IEEE INTERNET OF THINGS JOURNAL, VOL. 5, NO.

6, DECEMBER 2018 4409

An Internet-of-Things Enabled Smart Sensing

System for Nitrate Monitoring
Md. Eshrat E. Alahi , Student Member, IEEE, Najid Pereira-Ishak,
Subhas Chandra Mukhopadhyay, Fellow, IEEE, and Lucy Burkitt

Abstract—Monitoring the nitrate concentration in the field is a routine manner. This is done by expert staff at regular inter-
an excellent ability for a water-monitoring study. We report an vals to track the change of nitrate concentration [10]. However,
interdigital FR4-based capacitive sensor, which is characterized data taken too seldom may not be adequate to measure the
for nitrate concentration. The concentration range of nitrate is
0–40 ppm (mg/L). Different unknown samples were measured actual nitrate profile.
and validated with standard UV-Spectrometry. A smart sensing There are conventional standard methods available for
node has been developed which can collect water from a lake, nitrate measurement, such as UV/V is spectrometry, chro-
stream, or river, measure the instantaneous nitrate concentration, matography, HPLC, and capillary electrophoresis. They are
and transfer the data through the gateway to a user-defined unsuitable for creating a large-scale wireless sensor net-
cloud server. The system is completely autonomous and solar
powered, robust, and trialed in the field successfully. A simple work (WSN) due to their massive instrumentation, bulky fea-
moving-average algorithm is used to smooth the collected data in tures, and complex and sensitive measurement procedure; and
the cloud side. The LoRa protocol and WiFi protocol are com- above all they are costly. They also produce lots of chemical
pared in terms of power consumption. The proposed system is waste which might be harmful to the environment [11].
trialed in the field continuously and the result validated with A WSN consists of a number of dedicated sensor nodes
standard UV-Spectrometry. The developed smart system can be
easily deployable and friendly to use, and offers new possibilities with physical sensing and computing abilities, which can sense
for both spatial and temporal analysis for nitrate concentration. and monitor the surrounding physical parameters. A WSN has
a lot of essential characteristics and a few constraints, such
Index Terms—Interdigital sensor, Internet of Things (IoT), as limited energy and computational power. During the last
LoRa protocol, nitrate concentration, sensing node, water mon- decades, WSNs have been widely used in different applica-
itoring, WiFi protocol, wireless sensor network (WSN). tions related to water monitoring [12]–[14], forests [15], [16],
industrial [17], [18], and agricultural [19]–[22]. Some
researches were reported [23]–[25] to monitor the nitrate con-
I. I NTRODUCTION
centration through the WSN network [26]. Their limitation
ITROGEN is an important nutrient for plants and is
N contained in the building blocks of life, such as in
nucleotides, amino acids, and proteins [1]. The excessive use
was that most of them are laboratory-based and perform the
real-time analysis in the laboratory only. Capella et al. [24]
reported in-line analysis of nitrate concentration in river water
of fertilizers, urination due to animal farming, and indus- and emphasized the feasibility of development of WSNs-based
trial waste are the significant reasons for nitrate leaching in continuous monitoring.
water [2]–[7]. Excessive nitrate can hamper aquatic life and The Internet of Things (IoT) enables any physical objects
can lead to algal blooms and eutrophication [8]. According to communicate through the Internet and transfer data to
to the U.S. Environmental Protection Agency, the maximum a specific server for further processing. It requires pervasive
allowable nitrate concentration in drinking water should be computing, smart sensors, embedded devices, communication
44.2 ppm. Over this range of concentration it would be technologies, Internet protocols, and applications. IoT-enabled
considered as contaminated water [9]. WSNs help to connect more objects for monitoring purposes
The concentration of nitrate fluctuates in water, both spa- to build up smart cities, smart industries, smart agriculture,
tially and temporally. It also depends on the season, weather etc.
conditions, and the rainfall of any location. Currently, the In this paper, a low-cost interdigital FR4-based capacitive
regional council or local government collects water samples in sensor is proposed for nitrate measurement. The developed
sensor has been characterized and calibrated against standards
Manuscript received November 15, 2017; revised December 22, 2017
and February 3, 2018; accepted February 20, 2018. Date of publication to measure nitrate components. The measured results are val-
February 27, 2018; date of current version January 16, 2019. (Corresponding idated against the standard UV-spectrometry method. A smart
author: Md. Eshrat E. Alahi.) sensing node has been developed which is suitable for an IoT-
Md. E. E. Alahi, N. Pereira-Ishak, and S. C. Mukhopadhyay are with
the School of Engineering, Macquarie University, Sydney, NSW 2109, enabled WSN. The sensing node collects water from stream
Australia (e-mail: [email protected]; [email protected]; or lake and measures the nitrate concentration. The collected
[email protected]). water can be discharged automatically to stream or lake again.
L. Burkitt is with the School of Agriculture, Massey University, Wellington,
New Zealand (e-mail: [email protected]). A solar panel is used for energy harvesting and provides the
Digital Object Identifier 10.1109/JIOT.2018.2809669 system with the necessary energy for monitoring continuously.
2327-4662 c 2018 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.
See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
Authorized licensed use limited to: INTERNATIONAL ISLAMIC UNIVERSITY. Downloaded on August 12,2023 at 10:26:30 UTC from IEEE Xplore. Restrictions apply.
4410 IEEE INTERNET OF THINGS JOURNAL, VOL. 5, NO. 6, DECEMBER 2018

