Smart Water Quality Monitoring System
A.N.Prasad, K. A. Mamun, F. R. Islam, H. Haqva
School of Engineering and Physics
University of the South Pacific
Laucala, Fiji Islands
Abstract— Nowadays Internet of Things (IoT) and
Remote Sensing (RS) techniques are used in different area of
research for monitoring, collecting and analysis data from
remote locations. Due to the vast increase in global industrial
output, rural to urban drift and the over-utilization of land
and sea resources, the quality of water available to people has
deteriorated greatly. The high use of fertilizers in farms and
also other chemicals in sectors such as mining and
construction have contributed immensely to the overall
reduction of water quality globally. Water is an essential
need for human survival and therefore there must be
mechanisms put in place to vigorously test the quality of
water that made available for drinking in town and city
articulated supplies and as well as the rivers, creeks and
shoreline that surround our towns and cities. The availability
of good quality water is paramount in preventing outbreaks
of water-borne diseases as well as improving the quality of
life. Fiji Islands are located in the vast Paciﬁc Ocean which
requires a frequent data collecting network for the water
quality monitoring and IoT and RS can improve the existing
measurement. This paper presents a smart water quality
monitoring system for Fiji, using IoT and remote sensing
Keywords—Smart Water Quality Monitoring; Internetof
Things; Remote Sensing.
Over the past few decades, waters in and around Fiji
have gradually succumbed to a fair degree of pollution.
Chemical waste and oil spills are the major, primary forms
of water pollution threatening Fiji’s waterways. For
example, an article published in the Fiji Times on 24
December, 2014 reported on raw sewage seeping into the
Samabula River at a rate of 200 liters per second due to
broken pipes . Eliminating pollution altogether may
seem like an unfathomable notion but limiting its effects
when it does happen is certainly possible. The primary
objective of this project is to devise a method to monitor
seawater quality in an effort to aid in water pollution
control in Fiji with the help of IoT and RS technology.
The Smart Water Quality Monitoring System will
measure the following water parameters for analysis;
Potential Hydrogen (pH), Oxidation and Reduction
Potential (ORP), Conductivity and Temperature using a RS
technology. While monitoring these parameters, it is
perceived that one should receive a stable set of results.
Therefore a continuous series of anomalous measurements
would indicate the potential introduction of a water
pollutant and the user will be notified of this activity with
the aid of IoT technology. False positives, such as
anomalous readings over a short period of time, will be
recorded but not treated as an alert. Hence, with the
successful implementation of this monitoring approach, a
water pollution early warning system can be achieved with
a fully realized system utilising multiple monitoring
Initiatives have been taken all over the globe to develop
projects based on sampling water to aid in controlling
marine environments. It may not be specific to water
pollution monitoring but similar concepts are involved.
Libeliums Smart Water device monitors the status of an
aquarium’s health in Europe . It specifically monitors
parameters like pH, electro conductivity,
oxidation/reduction potential (ORP) and temperature. A
cloud based solution is developed to help in monitoring
data in real time providing a fast and effective reaction in
case of rising abnormalities.
A similar example to that of this project can be seen in
the coastal water pollution monitoring initiative in the Gulf
of Kachchh  with the only difference being in terms of it
having a much larger scope and vastly more expensive
protocols deployed to counter the effects of the industrial
Furthermore, locally there have been projects based
around the conservation of the coral reefs. The Mamanuca
Environment society’s (MES) Biannual Sea Water
Monitoring Program has been around for 4 years whereby
tests are carried out on seawater for faecal coliform (FC)
bacteria, salinity and nutrients which helps in ascertaining
the health of the surrounding reefs .
Research indicates that projects of this nature are
developed on a large scale with generous funding from
reputable organizations. There is little indication of small-
scale and inexpensive projects that have a similar role in
places like marine jetties, cities and industrial rivers to
preserve aquaculture and public health. By applying a
strategic, cheap and methodical technique this project
hopes to achieve this in an effort to sanitize our oceans.
III. INTERNET OF THINGS
The internet of Things (IoT) is a revolutionary new
concept that has the potential to turn virtually anything
“smart”. A Thing in this context could be defined as an
object such as a cardiac monitor to a temperature sensor.
This extraordinary event has captured the attention of
millions. Why is this so big today? So imagine a world
where machines function without any notion of human
interaction. A future where machines communicate with
other machines and make decisions based on the data
collected and all independent of an end user.
