3C Tecnología. Glosas de innovación aplicadas a la pyme. ISSN: 2254 – 4143 Ed. 38 Vol. 10 N.º 2 Junio - Septiembre 2021
117
A COST-EFFECTIVE SIMPLIFIED ENERGY MONITORING SYSTEM
USING IOT
Pierre E. Hertzog
Centre for Sustainable Smart Cities, Department of Electrical, Electronics and Computer Engineering.
Central University of Technology, Bloemfontein, (South Africa).
E-mail: phertzog@cut.ac.za ORCID: https://orcid.org/0000-0002-3396-6050
Arthur J. Swart
Centre for Sustainable Smart Cities, Department of Electrical, Electronics and Computer Engineering.
Central University of Technology, Bloemfontein, (South Africa).
E-mail: aswart@cut.ac.za ORCID: https://orcid.org/0000-0001-5906-2896
Recepción:
28/04/2021
Aceptación:
4/06/2021
Publicación:
14/06/2021
Citación sugerida:
Hertzog, P. E., y Swart, A. J. (2021). A cost-eective simplied energy monitoring system using IOT. 3C Tecnología.
Glosas de innovación aplicadas a la pyme, 10(2), 117-133. https://doi.org/10.17993/3ctecno/2021.v10n2e38.117-133
3C Tecnología. Glosas de innovación aplicadas a la pyme. ISSN: 2254 – 4143 Ed. 38 Vol. 10 N.º 2 Junio - Septiembre 2021
118
https://doi.org/10.17993/3ctecno/2021.v10n2e38.117-133
ABSTRACT
The management of any process really requires sensing, transferring and recording of relevant data for
analysis purposes so as to make an informed decision. In fact, smart manufacturing requires sucient
knowledge to be available to the decision-maker to ensure that any system runs eciently and eectively,
which may include an energy generation system. The purpose of this paper is to present a simplied
energy monitoring system, and its acquired data, that may be used to achieve such real-time results that
can be used to make an informed decision. An experimental methodology is applied where empirical
quantitative data is derived from a residential home in a suburb of a city having a population of around half
a million people. Nine electrical circuits are monitored systematically using a current sensor where data
is transferred to Google Sheets using an Arduino microprocessor and ESP8266 module. Custom made
power usage graphs are directly visible via any electronic device, for real-time results. It is recommended
to make more use of such cost-eective simplied systems in an attempt to provide sucient knowledge
to make an informed decision that may contribute to identifying energy discrepancies in any given
system with the aim of improving energy eciency.
KEYWORDS
Arduino, Cloud, Energy monitoring, IoT, Cost-Eective.
3C Tecnología. Glosas de innovación aplicadas a la pyme. ISSN: 2254 – 4143 Ed. 38 Vol. 10 N.º 2 Junio - Septiembre 2021
119
https://doi.org/10.17993/3ctecno/2021.v10n2e38.117-133
1. INTRODUCTION
“When I look at the solar system, I see the earth at the right distance from the sun to receive the proper
amounts of heat and light. This did not happen by chance” (Brainy Quote, 2020). These words, by
Sir Isaac Newton, a world-renowned English Mathematician, well indicate that things do not happen
by chance. There must be a designer, a creator, or a maker of some kind to bring forth a system that
contains order and stability. In fact, to ensure the ongoing stability, or sustainability, of any designed
system requires good management, and not mismanagement.
The term ‘management’ can be dened as a process of planning, organizing, directing, coordinating
and controlling material and non-material, or technology and physical resources toward achieving
organizational objectives (Amadike, 2018). The most appropriate denition is of course contingent upon
the context in which it is used. However, a key concept should always include monitoring any progress
or change.