Fig. 1. Electric field lines of parallel-plate capacitor and planar interdigital

sensor.

The LoRa protocol is used to transfer the measured data, and

saves energy. Finally, the measured data can be transferred to
an IoT-based cloud server, which is free for the users. Cloud Fig. 2. FR4 interdigital sensor and dimensions.
data is also analyzed further to identify the trend of nitrate
fluctuation.
B. WSNs
Section II explains the materials and methods, Section III
explains the experimental setup and study location, Section IV A WSN is formed from a number of small sensing systems,
explains all the collected results and discusses them. The last named sensing nodes, that collect information from the sur-
section has the conclusions from this paper and suggestions roundings through sensors. The sensing systems transmit the
for future work. collected information to a “gateway” through wireless commu-
nication. The collected data can be handled by the gateway and
sent to the cloud server where an “information management
II. M ATERIALS AND M ETHODS system (IMS)” analyzes the data in real time or for statistical
A. Working Principle of Interdigital Sensor analysis.
The operating principle of the planar interdigital sensor [27]
is similar to that of two parallel-plate capacitors, and the elec- C. Structure of WSN
trodes open up to one side to provide a one-sided access to the A WSN consists of three different subsystems. They are the
material under test (MUT) [27]. The generated electric-field nodes, the gateway, and the IMS.
lines penetrate into the MUT and the impedance of the sensor 1) Sensor Node: This is also called a sensing system with
will change. As a result, the impedance of the sensor becomes a low-cost, small low-power sensor, which can get appropriate
a function of the system properties. Therefore, the system measurements from the environment, process the measured
properties can be evaluated by measuring the impedance of data, and send them directly to the cloud server through the
the sensor. gateway. It consists of the following elements.
The term “interdigital” refer to a digit-like or finger-like • Microcontroller-Based System: This is the core of the
periodic pattern of parallel electrodes which helps to build sensing node, which will be a low-cost, small, and
up the impedance associated with the provided electric field low-power chip. Unfortunately, it is difficult to get all
that penetrates into the material sample [27]. An excitation these characteristics due to certain limitations, especially
alternating-current (ac) voltage is applied between the pos- regarding the computer power and memory.
itive terminal, and the negative terminal, and the electric • Power Supply Unit: Sensor nodes require an autonomous
field forms from the positive terminal to the negative termi- functioning system, which can provide continuous power
nal. Fig. 1 indicates the evolution of an interdigital sensor to run the system all day round. Sensor nodes always
toward a one-sided measurement. Interdigital sensors have depend heavily on the power supply unit.
been used in different applications for domestic and industrial • Wireless Communication Network: This network will
applications [21], [28], [29]. allow the necessary communication between the sen-
Fig. 2 shows the FR4 interdigital sensor used in the pro- sor nodes and the gateway. They use certain standard
posed sensing system and Fig. 3 shows its equivalent circuit. protocols which have different coverage regions, power
Rp represents the conductive properties of the sample water consumptions, and suitability for different applications.
which is under the effect of an electric field. It also explains the • Sensors: The sensor node also contains sensors, which
resistive nature of the sample water. Cp represents the capac- convert the physical parameter into an electrical signal
itance of the sample water to be measured. The fabrication to allow the microcontroller to process the filtered output
process is similar to the process used to develop a printed cir- data and send the data during transmissions.
cuit board. Electrodes are made from copper which is a good 2) Gateway: This is also called the “base station” of a net-
conductor. The dimensions of the sensing area are 33 mm × work. This is the core of the network, and collects data from
17 mm, which is easy to dip in to water. Tin oxide is layered the sensor node, processes and helps to store the data to a cloud
as a coating material on the copper electrodes to keep them server. It has also a computing system, which is based on
corrosion free. a high-power microcontroller or has high computing ability.

Authorized licensed use limited to: INTERNATIONAL ISLAMIC UNIVERSITY. Downloaded on August 12,2023 at 10:26:30 UTC from IEEE Xplore. Restrictions apply.
ALAHI et al.: IoT ENABLED SMART SENSING SYSTEM FOR NITRATE MONITORING 4411

Fig. 3. Equivalent circuit of interdigital sensor.