To understand how this revolution took shape we have
to travel back to the 1900’s with a profound prediction
from a well renowned inventor Nicolai Tesla in which he
stated that the world will be wirelessly connected to a
single brain. Every invention starts with a simple thought,
that’s all it takes to define history. Alan Turing, the
inventor of the computer, spoke about machines having
sensors and humans teaching the machines, what we know
today as Artificial Intelligence (AI). Then came the World
Wide Web (www), the flow of information that is available
to the public and this was exactly what was missing to
realise Teslas prediction. The term itself “internet of
things” was coined in 1999 by Kevin Ashton for linking
the idea of sensors with the internet . The IoT journey
has taken over a century to see light and it will undoubtedly
not stop here.
Fig. 1: The proposed schematic diagram of the smart
water quality monitoring system.
It might be difficult to see the significance of the IoT
but every advancement made is to make everyday life
simpler and safer. Examples of these are a baby monitor to
keep track of a baby’s health in real time , an IoT for
caregivers which collects behavioral data to improve care
 and a heart monitoring system that collects biometrics
data to track an aging patient’s health . These are just a
few examples of how IoT projects can improve the way of
life. Fig. 1 shows the proposed schematic diagram of the
smart water quality monitoring system using the IoT
The first task and a very integral one was to determine
which water parameters would provide a close indication
for water pollution. Through extensive research [9-11] the
parameters were chosen to be composed of pH, oxidation
and reduction potential (ORP) and temperature. The
reasoning behind these selections is discussed in section V
– Water Parameters.
Independently these parameters provide very little
information in terms of how polluted the seawater actually
is. Therefore, analysis will consider collective parameter
behavior in order to generate a valid output, which is either
polluted or not unpolluted. To put this into perspective, a
drop in pH of tap water alone is not a valid indication of
pollution, this only indicates a formation of acids but it
may still be consumable (e.g adding lemon juice to tap
The second step was the selection of locales that will
provide useful data. The area in question should be
susceptible to some chemical fluctuations by either marine
life or human interference since performing data readings
on clean, untouched waters would produce known results.
Therefore, the locations were narrowed down to industrial
areas, marine jetties, sewer waste openings and city lines
where human interference had a considerable impact.
Given that security was a factor, the site was chosen as the
USP jetty since the area is completely secure from theft and
The third obstacle was which form of data logging
would produce an acceptable format. An FTP solution was
developed initially on a local network, however without the
intervention of local Internet Service Providers this seemed
like the least convenient option. A cloud server has also
been considered to act as an intuitive and a more permanent
solution. Work is still in progress on this matter. Moving
on, since the equipment has an SD storage option, data
logging was ultimately done on the hardware itself in text
format which can easily be read by practically any
The final step was to decide on an acceptable, proficient
and accurate form of analysis. Seeing as the sea contains a
vast number of unknowns which will imminently
chemically alter the properties being measured. This will in
turn present erroneous readings. As previously mentioned,
changes in one measured parameter may be no indication
of the sea water actually being in the presence of
pollutants. The collective measured results had to be
consistent over period of time to be treated as a possible
threat. Moreover, to overcome this obstacle an intelligent
analytical system had to be designed in the manner of a
Neural Network model.
V. WATER PARAMETERS
It is important to record temperature alongside the
other parameters as this will be useful in behavioral
analysis of the parameters being measured. Relating to
temperature-relation theories, pH and conductivity have an
undesirable effect with large temperature changes. In
addition to this, extreme temperatures for pacific island
climates is of understandable concern.
The pH of a solution is the measure of the acidity or
alkalinity of that solution. The pH scale is a logarithmic
scale whose range is from 0-14 with a neutral point being
7. Values above 7 indicate a basic or alkaline solution and
values below 7 would indicate an acidic solution. The
majority of aquatic life prefers a pH level of 6.5 – 9.0.
Anything outside of this optimum range is considered fatal
to the marine ecosystem. Extreme pH values also increase
solubility of elements and compounds making them toxic
and therefore more likely to be absorbed by marine life.
Furthermore, temperature has an inverse relationship with
pH that is, as temperature increases pH levels decrease and
C. Oxidation- Reduction Potential
Oxidation-Reduction Potential is the measure of a
solutions oxidizing power. In simple terms it can be
described as the potential of a solutions ability to sanitize
itself. Higher ORP values would indicate more oxidizers
present. Likewise lower ORP equals more reducers
It is understood that a typical good value for aquatic
life should be in the vicinity of 100 mV to 200 mV.
