hatanaka, and S. Shimada, “Development of a sim-
machine learning,” in Proc. IEEE 2nd Int. Conf. Intell. Comput. Control
plified smart agriculture system for small-scale greenhouse farming,”
Syst. (ICICCS), Madurai, India, 2018, pp. 645–650.
Sens. Mater., vol. 31, no. 3, pp. 831–843, 2019.
[68] P. Dedeepya, U. Srinija, M. G. Krishna, G. Sindhusha, and T. Gnanesh,
“Smart greenhouse farming based on IoT,” in Proc. IEEE 2nd Int. Conf. [89] T. A. Ali, V. Choksi, and M. Potdar, “Precision agriculture monitoring
Electron. Commun. Aerospace Technol. (ICECA), Coimbatore, India, system using green Internet of Things (G-IoT),” in Proc. IEEE 2nd
2018, pp. 1890–1893. Int. Conf. Trends Electron. Informat. (ICOEI), Tirunelveli, India, 2018,
[69] U. Shafi, R. Mumtaz, J. García-Nieto, S. A. Hassan, S. A. R. Zaidi, pp. 481–487.
and N. Iqbal, “Precision agriculture techniques and practices: From [90] N. Silva et al., “Low-cost IoT LoRa solutions for precision
considerations to applications,” Sensors, vol. 19, no. 17, p. 3796, agriculture monitoring practices,” in EPIA Conference on Artificial
2019. Intelligence. Cham, Switzerland: Springer, 2019, pp. 224–235.
[70] S. Heble, A. Kumar, K. V. D. Prasad, S. Samirana, P. Rajalakshmi, [91] N. K. Nawandar and V. R. Satpute, “IoT based low cost and intelligent
and U. B. Desai, “A low power IoT network for smart agriculture,” module for smart irrigation system,” Comput. Electron. Agr., vol. 162,
in Proc. IEEE 4th World Forum Internet Things (WF-IoT), Singapore, pp. 979–990, Jul. 2019.
2018, pp. 609–614. [92] K. V. de Oliveira, H. M. E. Castelli, S. J. Montebeller, and
[71] V. Ramachandran, R. Ramalakshmi, and S. Srinivasan, “An automated T. G. P. Avancini, “Wireless sensor network for smart agriculture using
irrigation system for smart agriculture using the Internet of Things,” zigbee protocol,” in Proc. IEEE 1st Summer School Smart Cities (S3C),
in Proc. IEEE 15th Int. Conf. Control Autom. Robot. Vis. (ICARCV), Natal, Brazil, 2017, pp. 61–66.
Singapore, 2018, pp. 210–215. [93] T. Ojha, S. Misra, and N. S. Raghuwanshi, “Wireless sensor networks
[72] X. Zhang, J. Zhang, L. Li, Y. Zhang, and G. Yang, “Monitoring cit- for agriculture: The state-of-the-art in practice and future challenges,”
rus soil moisture and nutrients using an IoT based system,” Sensors, Comput. Electron. Agr., vol. 118, pp. 66–84, Oct. 2015.
vol. 17, no. 3, p. 447, 2017. [94] G. Mois, S. Folea, and T. Sanislav, “Analysis of three IoT-based
[73] B. Yimwadsana, P. Chanthapeth, C. Lertthanyaphan, and wireless sensors for environmental monitoring,” IEEE Trans. Instrum.
A. Pornvechamnuay, “An IoT controlled system for plant growth,” Meas., vol. 66, no. 8, pp. 2056–2064, Aug. 2017.
in Proc. IEEE 7th ICT Int. Student Project Conf. (ICT-ISPC),
[95] S. Al-Sarawi, M. Anbar, K. Alieyan, and M. Alzubaidi, “Internet of
Nakhonpathom, Thailand, 2018, pp. 1–6.
Things (IoT) communication protocols: Review,” in Proc. IEEE 8th
[74] M. Kuroda, H. Ibayashi, and H. Mineno, “Affordable 400MHz long-
Int. Conf. Inf. Technol. (ICIT), Amman, Jordan, 2017, pp. 685–690.
haul sensor network for greenhouse horticulture,” in Proc. IEEE Int.
