International Journal of Advanced Technology and Engineering Exploration, Vol 8(76)

ISSN (Print): 2394-5443 ISSN (Online): 2394-7454

Research Article
http://dx.doi.org/10.19101/IJATEE.2021.874023

Arduino-based battery monitoring system with state of charge and remaining

useful time estimation
Melbern Rose C. Maltezo 1,2,3*, August C. Thio-ac 3, Anna May C. Castillo 3, Leandro E. Gattu 3,
Carmine Ella A. Hernandez 3, Jonarld John C. Labuan 3, Leonard F. Navales 3, Erickson C. Sopeña
3
, Nilo M. Arago3, Edgar A. Galido3, Gilfred Allen M. Madrigal3, Cherry G. Pascion3, and Lean
Karlo S. Tolentino3, 4
Department of Computer Studies, College of Science, Technological University of the Philippines, Manila,
Philippines1
Information Management & Communication Engineering Department, Integrated Research and Training Center,
Technological University of the Philippines, Manila, Philippines2
Department of Electronics Engineering, College of Engineering, Technological University of the Philippines,
Manila, Philippines3
University Extension Services Office, Technological University of the Philippines, Manila, Philippines 4

Received: 08-February-2021; Revised: 24-March-2021; Accepted: 25-March-2021

©2021 Melbern C. Maltezo et al. This is an open access article distributed under the Creative Commons Attribution (CC BY)
License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly
cited.

Abstract
This paper presents a battery management system for lead-acid battery banks used in e-vehicle. It is incorporated with a
diagnostic, measurement, and monitoring system for improving Lead-acid battery performance up to its efficiency and
conservation. This matter calls the need for research on traction batteries as an insatiate demand exists for smaller
vehicles with lightweight and portable equipment. It is extensive that batteries are strictly assessed and diagnosed before
having them rented or exchanged for their condition to be highly maintained. The measurement of the battery’s State-of-
Charge and State-of-Health is derived from its load voltage, no-load voltage, load current, and temperature during
experimentation. The estimation of State-of-Charge, State-of-Health, Discharge Rate, and Remaining Useful Life are
then derived by utilizing the concept of correlation and regression from the yielded real-time parameters recorded to the
SD card module. This study paves the approach for the comprehensive and continuous progress of battery identification,
monitoring, and diagnosis that is a thorough advancement in the E-Vehicle industry.

Keywords
Battery management system, Lead-acid, Arduino-based management system, Electric vehicle, State of charge, State of
health, Remaining useful time, Discharge rate.

1.Introduction The Lead-acid battery is the type of traction battery

The utilization of commercial transportation is a facet that is commonly used in the Philippines based on the
of economic growth that the nation continues to statistics analyzed by the Electric Vehicle
progress. With developing technologies of Association of the Philippines (eVap), the production
transportation in the Philippines, e-vehicles turned of e-vehicles that use lead-acid batteries are greater
into a trend started last 2015, attracting than those vehicles that use lithium batteries. It is
manufacturers for production [1]. Alongside the practical considering it is low-cost in comparison to
thriving industry for e-vehicles, the traction battery other types [2]. In addition, it is uncomplicated to
industry, as the heart of its construction, will also be assemble as its construction of lead plates and
prominent. The battery in an e-vehicle structure is the electrolyte that is made of diluted sulfuric acid [3].
complement part of the fuel tank on a conventional
vehicle. Traction batteries are the alter for gasoline. Electromechanical batteries are constructed in
distinction to electrical analogy through means of
having a prominent network of electrical components
*Author for correspondence such as electromotive forces and passive components.
432
 International Journal of Advanced Technology and Engineering Exploration, Vol 8(76)