Fig. 4. EIS measurement in laboratory conditions.

It has to be static and plugged into the mains power supply,
as it requires more energy than the sensor nodes. It also should
sources are available, such as fluid flow, vibration, electromag-
have a wireless communication system, which is utilized by
netic fields, and so on. However, the most-used energy source
the sensing nodes. The gateway sends the collected data to the
for WSN applications is photovoltaic panels or solar panels.
cloud server through Ethernet, WiFi, or 3G/4G, etc.
The solar panels convert light (sunlight or artificial light) into
3) Cloud Server: This is the last component of a WSN electricity. Batteries can store the converted electricity and uti-
network. It consists of a database and suitable management lize the energy when there is no sunlight, such as on cloudy
software. It might be located in the gateway or any other days or at night. Sensor nodes and systems should run cleverly
remote computer. Users can get access to the management to extend the battery life, and therefore energy harvesting is
software through the Internet. an important factor in WSN application.

D. LoRaWAN Protocol for IoT III. E XPERIMENTAL S ETUP

Long-range, low-power wireless area network (LoRaWAN) A. Electrochemical Impedance Spectroscopy
is a data-link layer with long range, low power, and a low
Electrochemical impedance spectroscopy (EIS) [32]–[34] is
bit rate, which is a promising solution for IoT applications.
a highly sensitive method for impedance measurement. There
A LoRa-enabled sensor node consumes low energy and trans-
are many methods available to measure impedance, but a fre-
mits only a few bytes, and so is an excellent candidate for use
quency response analyzer (FRA) is considered the de facto
in many different applications (such as smart health care, smart
standard for EIS measurements. FRA requires a small AC volt-
cities, environmental monitoring, industry, etc.). There are two
age with an amplitude of 5–15 mV, and the frequency of the
distinct layers: 1) a physical layer, based on radio modulation
signal sweeps in a certain range based on a direct-current bias
called chirp spread spectrum and 2) a MAC-layer protocol
voltage. The signal is connected to the positive electrode, or
which is responsible to get access in LoRa architecture [30].
the working electrode, and a changed voltage is taken from the
LoRa modulation has the same characteristics as frequency
negative electrode, or sensing electrode. Due to the different
shift keying regarding the communication range in between
characteristics of different materials, their impedance profile is
gateway and sensor node. Thus, LoRaWAN is considered as
different and the following equation represents the impedance:
a communication protocol and network architecture, while
LoRa supports the long-range link. The node’s battery life, Z = R + jX (1)
the network capacity, the quality of service, security, and reli-
where Z is the total impedance (), R is the real part of the
ability are determined by the network architecture and the
impedance (), and X is the imaginary part of the impedance
defined protocol. It also supports virtualized wireless network-
(). The impedance profile can be represented graphically
ing technologies, where all base stations work together and are
with what is called a bode plot or Cole–Cole plot. The FR4-
collectively seen by the sensor nodes [30].
based interdigital sensor was characterized by a Hioki IM
3536 LCR meter where the frequency was swept from 10 Hz
E. Energy Harvesting to 100 kHz. Standard laboratory temperature and humidity was
Sensor nodes are rarely connected to a fixed power supply; maintained throughout these experiments. 1, 10, 20, 30, and
rather they are an independent, autonomous system with an 40 ppm nitrate solutions were taken as standard solutions.
energy-harvesting capability. Their energy consumption must Deionized water was taken as a control solution. The aver-
be limited, which can be achieved by having low-consumption age pH of the solution was 6.71, which is also maintained
operating modes. Therefore, chargeable batteries are always in creek water. Fig. 4 shows the laboratory setup for EIS
used in WSN applications with smart modes (sleep and active). data acquisition. Initially the sensor was characterized to get
For some applications, it is also required to have a provision the impedance profile to develop the calibration standard for
to collect the required energy from the environment. These nitrate measurement. All experiments were repeated five times
are called energy-harvesting techniques [31]. Energy sources to observe the impedance behavior, and average results were
should be clean and environmentally friendly. Different energy calculated.

Authorized licensed use limited to: INTERNATIONAL ISLAMIC UNIVERSITY. Downloaded on August 12,2023 at 10:26:30 UTC from IEEE Xplore. Restrictions apply.
4412 IEEE INTERNET OF THINGS JOURNAL, VOL. 5, NO. 6, DECEMBER 2018

Fig. 5. Block diagram of the data transmission.

Fig. 6. Circuit diagram of the sensing system.