Eective monitoring requires the systematic collection and analysis of appropriate data (Chapman, 2017)
that needs to be transformed into knowledge by linking it with other pockets of information to foster
understanding and create meaning (Swart, 2018). Systematic collection needs to be done over a period
of time using scientically validated techniques, where data is processed and presented in meaningful
graphs that may contribute to making informed decisions regarding the performance and health of any
given concept, or system. In engineering systems, this sensing, transferring, recording and analysing of
data is an integral, and indispensable part, of the management process.
How though can this process be implemented in a cost-eective and simplied manner to ensure that
systematic data is automatically processed and presented in meaningful graphs over an extended period
of time? This is indeed important, as the manual processing of data is often labour-intensive, time-
consuming and extremely costly (Weinstein, 2018). The purpose of this paper is to present a simplied
energy monitoring system, and its acquired data, that may be used to present real-time results that can be
3C Tecnología. Glosas de innovación aplicadas a la pyme. ISSN: 2254 – 4143 Ed. 38 Vol. 10 N.º 2 Junio - Septiembre 2021
120
https://doi.org/10.17993/3ctecno/2021.v10n2e38.117-133
used to make an informed decision about the performance and health of a given system. An experimental
methodology is applied where empirical quantitative data is derived for the energy usage in a residential
home of a city having a population of around half a million people. The paper rstly commences with a
brief discussion of the concept ‘smart’ and ‘IoT’ (Internet of Things). The experimental setup, research
site, methodology and results then follow. Succinct conclusions round o the paper.
1.1. THE BIRTH OF ‘SMART’ AND ‘IOT’
The term ‘smart’ as been applied in many dierent ways, and to many dierent concepts. It has been
stated that it refers to the collection, storage and evaluation of data (Genennig, 2020) where improvements
through digital technology can be applied to any process or service (Stolikj et al., 2016). It really implies
the automatic collection of desired information that needs to be transformed into meaningful graphs that
can be accessed from anywhere and at any time. This information should then be used to automatically
execute a command or enable a user to manually perform a required action.
The term “smart” was perhaps rst used in the 1990s, with the development of the information society,
where the primary emphasis was on ICT (Information and Communication Technologies) in relation
to modern infrastructures and their possible impact on society (Das, 2019). Some may assert that the
term was made famous by the mobile phone industry with the advent of smartphones during the birth
of the 21st century. Common terms now include “smart city”, “smart campus”, “smart home” and even
“smart student”.
Consider just the term “smart city” A Google Scholar search for this exact term reveals some 136 000
results. Performing an advanced search where only the title of the article is considered reveals 11 900
results, of which almost 9 000 occur within the last 5 years (as on 18 March 2020). This term has indeed
become a hot topic, or a “buzzword”, in academia, business, industry and society at large.
3C Tecnología. Glosas de innovación aplicadas a la pyme. ISSN: 2254 – 4143 Ed. 38 Vol. 10 N.º 2 Junio - Septiembre 2021
121
https://doi.org/10.17993/3ctecno/2021.v10n2e38.117-133
One could almost say that the term IoT is the twin brother or twin sister of the term ‘smart’. Many
believe that the term IoT was rst coined by Kevin Ashton in 1999 (Darwish et al., 2019), a well-
known British technology pioneer (Tyagi et al., 2019) to refer to a large set of named things that are
interconnected (Alagar & Wan, 2019). Even though the term has been circulating for the past two
decades, no commonly accepted denition exists. However, many denitions contain three key concepts,
namely sensors, communication and storage. Some or other sensor is used to collect specic data that
needs to be communicated to a centralized storage location from where it can be accessed anywhere and
at any time.
So, key similarities do exist between the term ‘smart’ and ‘IoT’. However, one could say that the term
‘IoT’ would focus more on the sharing of information, where the term ‘smart’ would go one step further
and include some or other form of control. Nevertheless, both terms should bring to the mind four key
systematic steps, namely sensing, transferring, recording and presentation of relevant data. This may be
accomplished by using a cost-eective simplied system, as outlined next.