Anything outside these limits is a case to be investigated.
The same can be said about tap water whose ORP levels
are very high, usually in the high 600 mV because of the
use of disinfectants such as chlorine. Anything outside this
range should be investigated.
Conductivity signifies the ionic strength of a solution.
In other words it is the ability of a solution to conduct
electricity with the typical unit for measurement being
micro-Siemens per centimeter (uS/cm). As the dissolved
ions increase in the water, conductivity increases.
Therefore, the conductivity of tap water is perceptibly low
at around 100 uS/cm. On the other hand, expected values
for sea water are 55000-60000 uS/cm due to its high ionic
content. Any further increase in the conductivity value
may be indicative of polluted waters, such as sewer leaks
or chemical wastes flooding into the water.
Moreover, conductivity is directly related to salinity
that is conductivity improves with high salinity.
Conductivity values outside of the optimum levels indicate
a possible negative scenario. Dead Sea is a prime example
of lethal concentrations of salt.
The temperature relation with conductivity is a
proportional one. A general assumption of a temperature-
conductivity relation is taken to be linear in nature with a
deviation of 2%/°C.
VI. OVERALL STRUCTURE
The main concept behind this project can be described
in a simple block diagram shown in Fig. 2 and Fig. 3
Shows the setup with the sensors and Waspmote
The system as a whole comprises of sensors, an
analogue to digital convertor (ADC), a microcontroller, an
SD storage and a GSM module. The data collected can
either be stored onboard via the SD card or sent to a File
Transfer Protocol (FTP) server or a cloud server. In the
case of this project, a cloud server in conjunction with a
local machine is utilised for data analysis.
A complete bundled set provided by Libelium  was
used for this project and included the sensors,
microcontroller and GSM communication. Furthermore,
since the deployment duration is expected to run for
months or even years, power conservation is imperative. To
achieve this, the system design incorporates sleep mode i.e
the system gets a 15 minutes sleep time after an hour of
continuous readings. To further extend battery life any idle
modules have been set to off mode. For instance, when an
SD card operation has finished the SD module switches off.
The same is realized with the GSM and serial
communication. Further to this, alerts have been set to
notify the user of certain conditions such as battery life and
Fig. 3: Deployment Setup with the sensors and Waspmote
VII. EXPERIMENTAL RESULTS
Four water samples from different water sources were
tested to establish a reference on the parameters for each
water type. The chosen water types were seawater, surface
water, Tap water and polluted creek water.
The four water samples were tested simultaneously
with four separate, identical systems at indoor ambient
temperature. Readings were taken at 1 hour intervals for a
total period of 12 hours. For security reasons the systems
were not deployed in the specific areas of interest, instead
water samples were collected and tested in a safe controlled
environment. However, the tap water sample was changed
every hour to see the consistency of Fiji tap water (supplied
by Fiji Water Authority) readings.
A. Reference for tap water
Fig. 4-7 shown the trends of the acquired data and are
consistent with the globally accepted values for pH,
conductivity and ORP. The temperature effect on pH and
conductivity is clearly observed.
Fig. 2 : Shows the overall block diagram of system operation.
Fig. 4: Graph of the solutions ambient temperature.
Fig. 5: Conductivity for tap water is shown.
Fig. 6: PH trend for tap water.
Fig. 7: ORP trend for tap water.
In addition, Temperature- Conductivity relation can be
seen to be linearly proportional.
B. Reference on sea water
A sample of fresh seawater, collected from the
shores of Sigatoka, was tested to provide a reference
on healthy sea water with little to no contamination.
Fig. 8: Temperature trend for seawater.
Fig. 9: Conductivity trend for seawater.
Fig. 10: PH trend for seawater.
Fig. 11: ORP trend for seawater.
The results shown in Fig. 8-11 indicate values
that are near to the researched data for acceptable sea
water parameters that can sustain aquaculture .
C. Reference on surface water
A sample of water was taken from Rewa River
(Suva, Fiji) to provide a reference on surface water.
The results obtained (Fig. 12-15) from Rewa River
were also consistent with the researched data available
on acceptable surface water parameters .