B. System Description
Fig. 7. Different parts of the proposed system.
A smart sensing system is proposed to carry out the in-
line nitrate analysis as per Fig. 5. Fig. 6 shows the circuit
diagram of the developed system. Arduino Uno and Arduino gateway is also responsible to send the data to the cloud server.
Uno WiFi [35] was used as the main microcontroller. An Thingspeak is used as an IoT-based cloud server, which is free
AD5933 [36] is used to get the impedance data from the sen- to use and can easily store data.
sor. The impedance analyzer gets the impedance and the phase WiFi and LoRa, both communication protocols, were used
shift of the sensor. The impedance analyzer has the ability to to examine the durability of the sensor nodes. The Arduino
sweep the frequency, which is not required in the current appli- Uno WiFi has a WiFi module, which is responsible for trans-
cation. The operating frequency is fixed and each measurement mitting and receiving data through the gateway. An LG01S is
is done five times. Only the average nitrate concentration is also used as a WiFi gateway. All the microcontrollers, sensor,
sent to the IoT cloud server. inlet and outlet pumps, rechargeable battery, solar charge con-
A Dragino LoRa shield is used as a long-range transceiver troller, and water reservoir are contained in a steel box which
to communicate with the gateway. It allows sending data and is robust and easy to install. Fig. 7 indicates the materials
reaching an extremely long range with low data rates. It is which are used in the sensing node. Fig. 8 shows the inside
based on the RFM95W/RFM98W [37] and used for 915-MHz of the steel box.
transmission/reception. An L298N [38] is used as a motor
driver to control the inlet and outlet pumps. The inlet pump
brings the water into the reservoir and the outlet pump empties C. Energy-Harvesting Technique
the reservoir after the measurement. A solar panel (model: ZM-9051), solar charge controller
LG01S [39] is used as the LoRa/WiFi gateway to commu- (MP-3750), and a sealed rechargeable battery (12 V, 12 AH)
nicate between the sensing node and the cloud server. The were used to provide energy continuously without any human

Authorized licensed use limited to: INTERNATIONAL ISLAMIC UNIVERSITY. Downloaded on August 12,2023 at 10:26:30 UTC from IEEE Xplore. Restrictions apply.
ALAHI et al.: IoT ENABLED SMART SENSING SYSTEM FOR NITRATE MONITORING 4413

Fig. 8. Inside of the sensor node.

Fig. 10. Study location and distance between gateway and the sensor node.

Fig. 9. Field installation of the smart sensor node.

intervention. All through the day the system is controlled from

the microcontroller in various operation modes (active, sleep,
and transmitting/receiving). Due to the steel structure, the
microcontroller had some difficulty communicating with the
gateway to send the data from the sensor node with the existing
antenna. Therefore, a VERT900 omni-directional antenna was
used and extended through the steel box. Finally, it was Fig. 11. Bode plots for different nitrate concentrations.
installed near the study location as per Fig. 9.

D. Study Location properties of different concentrations are different, which is

The study location is in Macquarie University near a small reflected in the bode plot [32]–[34].
creek. As is seen from Fig. 10, that sensor node is installed The real impedance and imaginary impedance are plot-
340 m from the gateway. The gateway is installed in such ted with respect to frequency, and it is noted that the real
a way that the sensor node has a clear line of sight and there impedance has a significant change unlike the imaginary
is no obstacle to data transmission. The blue marker in the map impedance. The change is stable and consistent when the
indicates the gateway location and the green marker indicates frequency is more than 500 Hz.
the sensor node’s location.
B. Calibration Standard
IV. R ESULTS AND D ISCUSSION
It is seen from Fig. 12 that the sensitive region for different
A. EIS Measurement for Nitrate Concentrations concentrations is 500 Hz to 100 kHz. There is a signifi-
The impedances of different concentrations of nitrate sam- cant change in the real impedance in that frequency range
ple are measured during an EIS measurement and plotted as for different nitrate concentrations. 1000 Hz is chosen as
Fig. 11. It is seen that the bode plots for concentrations of 1, the operating frequency to develop a calibration standard to
10, 20, 30, and 40 ppm are different from each other due to the measure unknown nitrate concentrations. At 1000 Hz, the
impedance change. The number of ions for different concen- real impedances of 1, 10, 20, 30, and 40 ppm of nitrate
trations are different from each other. Therefore, the electric concentrations were used to develop a calibration standard
field from the positive electrode bulges through the ions in the to measure any unknown concentration. From Fig. 13, it is
aqueous medium toward the negative electrode. The dielectric seen that the nitrate concentrations are plotted on the x-axis

Authorized licensed use limited to: INTERNATIONAL ISLAMIC UNIVERSITY. Downloaded on August 12,2023 at 10:26:30 UTC from IEEE Xplore. Restrictions apply.
4414 IEEE INTERNET OF THINGS JOURNAL, VOL. 5, NO. 6, DECEMBER 2018

TABLE I
U NKNOWN S AMPLE M EASUREMENT C OMPARED TO
L ABORATORY S TANDARD M ETHOD

Fig. 12. Frequency versus real impedance for different nitrate concentrations.