1.2. EXPERIMENTAL SETUP
A cloud-based energy monitoring system was designed as shown in Figure 1. Ten identical current
transformer sensors were used to sense ten dierent electrical circuits in an electrical distribution board of
a residential home in a city having a population of around half a million people. The current sensors were
coupled to a logging interface circuit that conditions the sensed current signals so that it can be interpreted
by the analogue-to-digital converters on an Arduino Mega microcontroller. The microcontroller receives
and processes the received signals from the login interface circuit, communicating power values for each
individual circuit via serial communication to the ESP8266 WiFi module. The software on this module
enables connection to a local WiFi network which facilitates the transfer of data from the Arduino Mega
to Google Sheets using a Google script. The recorded data is then analysed and presented in standard
graphs available in Google sheets.
3C Tecnología. Glosas de innovación aplicadas a la pyme. ISSN: 2254 – 4143 Ed. 38 Vol. 10 N.º 2 Junio - Septiembre 2021
122
https://doi.org/10.17993/3ctecno/2021.v10n2e38.117-133
Figure 1. Block diagram of the experimental setup.
Source: own elaboration.
Figure 2 shows the logging interface circuit as well as the ESP8266 WiFi module. The pink wires
connect the logging interface circuit to 10 x 3.5 mm audio jacks that provide a connection to the current
transformer sensors.
3C Tecnología. Glosas de innovación aplicadas a la pyme. ISSN: 2254 – 4143 Ed. 38 Vol. 10 N.º 2 Junio - Septiembre 2021
123
https://doi.org/10.17993/3ctecno/2021.v10n2e38.117-133
Figure 2. Logging interface circuit and WiFi module.
Source: own elaboration.
In Figure 3, a part of the electrical distribution board of a residential home can be seen. The current
transformer (CT current sensor) sensors were rstly coupled to the output of the circuit breakers with
no physical interruption in the electrical wires required. This limits the possibility of electrical shock
occurring to the installer, as no screwdriver or physical wire connection is required. The transformer
current sensor is simply clipped around the relevant current-carrying electrical wire.
Figure 3. Current sensors and circuit breakers.
Source: own elaboration.
3C Tecnología. Glosas de innovación aplicadas a la pyme. ISSN: 2254 – 4143 Ed. 38 Vol. 10 N.º 2 Junio - Septiembre 2021
124
https://doi.org/10.17993/3ctecno/2021.v10n2e38.117-133
The current transformer sensors are then plugged into the logging interface circuit where the sensed
data is sent to a WiFi module via the Arduino Mega microcontroller. The software program that resides
on the Arduino Mega microprocessor includes access details to the local WiFi network of the home, that
needs to be modied for other WiFi networks at dierent research sites.
2. METHODOLOGY
The research site for this study is located in Bloemfontein, in the centre of South Africa. This is a semi-
arid region which is characterized by deep red or yellow sandy loam soils, low annual rainfall, high
temperatures and high evaporation (Barnard & Smith, 2009). It has average daily radiation of between
4.5 and 7 kWh/m2 with an average annual rainfall of approximately 550 mm (Mulaudzi et al., 2012).
Bloemfontein was founded in 1896 and currently has an average population of 594 428 residents
(Mangaung Metro Municipality). It is the capital city of the Free State Province of South Africa. In
Dutch, "Bloem" means "ower" and "fontein" means "fountain" (SouthAfrica.TO, 2020). It is also known
as the "City of Roses" or Mangaung (which means "place of Cheetahs" in Sesotho). Bloemfontein is also
home to South Africa's Appeal Court.
An experimental study is undertaken where data was collected for 6 months, from June 2018 to December
2018. Experimental studies are usually subject to many factors, eg, materials, equipment, and costs, and
therefore limit the research speed (Zhang et al., 2018). The equipment associated with this study includes
nine current transformer sensors, a logging interface circuit, an Arduino Mega microprocessor and an
ESP8266 module that tallies a total cost of around $130. Ten electrical circuits in an averaged sized
residential home was sampled. This home consists of a living area (including 3 bedrooms), two guest
rooms, laundry room, an oce, a kitchen and an outside aquaponics system. Two geysers rated at 2 kW
each, also exist on the property.