Conf. Inf. Netw. (ICOIN), 2015, pp. 19–24. [96] “Sensor node or sensor mote-basics of sensor node or mote,” RF-
[75] R. K. Kodali and A. Sahu, “An IoT based soil moisture monitoring Wireless World, 2012. Accessed: Jan. 28, 2020. [Online]. Available:
on losant platform,” in Proc. IEEE 2nd Int. Conf. Contemp. Comput. https://www.rfwireless-world.com
Informat. (IC3I), Noida, India, 2016, pp. 764–768. [97] J. T. Adams, “An introduction to IEEE STD 802.15.4,” in Proc. IEEE
[76] I. Al-Bahadly and J. Thompson, “Garden watering system based on Aerospace Conf., Big Sky, MT, USA, 2006, p. 8.
moisture sensing,” in Proc. IEEE 9th Int. Conf. Sens. Technol. (ICST), [98] M. Bassoli, V. Bianchi, I. De Munari, and P. Ciampolini, “An IoT
Auckland, New Zealand, 2015, pp. 263–268. approach for an AAL Wi-Fi-based monitoring system,” IEEE Trans.
[77] F. Adamo, G. Andria, F. Attivissimo, and N. Giaquinto, “An acoustic Instrum. Meas., vol. 66, no. 12, pp. 3200–3209, Dec. 2017.
method for soil moisture measurement,” IEEE Trans. Instrum. Meas., [99] MaxStream. (2007). XBeeTM /XBee-PROTM OEM RF Modules,
vol. 53, no. 4, pp. 891–898, Aug. 2004. Product Manual v1.xAx-802.15.4 Protocol. Accessed: Jan. 20, 2020.
[78] S. Athani, C. Tejeshwar, M. M. Patil, P. Patil, and R. Kulkarni, “Soil [Online]. Available: https://www.sparkfun.com/datasheets/Wireless/
moisture monitoring using IoT enabled arduino sensors with neural Zigbee/XBee-Manual.pdf
networks for improving soil management for farmers and predict sea- [100] W. Xue-fen et al., “Smartphone accessible agriculture IoT node based
sonal rainfall for planning future harvest in north Karnataka–India,” on NFC and BLE,” in Proc. IEEE Int. Symp. Consum. Electron. (ISCE),
in Proc. IEEE Int. Conf. I-SMAC (IoT Soc. Mobile Anal. Cloud) Kuala Lumpur, Malaysia, 2017, pp. 78–79.
(I-SMAC), Palladam, India, 2017, pp. 43–48. [101] K. Tanaka, M. Murase, and K. Naito, “Prototype implementation of
[79] S. Khan and M. M. Hussain, “IoT enabled plant sensing systems for BLE based automated data collection scheme in agricultural measure-
small and large scale automated horticultural monitoring,” in Proc. ment system,” in Proc. 15th IEEE Annu. Consum. Commun. Netw.
IEEE 5th World Forum Internet Things (WF-IoT), Limerick, Ireland, Conf. (CCNC), Las Vegas, NV, USA, 2018, pp. 1–2.
2019, pp. 303–308.
[102] A. Paventhan, S. K. Allu, S. Barve, V. Gayathri, and N. M. Ram,
[80] T. Okayasu et al., “Affordable field environmental monitoring and
“Soil property monitoring using 6LoWPAN-enabled wireless sensor
plant growth measurement system for smart agriculture,” in Proc. 11th
networks,” in Proc. 3rd Conf. AIPA, Hyderabad, India, Aug. 2012,
Int. Conf. Sens. Technol. (ICST), Sydney, NSW, Australia, 2017,
pp. 278–282.
pp. 1–4.
[81] J. Hadabas, M. Hovari, I. Vass, and A. Kertész, IOLT Smart Pot: [103] Z. Suryady, M. H. M. Shaharil, K. A. Bakar, R. Khoshdelniat,
An IoT-Cloud Solution for Monitoring Plant Growth in Greenhouses, G. R. Sinniah, and U. Sarwar, “Performance evaluation of 6LoWPAN-
SciTePress, Setúbal, Portugal, 2019. based precision agriculture,” in Proc. IEEE Int. Conf. Inf. Netw.
[82] W. Slamet, N. Irham, and M. Sutan, “IoT based growth monitoring (ICOIN2011), Barcelona, Spain, 2011, pp. 171–176.
system of Guava (psidium guajava l.) fruits,” in Proc. IOP Conf. Series [104] X. Zhang, A. Andreyev, C. Zumpf, M. C. Negri, S. Guha, and
Earth Environ. Sci., vol. 147, no. 1, 2018, Art. no. 012048. M. Ghosh, “Thoreau: A subterranean wireless sensing network for
[83] K. Patil and N. Kale, “A model for smart agriculture using IoT,” in agriculture and the environment,” in Proc. IEEE Conf. Comput.