Fig. 14. Nitrate concentration is Thingspeak server.

was used to measure the unknown nitrate concentrations. It is

observed from Table I that the sensor and developed system
can measure the nitrate concentration with an error of less
Fig. 13. Calibration standard to measure any unknown nitrate concentration. than 5%. When the concentrations are on the higher side, the
error is smaller than with lower nitrate concentrations.

and the corresponding real impedances are plotted on the

y-axis. They follow a straight line, and the regression co- D. Data Transfer to Cloud Server
efficient is R2 > 0.98, which is adequate to calculate any It is seen from Fig. 14 that the measured nitrate concen-
unknown impedance. It also means that the actual impedance tration was transferred from the field location during the field
and predicted impedance are close to each other. trial. Seven days of sampling data are collected without any
Therefore, the calibration standard for an unknown sample interruption and the sampling time was 13 min. Though this
concentration is was a very dense data collection, the sampling time could be
R − 2212.3 easily controlled through the software programming. In 24 h,
C= (2) 112–113 batches of sampling data could be collected with
−51.77
sampling every 13 min. Table II shows the average nitrate
where C is the concentration (ppm) and R () is the measured
concentration and compares the data with laboratory mea-
real impedance from an unknown sample. It is seen that the
surements. Every morning, afternoon and evening, the sample
sensitivity of the sensor is −51.77 /ppm.
water was collected from the creek and measured in the lab-
oratory immediately. In a single day, the sampling water was
C. Unknown Sample Measurement collected and measured in the laboratory to maximize the accu-
To measure any unknown sample, different sample waters racy. It is obvious that the sampling frequency will be not be
are collected from different sampling locations, such as river, similar to the developed smart system. It is also observed that
lake, stream, and tap water. Naturally, the concentration of the error (less than 5%) of the sampling data from the devel-
nitrate was not high enough to measure the range to high con- oped system is close to the laboratory measurement. A similar
centrations. Therefore, a nitrate sample was added to elevate thing is also seen from Fig. 15.
the concentration of nitrate. The FR4-based sensor and sens- Fig. 16 illustrates the single-day nitrate concentrations for
ing system were used to measure the nitrate concentrations and each sample time. It varied between 1.8 and 1.9 ppm while the
compared with the laboratory standard method. Equation (2) average nitrate concentration of that single day was 1.9 ppm.

Authorized licensed use limited to: INTERNATIONAL ISLAMIC UNIVERSITY. Downloaded on August 12,2023 at 10:26:30 UTC from IEEE Xplore. Restrictions apply.
ALAHI et al.: IoT ENABLED SMART SENSING SYSTEM FOR NITRATE MONITORING 4415

TABLE II
AVERAGE N ITRATE C ONCENTRATION OF THE S TUDY L OCATION
C OMPARED W ITH L ABORATORY S TANDARD M ETHOD

Fig. 15. Comparison of daily evolution of nitrate concentration with standard

method.
TABLE III
C OMPARISON OF W I F I AND L O R A D RIVEN S ENSING S YSTEM

Fig. 16. Nitrate concentration of a single day.

In addition, the collected data have been smoothed with a sim- LoRa enabled microcontroller. Therefore, a LoRa enabled
ple moving-average algorithm. The smoothed data can provide sensing system was used to do the necessary field trial.
the trend of nitrate concentration, which might be useful over
a longer time. The creek does not carry a high level of nitrate
concentration as it is located inside the University and the V. C ONCLUSION
administration of the University control all kind of pollution An IoT-enabled smart nitrate sensor and sensing system is
(air, water, and environment) very carefully. proposed to monitor the nitrate concentration in real-time. An
FR4-based interdigital sensor is characterized and is extremely
useful for robust use during nitrate monitoring. The system
E. LoRa Protocol Over WiFi Protocol is autonomous and was trialed in the field for seven days
Components for remote IoT applications must draw as little without any interruption. The LoRa protocol was used to run
power as possible to maximize battery life. The difference of the system over longer periods than for the WiFi protocol.
The LoRa protocol over the WiFi RF protocol was to reduce The LoRa protocol is a low-power energy-saving protocol,
overall power consumption during sleep mode to less than half. which was implemented successfully. The system’s collected
It also allows transmissions to penetrate obstacles and allows data were also validated through the UV-spectrometry method.
the data to travel larger distances whilst consuming less power The collected data shows that few data have been lost dur-
than the standard WiFi protocol. It optimizes the data exchange ing transmission, and 98% of data are successfully collected
with the gateway, allowing for lower power consumption as through the gateway to the cloud server. The results show
compared to WiFi. The current drain from each component that the proposed smart sensing system can be very useful to
for each stage, comparing WiFi and LoRa, can be observed develop a WSN to monitor nitrate concentration in real time.
in Table III. It is observed that the LoRa protocol is consum- It also shows that, without human interaction, changes of both
ing less current, which helped to harvesting the energy for temporal and spatial evolutions of nitrate concentration can be
longer time. The consumption of energy for pumps and motor monitored successfully.
drivers are similar for both the system. However, WiFi enabled The future work would be to use more sensing nodes to
microcontroller consumes 2.22 times more energy than the create a WSN. The data would be collected for a longer time to