The construction of the cell is formed by two acquire real-time parameters with computational
schemes: design each part as an electrical element or analysis in obtaining the battery’s internal resistance,
set up the cell behavior as a black box. The approach SoC, SoH, Discharge Rate, and Depth-of-discharge
of a black box is the analysis of the terminal’s (DOD). In addition to the indicator installed on the
measurements [4]. Battery modeling can be collected device, which includes the proposed system, with the
from artificial intelligence which fuzzy logic is data acquisition specification of the device the
utilized to layout discharging of lead-acid batteries. researchers developed a user interface to extract and
The data of discharging are acquired through interpret data from the SD card installed in the
controlling the network among the battery’s open- device.
circuit terminal voltage, the output currents, and the
state of charge [5]. The general objective of this study is to develop a
smart Arduino-based e-vehicle battery management
Battery identification and diagnosis is a technology system with the remaining useful time prognosis
that is still not accessible to e-trike owners and e-trike deployed on an e-tricycle battery shop. Specifically,
battery shops in the Philippines. It is significant for this study aims: (1) to design and develop an
batteries to be diagnosed before engaging them into electronic circuit for the Arduino-based data
business for maintaining their good quality. Even acquisition system for the battery’s no-load voltage,
this, owners of e-trike and e-trike battery shops are load voltage, current, and temperature, (2) to develop
only capable of diagnosing the battery by visual an Arduino-based identification and diagnostic
observation for any physical deformity and using a system for batteries that stores and validates battery
common voltmeter for checking the battery’s health. performance through information storage for the
State of Charge (SOC) is commonly confused with initialization, reading, and writing of the battery’s
State of Health (SOH). Both stated techniques are performance history, and (3) predicting its remaining
inadequate assessment and resolution-making would useful life in terms of time.
be at risk.
The study only considered the sets of lead-acid
As a solution, the researchers of the study proposed a batteries in an E-bike Shop used for public utilization
system that exerts a profound understanding in Bacoor, Cavite, Philippines. The output of the
regarding the parameters to scrutinize in diagnosing a study was proposed to e-trike battery shops and even
battery like the Lead-acid battery. The researchers e-trike owners who offer battery renting and/or
specifically aims to: (1) design and develop an swapping. This study was focused on the
electronic circuit for the Arduino-based data identification, diagnosis, and monitoring of lead-acid
acquisition system for the battery’s no-load voltage, batteries that are available within the area. This study
load voltage, current, and temperature, and (2) dealt with the use of SD Card technology for security
develop an Arduino-based identification and over authentication in each-and-every battery
diagnostic system that stores and validates battery available. The goal of the proposed SD Card
performance through information storage for the Identification System is to be able to store large
initialization, reading, and writing of the battery’s amounts of data from the analog signal over a long
performance history used for the battery. period of time so that it can stay unattended for a
long time. Physical battery defects, such as wears and
A Management System solution is proposed to tears or smudges, are not considered as factors for
showcase how important measuring and monitoring diagnosis and monitoring. This study required
the battery SOH and SOC for life span predictions. manpower to ensure the security of data from the SD
Nowadays, lead-acid batteries for e-vehicle are Card till it makes to the database on the shop owner’s
widely used because of their versatility and low cost personal computer.
but degrade their performance without proper
management, measurement, and monitoring. As the The success of the study will be a source for the e-
lead-acid battery ages, its SOH and SOC are trike battery shops for improved diagnosis of battery
susceptible to a lot of factors affecting its life span. and business exhortation. In addition to that, it will
The software development of the proposed system provide a reference for the continuous progress of
used the application of Arduino IDE for Lead-acid battery identification and diagnosis that is
programming the microcontroller of the system, a profound upgrade for the e-vehicle industry.
MySQL for database, and Python for the system user
interface. The microcontroller is programmed to