Proc. IEEE Int. Conf. Global Trends Signal Process. Inf. Comput. Commun. Workshops (INFOCOM WKSHPS), Atlanta, GA, USA, 2017,
Commun. (ICGTSPICC), Jalgaon, India, 2016, pp. 543–545. pp. 78–84.
[84] N. Materne and M. Inoue, “IoT monitoring system for early detec- [105] D. Davcev, K. Mitreski, S. Trajkovic, V. Nikolovski, and N. Koteli,
tion of agricultural pests and diseases,” in Proc. IEEE 12th South “IoT agriculture system based on LoRaWAN,” in Proc. 14th IEEE
East Asian Techn. Univ. Consortium (SEATUC), vol. 1. Yogyakarta, Int. Workshop Factory Commun. Syst. (WFCS), Imperia, Italy, 2018,
Indonesia, 2018, pp. 1–5. pp. 1–4.

Authorized licensed use limited to: The University of British Columbia Library. Downloaded on September 23,2020 at 18:46:50 UTC from IEEE Xplore. Restrictions apply.
 210 IEEE JOURNAL OF RADIO FREQUENCY IDENTIFICATION, VOL. 4, NO. 3, SEPTEMBER 2020

[106] H.-C. Lee and K.-H. Ke, “Monitoring of large-area IoT sensors using [130] M. S. Mekala and P. Viswanathan, “A novel technology for smart agri-
a LoRa wireless mesh network system: Design and evaluation,” IEEE culture based on IoT with cloud computing,” in Proc. IEEE Int. Conf.
Trans. Instrum. Meas., vol. 67, no. 9, pp. 2177–2187, Sep. 2018. I-SMAC (IoT in Soc. Mobile Anal. Cloud) (I-SMAC), Palladam, India,
[107] X. Jia, Q. Feng, and C. Ma, “An efficient anti-collision protocol 2017, pp. 75–82.
for RFID tag identification,” IEEE Commun. Lett., vol. 14, no. 11, [131] P. Mell and T. Grance, The NIST Definition of Cloud Computing, Nat.
pp. 1014–1016, Nov. 2010. Inst. Stand. Technol., Gaithersburg, MD, USA, 2011.
[108] T. Wasson, T. Choudhury, S. Sharma, and P. Kumar, “Integration of [132] I. A. T. Hashem, I. Yaqoob, N. B. Anuar, S. Mokhtar, A. Gani, and
RFID and sensor in agriculture using IoT,” in Proc. IEEE Int. Conf. S. U. Khan, “The rise of ‘big data’ on cloud computing: Review and
Smart Technol. Smart Nation (SmartTechCon), Bangalore, India, 2017, open research issues,” Inf. Syst., vol. 47, pp. 98–115, Jan. 2015.
pp. 217–222. [133] S. Nativi, P. Mazzetti, M. Santoro, F. Papeschi, M. Craglia, and
[109] K. Finkenzeller, RFID Handbook: Fundamentals and Applications in O. Ochiai, “Big data challenges in building the global Earth obser-
Contactless Smart Cards, Radio Frequency Identification and Near- vation system of systems,” Environ. Model. Softw., vol. 68, pp. 1–26,
Field Communication. Chichester, U.K.: Wiley, 2010. Jun. 2015.
[110] M.-H. Liang, Y.-F. He, L.-J. Chen, and S.-F. Du, “Greenhouse environ- [134] J. Shao, Q. Wang, and H. Mei, “Model based monitoring and con-
ment dynamic monitoring system based on WiFi,” IFAC PapersOnLine, trolling for platform-as-a-service (PaaS),” Int. J. Cloud Appl. Comput.
vol. 51, no. 17, pp. 736–740, 2018. (IJCAC), vol. 2, no. 1, pp. 1–15, 2012.
[111] K. Bannister, G. Giorgetti, and S. Gupta, “Wireless sensor networking [135] T.-C. Hsu, H. Yang, Y.-C. Chung, and C.-H. Hsu, “A creative IoT agri-
for ‘hot’ applications: Effects of temperature on signal strength, data culture platform for cloud fog computing,” Sustain. Comput. Informat.
collection and localization,” in Proc. 5th Workshop Embedded Netw. Syst., to be published, 2018.