Authorized licensed use limited to: INTERNATIONAL ISLAMIC UNIVERSITY. Downloaded on August 12,2023 at 10:26:30 UTC from IEEE Xplore. Restrictions apply.
4416 IEEE INTERNET OF THINGS JOURNAL, VOL. 5, NO. 6, DECEMBER 2018

predict trends, which will be analyzed in the cloud side. More [21] M. E. E. Alahi, L. Xie, S. Mukhopadhyay, and L. Burkitt, “A temperature
low-power sensors can be included to monitor the humidity, compensated smart nitrate-sensor for agricultural industry,” IEEE Trans.
Ind. Electron., vol. 64, no. 9, pp. 7331–7341, Sep. 2017.
the temperature, or the conductivity of the water. [22] M. E. E. Alahi, L. Xie, A. I. Zia, S. Mukhopadhyay, and L. Burkitt,
“Practical nitrate sensor based on electrochemical impedance measure-
ment,” in Proc. IEEE Int. Instrum. Meas. Technol. Conf. (I2MTC),
ACKNOWLEDGMENT Taipei, Taiwan, 2016, pp. 1–6.
The authors would like to thank the School of Engineering, [23] M. R. Gartia et al., “The microelectronic wireless nitrate sensor network
for environmental water monitoring,” J. Environ. Monitoring, vol. 14,
Macquarie University, for providing the laboratory facility and no. 12, pp. 3068–3075, 2012.
supporting the field trial during the research. [24] J. V. Capella, A. Bonastre, R. Ors, and M. Peris, “In line river monitoring
of nitrate concentration by means of a wireless sensor network with
energy harvesting,” Sensors Actuators B Chem., vol. 177, pp. 419–427,
R EFERENCES Feb. 2013.
[25] M. E. E. Alahi, A. Nag, S. C. Mukhopadhyay, and L. Burkitt,
[1] J. Dechorgnat et al., “From the soil to the seeds: The long journey of “A temperature-compensated graphene sensor for nitrate monitoring in
nitrate in plants,” J. Exp. Botany, vol. 62, no. 4, pp. 1349–1359, 2011. real-time application,” Sensors Actuators A Phys., vol. 269, pp. 79–90,
[2] P. F. Swann, Environmental Carcinogenesis: Contributions of Basic Jan. 2018.
Research: Carcinogenic Risk From Nitrite, Nitrate and N-Nitrosamines [26] G. Xu, W. Shen, and X. Wang, “Applications of wireless sensor networks
in Food, Courtauld Inst. Biochem., Middlesex Hospital Med. School, in marine environment monitoring: A survey,” Sensors, vol. 14, no. 9,
London, U.K., 1977. pp. 16932–16954, 2014.
[3] H. J. Di and K. C. Cameron, “Nitrate leaching losses and pasture yields [27] A. V. Mamishev, K. Sundara-Rajan, F. Yang, Y. Du, and M. Zahn,
as affected by different rates of animal urine nitrogen returns and appli- “Interdigital sensors and transducers,” Proc. IEEE, vol. 92, no. 5,
cation of a nitrification inhibitor—A lysimeter study,” Nutrient Cycling pp. 808–845, May 2004.
Agroecosyst., vol. 79, no. 3, pp. 281–290, 2007. [28] S. C. Mukhopadhyay, C. P. Gooneratne, G. S. Gupta, and
[4] O. Bodansky, “Methemoglobinemia and methemoglobin-producing S. N. Demidenko, “A low-cost sensing system for quality monitor-
compounds,” Pharmacological Rev., vol. 3, no. 2, pp. 144–191, 1951. ing of dairy products,” IEEE Trans. Instrum. Meas., vol. 55, no. 4,
[5] C. S. Bruning-Fann and J. B. Kaneene, “The effects of nitrate, nitrite and pp. 1331–1338, Aug. 2006.
N-nitroso compounds on human health: A review,” Veterinary Human [29] M. E. E. Alahi, N. Afsarimanesh, S. Mukhopadhyay, L. Burkitt, and
Toxicol., vol. 35, no. 6, pp. 521–538, 1993. P.-L. Yu, “Highly selective ion imprinted polymer based interdigital sen-
[6] G. Ellis, I. Adatia, M. Yazdanpanah, and S. K. Makela, “Nitrite and sor for nitrite detection,” in Proc. 10th Int. Conf. Sens. Technol. (ICST),
nitrate analyses: A clinical biochemistry perspective,” Clin. Biochem., Nanjing, China, 2016, pp. 1–5.
vol. 31, no. 4, pp. 195–220, 1998. [30] J. de Carvalho Silva, J. J. P. C. Rodrigues, A. M. Alberti, P. Solic, and
[7] C. L. Walters, “The exposure of humans to nitrite,” Oncology, vol. 37, A. L. L. Aquino, “LoRaWAN—A low power WAN protocol for Internet
no. 4, pp. 289–296, 1980. of Things: A review and opportunities,” in Proc. 2nd Int. Multidiscipl.
[8] B. Sattelmacher, “Addiscott, T. M., A. P. whitmore and D. S. powlson: Conf. Comput. Energy Sci. (SpliTech), Split, Croatia, 2017, pp. 1–6.
Farming, fertilizers and the nitrate problem. CAB international, 1991; [31] M. T. Penella-López and M. Gasulla-Forner, Powering Autonomous