433
Melbern C. Maltezo et al.

This paper is organized as follows: Section 2 pertains You et al. [10] an illustration about the recurrent
to the gaps and limitations of the related researches, neural network (RNN) which is very suitable in
Section 3 defines the materials and methods used by working with sequential data like the ratio of current
the researchers, Section 4 explains the detection and and voltage during the charge cycle is carefully
database results of the study, and Section 5 declares constructed. There are many advantages with the use
the conclusion and enumerates possible future works of long-short term memory neural networks as the
of the research. improved variant of the standard recurrent neural
network. For the former data's neural network, the
2.Related works performance of a battery is used to gather samples
Recognizable battery degradation assessments have and is necessary to minimize the noise effect.
emerged even before the creation of electric vehicles.
Various methods and variables were introduced for Yang et al. [11] introduce a dynamic Peukerts law-
the estimation of measurements of the battery’s SOC based SOH estimation of batteries. The estimation
and SOH. In [6], researchers introduce a method that method is correlating Peukerts coefficient to the
adopts known cell-balancing circuits for individual battery’s capacity loss. It is computed that the
estimation of cell’s voltage and current from a battery Peukerts coefficient is a function of the battery’s
bank or battery string terminal. The methods include capacity loss. They experimented with seven samples
control strategies and algorithms by manipulating using one type of battery and compared the results
balancing circuits to observe battery subsystems. It is using the proposed method to the same number of
concluded that a large balancing current circuit is sample history data of the same battery type. The
ideal for accurate measurement. proposed method is eligible for direct SOH
estimation with missing data.
Diao et al. [7] determined if batteries have reached
their limit by using internal resistance to evaluate the Dong et al. [12] estimation for short-term SOH and
state of health (SOH). However, due to capacity long-term RUL was proposed using Brownian
independence from the internal resistance, it can Motion and Particle Filter. This study utilizes the
bring about contradicting outcomes for the SOH of degradation modeling using Brownian Motion with
the battery. The SOH is more accurately defined as drift and scale and Particle Filter is used for
the present battery status with relation to the capacity estimation of the drift parameters. Two sets of
and power SOH, in which degradation, internal different capacity of battery degradation in terms of
resistance, and inconsistency of capacity are all distance traveled were experimented for validating
considered. With the use of this method, a clear the results of the method. The proposed forecasting
advantage can be seen by analyzing data and by of the mentioned two parameters was verified to be
comparing it to other methods. accurate. The limitation of the method is that the
parameters are initialized off-line with the use of
Liu et al. [8], researchers utilized probability available measurements.
distribution and adopted the concept of Monotonic
echo sate Networks or MONESNs algorithm for Qu et al. [13] proposed a model for estimation of
tracking nonlinear degradation patterns of battery- SOH and prediction of RUL using a synergy of long
RUL estimation. A correlation model between health short-term memory (LSTM) network and particle
index and battery capacity is developed. Two sets of swarm. The optimization of key parameters of the
data of lithium-ion batteries are used to prove the particle swarm was employed. In forecasting of RUL
efficiency of the proposed method. of the lithium-ion batteries, the Complete Ensemble
Empirical Mode Decomposition with Adaptive Noise
Hou et al. [9] propose calculation and monitoring of (CEEMDAN) method is utilized to denoise raw data
the electric vehicle's SOH, SOC, and state of function for extracting the downtrend of SOH. Real-life
(SOF). The estimation of SOC shows that the datasets from NASA of the lithium-ion batteries life
maximum error is 0.334%. The present maximum cycle were used for validation. The results from the
capability to minimize the error of the SOC experimentation present that the proposed method is
estimation will therefore be redefined. Furthermore, better than the predictions of RNN, LSTM, and RVM
this article states that the SOF is established based on in monitoring SOH and prediction of RUL of
the SOC and the SOH will yield the system's driving lithium-ion batteries. The application of an accurate
power. and low computational demanding state-of-health
(SOH) estimation algorithm represents a key
434
 International Journal of Advanced Technology and Engineering Exploration, Vol 8(76)

challenge for the battery management systems in impedance. Dung proposed a time-constant-ratio
electric vehicle (EV) applications. Stroe and Schaltz measurement to improve issues on different SOH
proposed the suitability of the incremental capacity estimations that were mentioned.
analysis (ICA) technique for estimating the capacity
fade and subsequently the SOH of LMO/NMC-based 3.Methodology
EV Lithium-ion batteries. During the eleven months In this research, the battery management system was
of testing, ageing results were collected and they deployed in a battery rental shop for public
were able to accurately relate the capacity fade of the transportation. Figure 1 shows the block diagrams of
studied batteries to the evolution of the voltage value, the entire system. Batteries, with and without the
which corresponds to one of the incremental capacity proposed device for identification and measurement,
(IC) valleys, obtained using the ICA technique [14]. are subjected to be manually operated by the
personnel of the battery shop. The specifications of
Topan et al. [15] avoiding battery failure and keeping new batteries are stored in the database by the user
the battery lifespan, a system should be controlled by interface, from the user interface a new SD card will
considering several parameters of Battery undergo decoding for the battery specifications. The
Management System (BMS) such as State of Charge SD card module is applied as the battery
(SOH) and State of Health (SOH). The State of identification that collects data for every charging
Charge in Battery Management System provides the and discharging of the battery. The measuring system
percentage of battery capacity, while the State Of is an Arduino-based circuitry designed as a
Health measures the battery health. Polarization voltmeter, ohmmeter, ammeter, and thermometer. In
characteristic, dynamic behaviour of the battery and the diagnostic system, the data from the SD card
estimation through Kalman Filter (KF) are defined by module, battery specifications and the measured
the use of Thevenin battery model. Parameters in the parameters were all transferred to the database of the
model were accumulated using Recursive Least system. The computation and prediction of the
Square. remaining useful time of the battery were taken from
a mathematical model using the measured values.
Dung et al. [16] Dung proposed an SOH estimation The monitoring system comprises a display of real-
system based on time constant-ratio measurement. time measured parameters, the produced critical
Time-constant represents the response speed of parameter, the SOC and the estimation of the battery
terminal voltage during charging and discharging and useful time by means of hours and minutes. The data
the change of internal chemical reactions. There have acquired from the Lead-acid battery over time is also
been different issues regarding traditional SOH presented in the display to show the performance of
estimations. The first SOH estimation is a full- the battery in every charging and discharging cycle.
charge-capacity (FCC) estimation. Unfortunately, to The Figure 2 is the system architecture that illustrates
access FCC information, it required a long-term the connectivity of the technology to be used in this
charging and discharging tests and therefore failed on study.
estimation speed. The second estimation is internal
impedance estimation. However, this SOH estimation
may cause large error because of environmental