Sens. (HotEmNets’08), 2008. [136] M. Gayatri, J. Jayasakthi, and G. A. Mala, “Providing smart agricul-
tural solutions to farmers for better yielding using IoT,” in Proc. IEEE
[112] C. A. Boano, N. Tsiftes, T. Voigt, J. Brown, and U. Roedig, “The
Technol. Innov. ICT Agr. Rural Develop. (TIAR), Chennai, India, 2015,
impact of temperature on outdoor industrial sensornet applications,”
pp. 40–43.
IEEE Trans. Ind. Informat., vol. 6, no. 3, pp. 451–459, Aug. 2010.
[137] A. Abdullah, S. Al Enazi, and I. Damaj, “AgriSys: A smart and ubiqui-
[113] P. Sheng, Y. Guo, and P. Li, “Intelligent measurement and control tous controlled-environment agriculture system,” in Proc. 3rd MEC Int.
system of facility agriculture based on ZigBee and 3G,” Trans. Chin. Conf. Big Data Smart City (ICBDSC), Muscat, Oman, 2016, pp. 1–6.
Soc. Agr. Mach., vol. 43, no. 12, pp. 229–233, 2012. [138] H. Channe, S. Kothari, and D. Kadam, “Multidisciplinary model
[114] X. Wang, C. Yang, and S. Mao, “TensorBeat: Tensor decomposition for for smart agriculture using Internet-of-Things (IoT), sensors, cloud-
monitoring multiperson breathing beats with commodity WiFi,” ACM computing, mobile-computing & big-data analysis,” Int. J. Comput.
Trans. Intell. Syst. Technol. (TIST), vol. 9, no. 1, p. 8, 2017. Technol. Appl., vol. 6, no. 3, pp. 374–382, 2015.
[115] L. Vangelista, A. Zanella, and M. Zorzi, “Long-range IoT technolo- [139] S. Sicari, A. Rizzardi, L. A. Grieco, and A. Coen-Porisini, “Security,
gies: The dawn of LoRaTM ,” in Proc. 1st EAI Int. Conf. Future Access privacy and trust in Internet of Things: The road ahead,” Comput. Netw.,
Enablers Ubiquitous Intell. Infrastruct. Ohrid, North Macedonia, vol. 76, pp. 146–164, Jan. 2015.
Sep. 2015, pp. 51–58. [140] S. Jebri, M. Abid, and A. Bouallegue, “STAC-protocol: Secure and trust
[116] J. Dilley, B. Maggs, J. Parikh, H. Prokop, R. Sitaraman, and B. Weihl, anonymous communication protocol for IoT,” in Proc. IEEE 13th Int.
“Globally distributed content delivery,” IEEE Internet Comput., vol. 6, Wireless Commun. Mobile Comput. Conf. (IWCMC), Valencia, Spain,
no. 5, pp. 50–58, Sep./Oct. 2002. 2017, pp. 365–370.
[117] M. Satyanarayanan, V. Bahl, R. Caceres, and N. Davies, “The case for [141] D. Shadrin, A. Menshchikov, A. Somov, G. Bornemann, J. Hauslage,
VM-based cloudlets in mobile computing,” IEEE Pervasive Comput., and M. Fedorov, “Enabling precision agriculture through embedded
vol. 8, no. 4, pp. 14–23, Oct.–Dec. 2009. sensing with artificial intelligence,” IEEE Trans. Instrum. Meas., early
[118] F. Bonomi, R. Milito, J. Zhu, and S. Addepalli, “Fog computing and access, Oct. 14, 2019, doi: 10.1109/TIM.2019.2947125.
its role in the Internet of Things,” in Proc. 1st Ed. MCC Workshop [142] Y. LeCun, Y. Bengio, and G. Hinton, “Deep learning,” Nature, vol. 521,
Mobile Cloud Comput., 2012, pp. 13–16. no. 7553, pp. 436–444, 2015.