176 seiten. Paperback, £12.95. ISBN 0 85198 658 7,” J. Plant Nutrition Sensors: An Integral Approach With Focus on Solar and RF Energy
Soil Sci., vol. 156, no. 1, pp. 97–98, 1993. Harvesting. Amsterdam, The Netherlands: Springer, 2011.
[9] E. P. Agency. (Feb. 27, 2017). Ground Water and Drinking Water. [32] M. E. Orazem and B. Tribollet, Electrochemical Impedance
[Online]. Available: https://www.epa.gov/ground-water-and-drinking- Spectroscopy, vol. 48. New York, NY, USA: Wiley, 2011.
water/table-regulated-drinking-water-contaminants [33] B.-Y. Chang and S.-M. Park, “Electrochemical impedance spectroscopy,”
[10] K. S. Johnson, J. A. Needoba, S. C. Riser, and W. J. Showers, “Chemical Annu. Rev. Anal. Chem., vol. 3, pp. 207–229, Jul. 2010.
sensor networks for the aquatic environment,” Chem. Rev., vol. 107, [34] E. Ivers-Tiffée, A. Weber, and H. Schichlein, “Electrochemical
no. 2, pp. 623–640, 2007. impedance spectroscopy,” in Handbook Fuel Cells. Hoboken, NJ, USA:
[11] M. J. Moorcroft, J. Davis, and R. G. Compton, “Detection and deter- Wiley, 2010.
mination of nitrate and nitrite: A review,” Talanta, vol. 54, no. 5, [35] Arduino. (Sep. 4, 2017). Arduino Uno WiFi. [Online]. Available:
pp. 785–803, 2001. https://store.arduino.cc/usa/arduino-uno-wifi
[12] H.-C. Lee, A. Banerjee, Y.-M. Fang, B.-J. Lee, and C.-T. King, “Design [36] Analog Devices. (Sep. 4, 2017). AD5933: Impedance Analyzer.
of a multifunctional wireless sensor for in-situ monitoring of debris [Online]. Available: http://www.analog.com/media/en/technical-
flows,” IEEE Trans. Instrum. Meas., vol. 59, no. 11, pp. 2958–2967, documentation/data-sheets/AD5933.pdf
Nov. 2010. [37] HOPERF Electronic. (Nov. 13, 2017). RFM95/96/97/98(W)—Low
[13] P. Jiang, H. Xia, Z. He, and Z. Wang, “Design of a water environment Power Long Range Transceiver Module V1.0 (V1.0 Edition). [Online].
monitoring system based on wireless sensor networks,” Sensors, vol. 9, Available: http://www.hoperf.com/upload/rf/RFM95_96_97_98W.pdf
no. 8, pp. 6411–6434, 2009. [38] Sparkfun. (Sep. 13, 2000). Dual Full Bridge Driver. [Online]. Available:
[14] T. P. Lambrou, C. C. Anastasiou, C. G. Panayiotou, and https://www.sparkfun.com/datasheets/Robotics/L298_H_Bridge.pdf
M. M. Polycarpou, “A low-cost sensor network for real-time monitoring [39] Dragino. (Sep. 13, 2017). LG01-S LoRa Gateway. [Online]. Available:
and contamination detection in drinking water distribution systems,” http://www.dragino.com/products/lora/item/119-lg01-s.html
IEEE Sensors J., vol. 14, no. 8, pp. 2765–2772, Aug. 2014.
[15] A. Bayo, D. Antolín, N. Medrano, B. Calvo, and S. Celma, “Early detec-
tion and monitoring of forest fire with a wireless sensor network system,”
Procedia Eng., vol. 5, pp. 248–251, Sep. 2010.
[16] Y. E. Aslan, I. Korpeoglu, and Ö. Ulusoy, “A framework for use of wire-
less sensor networks in forest fire detection and monitoring,” Comput.
Environ. Urban Syst., vol. 36, no. 6, pp. 614–625, 2012.
[17] I. Silva, L. A. Guedes, P. Portugal, and F. Vasques, “Reliability Md. Eshrat E Alahi (S’15) was born in Dhaka,
and availability evaluation of wireless sensor networks for industrial Bangladesh. He received the B.Sc. degree in electri-
applications,” Sensors, vol. 12, no. 1, pp. 806–838, 2012. cal engineering from the Military Institute of Science
[18] G. Zhao, “Wireless sensor networks for industrial process monitor- and Technology, Dhaka, in 2007, and the M.Sc.
ing and control: A survey,” Netw. Protocols Algorithms, vol. 3, no. 1, degree in information and automation engineering
pp. 46–63, 2011. from the University of Bremen, Bremen, Germany,
[19] R. Morais et al., “Sun, wind and water flow as energy supply for small in 2013. He is currently pursuing the Ph.D. degree
stationary data acquisition platforms,” Comput. Electron. Agric., vol. 64, in electronic engineering at Macquarie University,
no. 2, pp. 120–132, 2008. Sydney, NSW, Australia.
[20] X. Li, Y. Deng, and L. Ding, “Study on precision agriculture monitoring His current research interests include smart sens-
framework based on WSN,” in Proc. 2nd Int. Conf. Anti Counterfeiting ing systems, Internet of Things, and embedded
Security Identification (ASID), Guiyang, China, 2008, pp. 182–185. systems.