Figure 1 System block diagram

435
Melbern C. Maltezo et al.

Figure 2 System architecture

3.1Research locale an Arduino-based circuitry is used to be the

Shops in Bacoor, Cavite, Philippines that offer rental equipment to measure real-time parameters of the
and/or swapping of lead acid batteries were located. battery. Initial diagnosis will undergo the Arduino-
For that opted location, the study device was used on based voltmeter, ohmmeter, ammeter, and
battery banks used in e-trikes for testing and battery thermometer circuits that will be stored in the
diagnosis. The acquired data and evaluation would be integrated microSD card module. The integration of
disclosed to the owners of the business. the microSD card module functions also as the
unique identification for the system. It will be given a
3.2Hardware development serial number for a set of the battery bank. The flow
The hardware development consists of measuring and chart illustrated in Figure 3 is the process of data
identification technology. In the measuring section, transfer in the device.

Figure 3 Data transfer in the device

436
 International Journal of Advanced Technology and Engineering Exploration, Vol 8(76)

3.2.1Identification system 3.3Software development

Batteries, with and without the proposed device for The software development of the proposed system
identification and measurement, are subjected to be used the application of Arduino IDE for
manually selected by the owner of the battery shop. programming the microcontroller of the system,
The specifications of new batteries (without the MySQL for database and Python for the system user
device) are stored in the database then, from the user interface. The microcontroller is programmed to
interface a new SD card will undergo decoding for acquire real-time parameters with computational
the battery specifications. The SD card module analysis in obtaining the battery’s internal resistance,
served as the identifier of the battery that also records SoC, SoH, Discharge Rate, and Depth-of-Discharge
data every charging and discharging. (DOD).
3.2.2Measuring system
Measuring the battery’s parameters was real-time and Figure 4 shows the data transfer on the device alone
saved on the SD card contained in the prototype. The of SOC and SOH Measurement. With the process-
measuring system is an Arduino-based voltmeter, flow presented measured real-time variables are also
ohmmeter, ammeter, and thermometer circuit used programmed to be shown on the LCD. Meanwhile,
for the diagnosis and estimation of battery life. In the Figure 5 is the illustration of the user interface flow
measuring system, an Arduino-based voltmeter, chart. The user interface must be manipulated by
ohmmeter, ammeter, and thermometer circuit capable personnel to choose whether to diagnose a battery
of measuring the parameters needed in diagnosis and bank that is already on the database (with device) or
the battery end-of-life prediction is used. Two clips not. MySQL technology will be used for the
were connected to the positive and negative terminals development of databases. If a battery has an SD card
of the lead-acid battery for measuring no-load module on it, then it is already subjected to diagnosis
voltage, load voltage, internal resistance, and current by taking its measured parameters and storing the
of the battery. A thermal sensor was used for data for monitoring.
measuring the battery bank temperature and its
surrounding.

Figure 4 Measurement of SOC and SOH flow chart

437
Melbern C. Maltezo et al.

Figure 5 User interface flow chart

3.4Computational analysis capacity. Fully charged new battery SOCo is equal to

The researchers classify the health of the battery as 100% since QR is same as QM, as declared in
they know critical parameters, the SOC and SOH. Equation 1.
SOC is the ratio of the charge remaining to the SOCo = (QM / QM) x 100% (3)
battery to its fully charged state. In alternative, SOC SOCo = 100% (4)
is the percentage of the charge remaining in the
battery to the maximum charge of the battery. To The SOC can be computed also using cell voltage
calculate the SOC: (VCELL) and the standard battery voltage prior from
SOC = (QR / QM) x 100% (1) having internal resistance (VO), expected from the
where, direct proportionality of charge and voltage.
SOC, State-of-Charge of the battery SOCV = (5)
QR, Remaining Charge in the battery
QM, Maximum Charge of the battery
VCELL = VBAT – VINTERNAL RESISTANCE (6)
SOC = SOCo – ( ∫ ) x 100% (2) where,
where, VCELL, Cell Voltage
SOC0, Initial State-of-Charge of full charge battery VBAT, Battery Voltage (12 V)
VIR, Battery Voltage having Internal Resistance
The initial state of charge (SOCo) of new fully
charged batteries can be simply computed The battery voltage will increment for a certain value
considering the new battery will pass its maximum (α, normally 0.003V) to each cell of a battery bank
438
 International Journal of Advanced Technology and Engineering Exploration, Vol 8(76)

when the battery temperature is below the ambient SOHR (13)

temperature and reciprocally, where n is equal to the
number of cells.
The degradation of the battery is the decrease of the
SOCV = (7) SOH over a period of time.