[119] M. Satyanarayanan, “The emergence of edge computing,” Computer, [143] S. Liu, S. Cossell, J. Tang, G. Dunn, and M. Whitty, “A com-
vol. 50, no. 1, pp. 30–39, Jan. 2017. puter vision system for early stage grape yield estimation based
[120] X. Chen, Q. Shi, L. Yang, and J. Xu, “ThriftyEdge: Resource-efficient on shoot detection,” Comput. Electron. Agr., vol. 137, pp. 88–101,
edge computing for intelligent IoT applications,” IEEE Netw., vol. 32, May 2017.
no. 1, pp. 61–65, Jan./Feb. 2018. [144] S. R. Dubey and A. S. Jalal, “Adapted approach for fruit disease iden-
[121] M. A. Zamora-Izquierdo, J. Santa, J. A. Martínez, V. Martínez, and tification using images,” Int. J. Comput. Vis. Image Process. (IJCVIP),
A. F. Skarmeta, “Smart farming IoT platform based on edge and cloud vol. 2, no. 3, pp. 44–58, 2012.
computing,” Biosyst. Eng., vol. 177, pp. 4–17, Jan. 2019. [145] D. J. A. Rustia and T.-T. Lin, “An IoT-based wireless imaging and
[122] D. Fan and S. Gao, “The application of mobile edge computing in agri- sensor node system for remote greenhouse pest monitoring,” Chem.
cultural water monitoring system,” in Proc. IOP Conf. Earth Environ. Eng. Trans., vol. 58, pp. 601–606, Jun. 2017.
Sci., vol. 191, 2018, Art. no. 012015. [146] S. D. Khirade and A. Patil, “Plant disease detection using image pro-
cessing,” in Proc. IEEE Int. Conf. Comput. Commun. Control Autom..
[123] A. Reziouk, E. Laurent, and J.-C. Demay, “Practical security overview
Pune, India, 2015, pp. 768–771.
of IEEE 802.15.4,” in Proc. IEEE Int. Conf. Eng. MIS (ICEMIS),
[147] P. Palli et al., “Robotic gantry for automated imaging, sensing, crop
Agadir, Morocco, 2016, pp. 1–9.
input application, and high-throughput analysis,” in Proc. ASABE Annu.
[124] IEEE 802.15.4 Standard: A Tutorial/Primer, electronicsnotes, Int. Meeting, 2019, p. 1.
Apr. 2020. Accessed: Jan. 29, 2020. [Online]. Available: [148] M. Nachidi, F. Rodríguez, F. Tadeo, and J. Guzman, “Takagi–Sugeno
https://www.electronics-notes.com control of nocturnal temperature in greenhouses using air heating,” ISA
[125] T. Kumar and P. Mane, “ZigBee topology: A survey,” in Proc. IEEE Trans., vol. 50, no. 2, pp. 315–320, 2011.
Int. Conf. Control Instrum. Commun. Comput. Technol. (ICCICCT), [149] M. Azaza, C. Tanougast, E. Fabrizio, and A. Mami, “Smart
Kumaracoil, India, 2016, pp. 164–166. greenhouse fuzzy logic based control system enhanced with
[126] “ZigBee control your world, ZigBee specification,” ZigBee Alliance, wireless data monitoring,” ISA Trans., vol. 61, pp. 297–307,
Rep. 05-3474-21, Aug. 2015. Accessed: Jan. 25, 2020. [Online]. Mar. 2016.
Available: https://zigbeealliance.org/wp-content/uploads/2019/11/docs- [150] R. R Shamshiri et al., “Advances in greenhouse automation and con-
05-3474-21-0csg-zigbee-specification.pdf trolled environment agriculture: A transition to plant factories and
[127] J. Kim, S.-K. Kang, and J. Park, “Bluetooth-based tree topology urban agriculture,” Int. J. Agr. Biol. Eng., vol. 11, no. 1, pp. 1–22,
network for wireless industrial applications,” in Proc. 15th Int. 2018.
Conf. Control Autom. Syst. (ICCAS), Busan, South Korea, 2015, [151] E. Heuvelink, “Evaluation of a dynamic simulation model for tomato
pp. 1305–1308. crop growth and development,” Ann. Bot., vol. 83, no. 4, pp. 413–422,
[128] H. A. Al-Kashoash and A. H. Kemp, “Comparison of 6LoWPAN and 1999.