Authorized licensed use limited to: INTERNATIONAL ISLAMIC UNIVERSITY. Downloaded on August 12,2023 at 10:26:30 UTC from IEEE Xplore. Restrictions apply.
ALAHI et al.: IoT ENABLED SMART SENSING SYSTEM FOR NITRATE MONITORING 4417

Najid Pereira-Ishak was born in Sydney, NSW, Lucy Burkitt was born in Melbourne, Australia. She
Australia. He has recently completed his final year received the Bachelor of Agricultural Science degree
of study as an undergraduate student with Macquarie and the Ph.D. degree in soil science from LaTrobe
University, Syndney, graduating with a Bachelor of University, Melbourne, VIC, Australia, in 1998 and
Engineering (with Honours) in mechatronics. 2003, respectively.
He is currently with CSIRO, as a Research She was a Researcher with the University of
Assistant with the Department of Agriculture and Tasmania, Hobart, TAS, Australia, for ten years,
Food. He has interests in Internet of Things (IoT) specializing in soil nutrient management and nutri-
systems, smart systems applications, data acquisi- ent loss. She is currently a Senior Researcher with
tion, and data analysis. Massey University, Wellington, New Zealand, spe-
cializing in reducing the impacts of agriculture on
water quality and high frequency water quality monitoring.
Dr. Burkitt was an Editor of the joint Australian and New Zealand Society of
Soil Science Conference Proceedings in 2012, and a member of the Organizing
Committee for this conference in 2012 and 2016.
Subhas Chandra Mukhopadhyay (A’97–M’99–
SM’02–F’11) received the B.E.E. (gold medalist),
M.E.E., Ph.D., and the Doctor of Engineering
degrees.
He is currently a Professor of mechanical/
electronics engineering with Macquarie University,
Sydney, NSW, Australia, and the Discipline
Leader of the Mechatronics Engineering Degree
Programme. Before joining Macquarie University,
he was a Professor of sensing technology with
Massey University, Wellington, New Zealand. He
has over 26 years of teaching, industrial, and research experience. He has
supervised over 40 postgraduate students and over 100 honors students. He
has examined over 50 postgraduate theses. He has authored or co-authored
over 400 papers in different international journals and conference proceedings,
written 6 books and 36 book chapters, and edited 15 conference proceedings.
He has also edited 25 books with Springer-Verlag and 17 journal special
issues. He has delivered 292 presentations including keynote, invited, tutorial
and special lectures. His current research interests include smart sensors and
sensing technology, instrumentation techniques, wireless sensors and network,
Internet of Things, numerical field calculation, and electromagnetics.
Dr. Mukhopadhyay is a Topical Editor of the IEEE S ENSORS J OURNAL,
and an Associate Editor of the IEEE T RANSACTIONS ON I NSTRUMENTATION
AND M EASUREMENTS . He is a Distinguished Lecturer of the IEEE Sensors
Council from 2017 to 2019. He chairs the IEEE IMS Technical Committee
18 on Environmental Measurements. He has organized over 20 international
conferences as either the General Chair/Co-Chair or the Technical Programme
Chair. He is a Fellow of IET, U.K. and IETE, India.

Authorized licensed use limited to: INTERNATIONAL ISLAMIC UNIVERSITY. Downloaded on August 12,2023 at 10:26:30 UTC from IEEE Xplore. Restrictions apply.