From Equation 2 and Equation 5, where SOCV = 3.5Device setup

SOCo, an equation would be derived: Figures 6 to 9 present the different views of the
SOC = ( – ∫ )×100% (8) developed device for the proposed battery
management system by the proponents. Dimensions
are measured in terms of inches (in). The following
where,
figures show the integration of the material into a
VMeasured, Voltage measured by BattMan
N, Number of Battery/Cell compact device.
α, Voltage increment (at 0.003 V)

The initial discharge (QINITIAL DISCHARGE) is equal to 0

for new fully charged battery so that,
∫ (9)

Using Equation 9 to Equation 8, Cn is the original

capacity (QO),

SOC = ( – )x 100%
(10)
where, Figure 6 BattMan front view
QO, Original Charge Capacity of the battery

The remaining useful time of the battery is calculated

in terms of hours and minutes. The taverage is equal to
the average operation time of the battery that is
normally 6 hours.
tremaining = taverage x SOC (11)

where,
tremaining, Remaining time of the battery in hours and
mintes
taverage, Average Operation of the battery
Figure 7 BattMan right side view
Another parameter to be calculated for assessing the
health of the battery is the SOH. There are various
ways to calculate the SOH such as using internal
resistance, voltage, or charge.
SOHR
(12)
where,
REOL, End-of-Life Resistance of the battery
RIR, Rated Internal Resistance of the battery

The SOHR is the state of health internal resistance of

the battery, REOL is the resistance of the battery to
reach its end of life, and RO is the rated internal
resistance of the battery. Figure 8 BattMan left side view

439
Melbern C. Maltezo et al.

batteries became a challenge which is why the actual

DOD gathered was slightly different than the
expected. Listed also were the mean and maximum
current from 1st cycle up to the 10th cycle.

Table 3 shows the third experimentation for State-of-

Health and Depth of Discharge between BattMan
3M1 and BattMan 3M2. Also, the expected DOD
upon experimentation was 90%, 60% and 30% but
strict monitoring on the batteries was still a challenge
due to enhanced quarantine protocol which made it
almost the same as the second experimentation.

Table 4 shows the new sets of data like Days Used

and Predicted Remaining Service Life. Used Days is
the number of days passed from the purchase date.
Figure 9 BattMan back view Predicted EOL is defined as the number of days till
the battery wears out. The prediction is based on the
4.Results current remaining SOH, which is determined thru the
4.1Data and results battery’s load voltage, no-load voltage, mean current,
Table 1 and 2 are the two experiments between maximum current, maximum temperature, and DOD.
batteries using varied current. Table 1 is the From the five (5) sample batteries, five sets of data
measurements with low current and Table 2 is the gathering/report were taken out based on a different
measurements with high currents. The expected DOD age.
upon experimentation from Table 1 and 2 was 90%,
60% and 30%. However, strict monitoring on the

Table 1 First testing: state-of-health and depth of discharge

Battery BATT MAN-N1 BATT MAN-N2 BATT MAN-N3
Decreased in SOH (%) 1st Cycle 0.09 0.09 0.09
5th Cycle 0.09 0.08 0.09
10th Cycle 0.08 0.09 0.09
Average DOD (%) 91.93 61.46 31.78
Average Current (A) 1.29 1.33 1.26

Table 2 Second testing: state-of-health and depth of discharge

Battery BATTMAN-3M1 BATTMAN-3M2
Report-1 0.14 0.18
Report-2 0.14 0.16
Report-3 0.13 0.17
Decreased in SOH (%)
Report-4 0.13 0.18
Report-5 0.13 0.17
Average DOD (%) 60.46 91.18

Table 3 Third testing: state-of-health and depth of discharge

Battery BATTMAN-3M1 BATTMAN-3M2
Report-1 0.15 0.17
Report-2 0.14 0.17
Decreased in SOH (%) Report-3 0.14 0.18
Report-4 0.14 0.16
Report-5 0.13 0.13
Average DOD (%) 61.07 90.40
Average Current (A) 13.42 13.40