Fig. 12: Temperature trend for Surface water.
Fig. 13: Conductivity trend for surface water.
Fig. 14: PH trend for surface water.
Fig. 25: ORP trend for surface water.
D. Polluted water test
To get a fair idea of how the parameters of polluted
water should look a sample of water was collected from
the Nabukalou Creek an extremely polluted waterway in
the heart of Suva City. The results are shown in Fig. 16-
Fig. 36: Temperature trend for Nabukulau creek.
Fig. 47: Conductivity trend for Nabukalou Creek.
Fig. 58: PH trend for Nabukalou Creek
Fig. 69: ORP trend for Nabukulau Creek.
E. Summary of the tested properties between the water
A comparison can be made with the collected data
between tap water, river water, seawater and polluted
creek water. The pH levels for all were fairly similar with
the only change being in relation with temperature.
Conductivity for the water samples differed significantly
because of the different salinity concentrations for
different water types. The highest conductivity being
58000 uS/cm for sea water and the lowest being that of tap
water with conductivity value of 58 uS/cm. ORP for sea
water and river water were similar with results being in the
low 100-200mV range. ORP for tap water was observed to
be 350 mV which is fine considering that the acceptable
range is from 300-650mV.
The data obtained for polluted water has some
interesting values for ORP and conductivity. A very low
ORP value was observed, averaging at -2mV which is an
indication of overpowering reductants. This is an expected
value considering the background of Nabukalou Creek
having waste lines connected to the creek. The
conductivity value was in the 40000 range indicating that
water samples likely contained traces of pollution. A
summary is also presented in table format shown in Table
TABLE I. SUMMARIZED RESULTS
This research demonstrates a smart water quality
monitoring system. Four different water sources were
tested within a period of 12 hours at hourly intervals to
validate the system measurement accuracy. The results
obtained matched with the expected results obtained
through research. The temperature relation with pH and
conductivity were also observed for all the water samples.
GSM technology has been successfully implemented to
send alarm based on reference parameter to the ultimate
user for immediate action to ensure water quality.
Additionally, the parameter references obtained from all
the different water sources will be used to build classifiers
which will be used to perform automated water analysis in
the form of Neural Network Analysis.
In a nutshell, the system has proved its worth by
delivering accurate and consistent data throughout the
testing period and with the added feature of incorporating
IoT platforms for real time water monitoring, this should
be an excellent contender in real time water monitoring
The research team would like to thank the School of
Engineering at the University of the South Pacific for
funding this research project.
 Madigibuli, A., “Sewere Spill”, Retrieved from Fiji Times Online:
http://www.fijitimes.com/story.aspx?id=289915, Accessed on:
November 4, 2015.
 Libelium., “Water Quality Monitoring in Europe’s Largest Fluvial
Aquarium”, Retrieved from Libelium:
fluvial-aquarium-zaragoza/, Accessed on: November 4, 2015.
 “Sea Water Monitoring For Chemical Parameters”, Retrieved from
Gujarat State Project Management Unit:
parameters.aspx, Accessed on: November 4, 2015.
 Vakacola, M., “ Biannual Sea Water Monitoring Program”, Retrieved
from Mamanuca Environment Society: http://mesfiji.org/biannual-
sea-water-monitoring-program, Accessed on: November 4, 2015.
 Postscapes, “Retrieved from History of the Internet of Things”,
Retrived from Online: http://postscapes.com/internet-of-things-
history, Accessed on: November 4, 2015.
 K. A. Mamun, Sharma, A., A. S. M. Hoque, T. Szecsi, “Remote
Patient Physical Condition Monitoring Service Module for iWARD
Hospital Robots”, Asia-Pacific World Congress on Computer
Science and Engineering, 2014.
 Atzori, Luigi, Antonio Iera, and Giacomo Morabito. "The internet of
things: A survey." Computer networks 54, no. 15, 2787-2805, 2010.
 “Preventive Solutions”, Retrieved from
http://www.preventicesolutions.com, Accessed on: November 4,
 Luke Mosley, S. S., “Water Quality Monitoring in the Pacific Island
Countries”, Report, University of the South Pacific, 2005.
 3riversquest.org., “Sampling Parameters and Descriptions”, Online:
3riversquest.org, Accessed on: November 4, 2015.
 Ministry of health, Fiji Island, Fiji National Drinking Water Quality
0 to -3mV