3C Tecnología. Glosas de innovación aplicadas a la pyme. ISSN: 2254 – 4143 Ed. 38 Vol. 10 N.º 2 Junio - Septiembre 2021
125
https://doi.org/10.17993/3ctecno/2021.v10n2e38.117-133
3. RESULTS
The experimental setup was used to record the power consumption of nine electrical circuits in an
average sized residential home. In order to make informed decisions regarding the power usage, the
sampling period for the data was set to 10 minutes. Figure 4 presents the power consumption recorded
for a one-day period. The kitchen used 17.8% of the total power, which excludes the stove that was
monitored separately (3.1% for the stove that brings the total power consumption for the kitchen to
20.9%). The aquaponics system, featuring two electrical pumps, consumed 20.8% of the total power for
that day. The oce that also includes an air conditioner consumed 16.9% of the total power.
The two hot water geysers used 12.4 and 9.2% of the total power. This is dierent from previous
research published in 1992 that showed that some geysers consume approximately 45% of the total
electricity used in homes (Meyer et al., 1992). However, a 2016 study in Norway showed that newer
geysers consume 10% of household electricity consumption (Shan et al., 2016). The lower percentage
power usage by the geysers may be attributed to the fact that the geysers are well isolated and that the
total power used would be higher than a home that does not have an aquaponics system, that consumed
one fth of the total power.
The total power consumption for both geysers dedicated to water heating was 21.6%. The rest of the
power consumption was spread between the guest rooms (11.1%), laundry room (1%) and the living area
(7.7%).
Figure 5 shows the total recorded power curve for the one-day period. The maximum power that was
drawn around 4 a.m. can be attributed to the timer that switches both geysers on as residents awake and
start preparing for the days’ work ahead. A second signicant spike around 1:30 p.m. can be attributed
to appliances in the kitchen as lunch is prepared.
3C Tecnología. Glosas de innovación aplicadas a la pyme. ISSN: 2254 – 4143 Ed. 38 Vol. 10 N.º 2 Junio - Septiembre 2021
126
https://doi.org/10.17993/3ctecno/2021.v10n2e38.117-133
Figure 4. Distribution of recorded power consumption for one day.
Source: own elaboration.
The percentage power consumption distribution of the nine electrical circuits for a six-month period is
visible in Figure 6. It is noteworthy that its distribution is similar to that for the six-month period. This
relates to the validity of the simplied cost-eective energy monitoring system and reliability of the data.
Reliability refers to the consistency, stability and repeatability of results while validity refers to the degree
to which a test, or system, appears to measure what it claims to measure (Mohajan, 2017). For instance,
the one-day consumption for the guest room was 11.1% in Figure 4, and 10.7% for the six-month period
in Figure 6. The aquaponics system power consumption varied with 0.5% from 20.8% for one-day to
20.3% for the six-month period. Similar results are observed for the living area.
3C Tecnología. Glosas de innovación aplicadas a la pyme. ISSN: 2254 – 4143 Ed. 38 Vol. 10 N.º 2 Junio - Septiembre 2021
127
https://doi.org/10.17993/3ctecno/2021.v10n2e38.117-133
Figure 5. Recorded power prole for one day.
Source: own elaboration.
There was a slightly larger dierence in the power consumption for the kitchen, being 17.8% on one-
day and 14.1% over the six-month period. However, a signicant dierence in power consumption was
determined for geyser 2, being 9.2% for one-day and 14.8% over the six-month period. This may be
attributed to the fact that this geyser is older than geyser one and may not be as well isolated. New water
heaters (or geysers) are much more energy-ecient than older models and replacing one that is 7 years
old is a good idea (Smith, 2019). Almost all new units have a 2-inch-thick polyurethane foam in their
jacket that contributes signicantly to an improvement in thermal insulation.