440
 International Journal of Advanced Technology and Engineering Exploration, Vol 8(76)

Table 4 Data report: predicted EOL of different tested batteries

Battery Report number Report date Days used Predicted EOL
48 23-Jun-20 7 169
54 24-Jun-20 8 167
BATTMAN-N2 60 27-Jun-20 11 162
66 29-Jun-20 13 160
72 2-Jul-20 16 155
46 22-Jun-20 30 147
58 26-Jun-20 34 141
BATTMAN-1M 67 29-Jun-20 37 136
74 3-Jul-20 41 130
77 5-Jul-20 45 127
49 23-Jun-20 60 117
61 27-Jun-20 64 111
BATTMAN-2M 70 30-Jun-20 67 106
75 3-Jul-20 71 101
78 5-Jul-20 73 98
45 22-Jun-20 90 78
51 25-Jun-20 93 73
BATTMAN-3M2 57 26-Jun-20 94 70
63 28-Jun-20 96 67
69 30-Jun-20 98 64
52 25-Jun-20 120 57
64 28-Jun-20 123 52
BATTMAN-4M 73 2-Jul-20 127 45
76 3-Jul-20 128 43
79 5-Jul-20 130 39

4.2Data of report
In addition to the indicator installed on the device The Second experimentation is about the
with the data acquisition specification of the device measurement of battery degradation between
the researchers developed a user interface to extract BattMan1 and BattMan2 using high current. Just like
and interpret data from the SD card installed in the the first experimentation, BattMan N1 had 60% DOD
device. Stated technologies: MySQL and Python and BattMan N2 had 90% DOD. Both BattMan N1
were used for interface development. The interface is and BattMan N2 are brand new but have undergone
programmed to diagnose the data for the prediction series of charge and discharge. For a better
of EOL as presented on Table 4. comparison of the said batteries, data were gathered
simultaneously. It is observed that the SOH of
5.Discussion BattMan N2 decreased greater than BattMan N1.
Calculating the Remaining Useful Life of the e- With high current applied, the Decrease in SOH
vehicle battery in terms of cycle and time using between the two batteries is different as well as
SOHR is used, since the degradation of the battery is DOD.
proportional to the number of cycles used. Cycles 1,
5 and 10 were chosen to serve as the initial, middle In the third experimentation, two used batteries (3-
and final experimental setup. The first month old) were utilized. Both batteries also
experimentation is to make a comparison of the displayed different DOD like BattMan N1 and
battery’s degradation; for BattMan1, BattMan2, and BattMan N2. The data gathered from the used
BattMan3, using low current. Each battery has a batteries and the brand-new ones were almost the
different average Depth of Discharge. In each battery, same. The measurement acquired was validated
three (3) SOH samples were taken on different against a multimeter as no government body or
cycles. It is greatly observed that the Decrease in private entity available for checking. Since the study
SOH is very little among the three batteries due to the is a pioneer in the battery industry, its reference or
small difference in current. Also, the big difference in standards being followed are basically from a
the DOD among the three batteries didn’t affect the multimeter.
Decrease in SOH.