LPWAN for the Internet of Things,” Aust. J. Elect. Electron. Eng., [152] P. Abreu, J. Meneses, and C. Gary, “Tompousse, a model of yield
vol. 13, no. 4, pp. 268–274, 2016. prediction for tomato crops: Calibration study for unheated plastic
[129] “IoT standards and protocols,” Postscapes, Seattle, WA, USA, greenhouses,” in Proc. XXV Int. Horticul. Congr. 9 Comput. Autom.
Feb. 2020. [Online]. Available: https://www.postscapes.com Electron. Inf. Horticul., 1998, pp. 141–150.

Authorized licensed use limited to: The University of British Columbia Library. Downloaded on September 23,2020 at 18:46:50 UTC from IEEE Xplore. Restrictions apply.
 RAYHANA et al.: INTERNET OF THINGS EMPOWERED SMART GREENHOUSE FARMING 211

[153] C. Gary, A. Baille, M. Navarrete, and R. Espanet, “Tompousse, un Zheng Liu (Senior Member, IEEE) received the first
modèle simplifié de prévision du rendement et du calibre de la tomate,” Ph.D. degree in engineering from Kyoto University,
Actes du Séminaire de l’AIP intersectorielle" Serres", INRA, Avignon, Kyoto, Japan, in 2000, and the second Ph.D. degree
1997, pp. 100–109. from the University of Ottawa in 2007. From 2000
[154] A. S. Patil, B. A. Tama, Y. Park, and K.-H. Rhee, “A framework for to 2001, he was a Research Fellow with Nanyang
blockchain based secure smart green house farming,” in Advances in Technological University, Singapore. Then he joined
Computer Science and Ubiquitous Computing. Singapore: Springer, the Institute for Aerospace Research (IAR), National
2017, pp. 1162–1167. Research Council (NRC) Canada, Ottawa, ON,
Canada, as a Governmental Laboratory Visiting
Fellow (nominated by NSERC). After being with
IAR for five years, he transferred to the NRC
Institute for Research in Construction, where he held a research officer
Rakiba Rayhana (Student Member, IEEE) received position. From 2012 to 2015, he worked as a Full Professor with Toyota
the B.Sc. degree in electrical and electronic engi- Technological Institute, Nagoya, Japan. In August 2015, he joined the
neering from BRAC University, Bangladesh, and the University of British Columbia (Okanagan campus), Kelowna, BC, Canada.
M.Sc. degree in electronic and electrical engineer- His research interests include condition-based maintenance, condition assess-
ing from the University of Strathclyde, Glasgow, ment, nondestructive inspection and evaluation, prognostic health manage-
U.K. She is currently pursuing the Ph.D. degree in ment, data/information fusion, computer/machine vision, pattern recognition,
electrical engineering with the University of British machine learning, and sensor/sensor network. He is chairing the IEEE IMS
Columbia, Canada. She is also a Student Member of TC-36 and serving as the VP publication of the IEEE Instrumentation and
SPIE Organization. Measurement Society from 2016 to 2017. He holds a Professional Engineer
licenses in both Ontario and British Columbia. He also serves on the edi-
torial boards for journal IEEE T RANSACTIONS ON I NSTRUMENTATION
AND M EASUREMENT , Information Fusion (Elsevier), Machine Vision and
Applications (Springer), and Intelligent Industrial Systems (Springer). He is a
Gaozhi (George) Xiao (Fellow, IEEE) received member of SPIE.
the Ph.D. degree from Loughborough University,
U.K., in 1995. He is a Principle Research Officer
with the National Research Council of Canada.
He has published more than 100 papers in refer-
eed journals and refereed conference proceedings,
and one co-edited book. He is currently holding
5 U.S. patents. He has brought several technolo-
gies from concepts to products. He has managed
large Research and Development projects in indus-
tries, academics and government labs covering areas,
including RFID/NFC, flexible/printable/wearable electronics, fiber optic sen-
sor systems, photonic sensing and measurement, structural health monitoring,
indoor air quality monitoring, structural materials, and smart materials. He
received the 2014 Technical Award from the IEEE Instrumentation and
Measurement Society. He is currently an Editor-in-Chief of the IEEE Journal
of Radio Frequency Identification and an Associate Editor-in-Chief of IEEE
T RANSACTIONS ON I NSTRUMENTATION AND M EASUREMENT.

Authorized
View publicationlicensed
stats use limited to: The University of British Columbia Library. Downloaded on September 23,2020 at 18:46:50 UTC from IEEE Xplore. Restrictions apply.