3C Tecnología. Glosas de innovación aplicadas a la pyme. ISSN: 2254 – 4143 Ed. 38 Vol. 10 N.º 2 Junio - Septiembre 2021
128
https://doi.org/10.17993/3ctecno/2021.v10n2e38.117-133
Figure 6. Distribution of recorded power for a six-month period.
Source: own elaboration.
Furthermore, the six-month period was from June 2018 to December 2018, which included a signicant
portion of the winter period. Geysers are known to require higher amounts of thermal energy in winter
than in summer, due to the temperature dierence of the water that needs to be heated (Kalogirou,
2009). More energy consumption would therefore be required in the winter period (June to July) than in
the summer period (November to December). The percentage dierence between the on-day and six-
month power consumption data for the nine circuits can be seen in Figure 7.
3C Tecnología. Glosas de innovación aplicadas a la pyme. ISSN: 2254 – 4143 Ed. 38 Vol. 10 N.º 2 Junio - Septiembre 2021
129
https://doi.org/10.17993/3ctecno/2021.v10n2e38.117-133
Figure 7. The percentage difference between the one-day and six-month power consumption data.
Source: own elaboration.
4. CONCLUSIONS
The purpose of this paper was to present a simplied energy monitoring system, and its acquired data,
that may be used to present real-time results that can be used to make an informed decision about
the performance and health of a given system. The results from this study showed that the energy
monitoring system is indeed reliable, as consistent power consumption readings were obtained over a
period of six-months that reect the power consumption of a one-day period. This was determined for
the living room, two guest rooms, laundry room, oce, kitchen and an aquaponics system. However, a
5.6% dierence was determined for the main geyser in the home, that may be attributed to its age or the
change in season, where more electrical energy is required in winter than in summer to heat the water
to a pre-determined temperature. It is recommended to make more use of such cost-eective simplied
systems in an attempt to provide sucient knowledge to make an informed decision that may contribute
to identifying energy discrepancies in any given system with the aim of improving energy eciency.
3C Tecnología. Glosas de innovación aplicadas a la pyme. ISSN: 2254 – 4143 Ed. 38 Vol. 10 N.º 2 Junio - Septiembre 2021
130
https://doi.org/10.17993/3ctecno/2021.v10n2e38.117-133
REFERENCES
Alagar, V., & Wan, K. (2019). Understanding and Measuring Risk due to Uncertainties in IoT. In
2019 IEEE International Conference on Smart Internet of Things (SmartIoT), 484-488. https://doi.
org/10.1109/SmartIoT.2019.00088
Amadike, C. M. (2018). The Eect Of Recruitment And Selection Policy As A Tool For Achieving Higher Employee
Productivity In Manufacturing Organizations: A Study Of Dozzy Group Of Companies (Thesis). Amadike,
Chinaza Marypamela. http://eprints.gouni.edu.ng/571/
Barnard, A. & Smith, M. F. (2009). The eect of rainfall and temperature on the preharvest sprouting
tolerance of winter wheat in the dryland production areas of the Free State Province. Field Crops
Research, 112(2-3), 158-164. https://doi.org/10.1016/j.fcr.2009.02.011
Brainy Quote. (2020). Homepage. http://www.brainyquote.com/quotes/
Chapman, A. R. (2017). A “Violations Approach” for Monitoring the International Covenant on
Economic, Social and Cultural Rights 1. Human Rights. Routledge.
Darwish, A., Hassanien, A. E., Elhoseny, M., Sangaiah, A. K., Muhammad, K., & Computing,
H. (2019). The impact of the hybrid platform of internet of things and cloud computing on
healthcare systems: opportunities, challenges, and open problems. Journal of Ambient Intelligence, 10,
4151-4166. https://doi.org/10.1007/s12652-017-0659-1
Das, D. (2019). Smart City. In Orum, A. (ed.) The Wiley Blackwell Encyclopedia of Urban Regional Studies.