441
Melbern C. Maltezo et al.

6.Conclusion and future work Acknowledgment

In determining the battery’s End-of-Life, parameters The University Research and Development Services of the
like no-load voltage, load voltage, mean current, Technological University of the Philippines is
acknowledged for providing funding and support.
maximum current, maximum temperature, DOD, and
Charge- Discharge cycle should be carefully Conflicts of interest
measured. Large battery consumption or DOD will The authors have no conflicts of interest to declare.
occur if the discharge current is low. It showed in the
data that DOD doesn’t affect the battery life since the References
discharged current was low. However, based on the [1] http://boi.gov.ph/wp-
data, with low discharge current and low DOD, the content/uploads/2018/03/Electric-Vehicle-Industry-
battery’s SOH continues to decrease due to its Roadmap.pdf? fbclid=IwAR0y1Q-
charge-discharge cycle. SOH still decrease no matter UnwulEIySuyKszxtV3uRhCoe_dDTYAQJhgu23CT0
how efficient the battery was used. If the battery was 5pM-zYiCGbhU. Accessed 30 January 2019.
discharged with a high current, the decrease in SOH [2] Buchmann I. Batteries in a portable world: a handbook
on rechargeable batteries for non-engineers, Fourth
is great and evident. The data showed that battery Edition. U.S. Department of Energy Office of
aging doesn’t affect the decrease in SOH. The Scientific and Technical Information 2002:1-292.
battery gets consumed easily as the SOH decreases [3] Lipschultz A. Batteries in a portable world.
continuously. In addition, the increase in the charge- Biomedical Instrumentation & Technology. 2015;
discharge cycle of the battery contributed to the 49(2).
decrease in SOH. The temperature, when high, [4] Azzollini IA, Di Felice V, Fraboni F, Cavallucci L,
affects battery consumption and may also cause Breschi M, Dalla Rosa A, et al. Lead-acid battery
serious damage like overheating. High discharge modeling over full state of charge and discharge
current was found to consume the battery easily. It range. IEEE Transactions on Power Systems. 2018;
33(6):6422-9.
was confirmed that battery life was prolonged when [5] Burgos C, Saez D, Orchard ME, Cárdenas R. Fuzzy
the e-trike kept constant speed than speeding up then modelling for the state-of-charge estimation of lead-
stopped and vice versa. It was observed that e-trike acid batteries. Journal of Power Sources. 2015;
with varying speed consumes high current which 274:355-66.
greatly degrades the battery life. Also, overloading an [6] Polis MP, Yin GG, Chen W, Fu Y, Mi CC. Battery
e-trike experience is the same. The battery life was cell identification and SOC estimation using string
also prolonged when the difference between the terminal voltage measurements. IEEE Transactions on
ambient (external) and internal temperature was kept Vehicular Technology. 2012; 61(7):2925-35.
at 10 degrees Celsius. Not completely exhausting the [7] Diao W, Jiang J, Zhang C, Liang H, Pecht M. Energy
state of health estimation for battery packs based on
battery and recharging it when 20% charge remains the degradation and inconsistency. Energy Procedia.
added to prolonging the battery life. 2017; 142:3578-83.
[8] Liu D, Xie W, Liao H, Peng Y. An integrated
For further research related to this study, here are probabilistic approach to lithium-ion battery
some of the recommendations from the proponents: remaining useful life estimation. IEEE Transactions
● Estimation of Remaining Useful Distance. The on Instrumentation and Measurement. 2014;
study only developed BattMan capable of 64(3):660-70.
predicting the Remaining Useful Life in terms of [9] Hou ZY, Lou PY, Wang CC. State of charge, state of
time and not what’s left to travel. health, and state of function monitoring for EV BMS.
In international conference on consumer electronics
● More sample data acquisition per second. The 2017 (pp. 310-1). IEEE.
group had a hard time gathering data because of [10] You GW, Park S, Oh D. Diagnosis of electric vehicle
the pandemic. batteries using recurrent neural networks. IEEE
● Testing of different battery capacities. Transactions on Industrial Electronics. 2017;
Manufacturers have produced e-trikes of higher 64(6):4885-93.
battery capacity for larger accommodation. [11] Yang A, Wang Y, Tsui KL, Zi Y. Lithium-ion battery
● Real-time data acquisition to the database using SOH estimation and fault diagnosis with missing data.
IoT. This study requires human force to ensure the In international instrumentation and measurement
technology conference 2019 (pp. 1-6). IEEE.
security of data from the SD Card till it makes it to
[12] Dong G, Chen Z, Wei J, Ling Q. Battery health
the database on the shop owner’s personal prognosis using Brownian motion modeling and
computer. particle filtering. IEEE Transactions on Industrial
● Coordinate trackers for e-trike. Electronics. 2018; 65(11):8646-55.

442
 International Journal of Advanced Technology and Engineering Exploration, Vol 8(76)

[13] Qu J, Liu F, Ma Y, Fan J. A neural-network-based Leandro E. Gattu finished Bachelor of

method for RUL prediction and SOH monitoring of Science in Electronics and
lithium-ion battery. IEEE Access. 2019; 7:87178-91. Communication Engineering major in
[14] Stroe DI, Schaltz E. SOH estimation of LMO/NMC- Microelectronics at Technological
based electric vehicle lithium-ion batteries using the University of the Philippines, Manila
incremental capacity analysis technique. In energy last 2020.
conversion congress and exposition 2018 (pp. 2720-5).
IEEE.
[15] Topan PA, Ramadan MN, Fathoni G, Cahyadi AI, Email: [email protected]
Wahyunggoro O. State of Charge (SOC) and State of
Health (SOH) estimation on lithium polymer battery Carmine Ella A. Hernandez finished
via Kalman filter. In international conference on her studies in high school in Our Lady
science and technology-computer 2016 (pp. 93-6). of the Pillar Catholic School. She
IEEE. received her Bachelor of Science in
[16] Dung LR, Wu SH, Yuan HF. An SOH estimation Electronics Engineering major in
system based on time-constant-ratio measurement. In Communications degree at
23rd international symposium on industrial electronics Technological University of the
2014 (pp. 1784-7). IEEE. Philippines, Manila last 2020.
Email: [email protected]
Dr. Melbern Rose C. Maltezo is a
Professor at the Integrated Research Jonarld John C. Labuan received his
and Training Center of the bachelor's degree in Electronics
Technological University of the Engineering from Technological
Philippines. She is the current Head of University of the Philippines, Manila in
the Information Management & 2020. He is currently pursuing Data
Communication Engineering Engineering under Dubbed Smarter
Department. Her research and Philippines through Data Analytics
publication interest includes Radio Frequency Research and Development, Training
Identification, Sensor Application and Monitoring Systems. and Adoption (Project SPARTA) by the Department of
Email: [email protected] Science and Technology – Philippine Council for Industry,
Energy and Emerging Technology Research and
August C. Thio-ac is currently taking Development (DOST-PCIEERD). His interests include
up his Doctor of Philosophy in building and managing of data workflows, pipelines, ETL
Technological Management at processes, and platforms and mainly responsible for the
Technological University of the management of the entire data lifecycle: ingestion,
Philippines, Manila. He finished Master processing, surfacing, and storage.
of Engineering major in Electronics Email: [email protected]
Engineering and Bachelor of Science in
Electronics and Communication Leonard F. Navales finished his
Engineering at Technological University of the Philippines, primary education at Bagong Barrio
Manila. He passed the Electronics and Communications Elementary School and completed his
Engineering Licensure Examination in November 1999 and secondary school at Quezon City
Licensure Examination for Teachers in August 2004. He Academy. He passed the board exam
also passed PhilNITS examinations Information and received his license in Electronics
Technology Passport (IP) in April 2018 and Fundamental Technician last October 2018. He
IT Engineer (FE) in October 2019 received his bachelor’s degree in
Email: [email protected] Electronics Engineering at Technological University of the
Philippines, Manila, Last May 2020.
Anna May C. Castillo completed her Email: [email protected]
secondary school at Navotas National
High School. She passed the board Erickson C. Sopeña finished
exam and earned her license as Computer Engineering Technology, a
Electronics Technician last October 3-year technology course at
2018. She took Bachelor of Science Technological University of the
degree in Electronics Engineering at Philippines, Cavite last 2017. He
Technological University of the passed the board exam and received his
Philippines – Manila and graduated last May 2020. license as Electronics Technician last
Email: [email protected] October 2018. He also finished his

443
Melbern C. Maltezo et al.

Bachelor of Science in Electronics Engineering major in Cherry G. Pascion is currently an

Communication at Technological University of the Assistant Professor in the Electronics
Philippines, Manila and graduated last 2020. Engineering Department at
Email: [email protected] Technological University of the
Philippines (TUP). She graduated from
Nilo M. Arago is an Associate her Master’s in Engineering Program at
Professor II at Technological TUP and Bachelor of Science in
University of the Philippines Manila. Electronics and Communication
He graduated with his Master of Engineering (BSECE) at Lyceum of the Philippines
Engineering in 2011 and Bachelor of University (LPU). She gained her industry experienced
Science in Electronics and working as a Quality Assurance (QA) Engineer at a PCB
Communications Engineering in 2001. Manufacturing Company. She was also a former faculty
His expertise is more on sensor member of Far Eastern University Institute of Technology
Engineering and Technology Applications. (FEU-FIT) and Lyceum of the Philippines University
Email: [email protected] (LPU). As teaching and research are her passion, she was
with the academe for more than 20 years.
Edgar A. Galido is currently a Faculty Email: [email protected]
Member of the Department of
Electronics Engineering at Lean Karlo S. Tolentino was the Head
Technological University of the of the Department of Electronics
Philippines - Manila. He graduated his Engineering, Technological University
Bachelor of Science in Electronics and of the Philippines (TUP) in Manila
Communications Engineering at Cebu from 2017 to 2019. He is currently the
Institute of Technology. Director of the University Extension
Email: [email protected] Services Office and Assistant Professor
at TUP since 2019. He is a graduate of
Bachelor of Science in Electronics and Communications
Gilfred Allen M. Madrigal is Engineering at TUP last 2010 and a graduate of Master of
currently a Faculty Member of the Science in Electronics Engineering major in
Department of Electronics Engineering Microelectronics at Mapua University last 2015. He is
at Technological University of the currently pursuing his PhD in Electrical Engineering at
Philippines (TUP). He is currently National Sun Yat-Sen University, Taiwan. He is currently
taking his Master of Engineering at one of the members of the Technical Committee on Audio,
TUP Manila. He is currently one of the Video, and Multimedia Equipment (TC 59) and Technical
members of the Technical Committee Committee on Electromagnetic Compatibility (TC 74) of
on Cleanrooms (TC 83) of the Bureau of Product Standards the Bureau of Product Standards of the Philippines’
of the Philippines’ Department of Trade and Industry. Department of Trade and Industry. He is a Regular Member
Email: [email protected] of the National Research Council of the Philippines. He is
an active member of IEEE and IECEP. His research
interests include microelectronics and information and
computing technologies.
Email: [email protected]

444