John Wiley and Sons.
Genennig, S. M. (2020). Research Context: Foundations of the Research. Realizing Digitization-Enabled
Innovation. Springer.
3C Tecnología. Glosas de innovación aplicadas a la pyme. ISSN: 2254 – 4143 Ed. 38 Vol. 10 N.º 2 Junio - Septiembre 2021
131
https://doi.org/10.17993/3ctecno/2021.v10n2e38.117-133
Kalogirou, S. (2009). Thermal performance, economic and environmental life cycle analysis
of thermosiphon solar water heaters. Solar energy, 83(1), 39-48. https://doi.org/10.1016/j.
solener.2008.06.005
Mangaung Metro Municipality. (2019). Integrated Development Plan 2018/19 http://www.mangaung.
co.za/2019/04/09/draft-integrated-development-plan-2019-20-idp-sectoral-plans-mtref-
budget-2019-20-2021-22-budget-related-policies/
Meyer, J. P., & Greyvenstein, G. (1992). Inuence of price changes on the viability of heat
pumps for heating water in South African homes. Energy conversion, 33(1), 41-49. https://doi.
org/10.1016/0196-8904(92)90145-M
Mohajan, H. K. (2017). Two criteria for good measurements in research: Validity and reliability. Annals
of Spiru Haret University. Economic Series, 17(4), 59-82. https://doi.org/10.26458/1746
Mulaudzi, S. K., Muchie, M., & Makhado, R. (2012). Investigation of the solar energy production
and contribution in South Africa: research note. African Journal of Science, Technology, Innovation and
Development, 4, 233-254. https://www.researchgate.net/publication/262527564_Investigation_
of_the_Solar_Energy_Production_and_Contribution_in_South_Africa
Shan, K., Wang, S., Yan, C., & Xiao, F. (2016). Building demand response and control methods for
smart grids: A review. Science and Technology for the Built Environment, 22(6), 692-704. https://doi.org
/10.1080/23744731.2016.1192878
Smith, B. (2019). Tucson regional strategies toward a more sustainable home. The University of Arizona.
Southafrica.TO. (2020). Bloemfontein https://www.southafrica.to/provinces/FreeState/towns/
Bloemfontein/Bloemfontein.php
Stolikj, M., Lukkien, J. J., Cuijpers, P. J., & Buchina, N. (2016). Nomadic service discovery in smart
cities. Smart Cities and Homes. Elsevier.
3C Tecnología. Glosas de innovación aplicadas a la pyme. ISSN: 2254 – 4143 Ed. 38 Vol. 10 N.º 2 Junio - Septiembre 2021
132
https://doi.org/10.17993/3ctecno/2021.v10n2e38.117-133
Swart, A. J. (2018). An Analysis of Master Dissertations: A Case Study of Central University of
Technology, South Africa. AJLAIS, African Journal of Library, Archives and Information Sciences, 28(2),
211-223. https://www.ajol.info/index.php/ajlais/article/view/183188
Tyagi, S., Joshi, M., Ansari, N., Singh, V., & Data, B. (2019). Impact of IoT to Accomplish a Vision
of Digital Transformation of Cities. In Solanki, V. K., Díaz, V. G. & Davim, P. (eds.) Handbook of
IoT.
Weinstein, B. G. (2018). A computer vision for animal ecology. Journal of Animal Ecology, 87(3), 533-545.
https://doi.org/10.1111/1365-2656.12780
Zhang, Y., Zhang, N., & Ge, C. (2018). First-Principles Studies of Adsorptive Remediation of Water
and Air Pollutants Using Two-Dimensional MXene Materials. Materials, 11(11), 2281. https://
doi.org/10.3390/ma11112281
