2023 International Conference on Innovative Computing, Intelligent Communication and Smart Electrical Systems (ICSES) | 979-8-3503-1920-0/23/$31.

00 ©2023 IEEE | DOI: 10.1109/ICSES60034.2023.10465538

Electric Vehicle Management and Tracking using

AI and IoT
A.Packialatha1, Bindu Vadlamudi2, B.V. Sai Thrinath3, C.S. Sundar Ganesh4, Mohammad Ishrat5, Archana Bhaskar6
1
Department of Computer Science and Engineering, Vels Institute of Science, Technology & Advanced Studies, Chennai,
Tamil Nadu 600117, India.
2
Department of Electrical and Electronics Engineering, Velagapudi Ramakrishna Siddhartha Engineering College, Vijayawada,
Andhra Pradesh 520007, India.
3
Department of Electrical and Electronics Engineering, School of Engineering, Mohan Babu University, Tirupati, Andhra Pradesh
517102, India.
4
Department of Electrical and Electronics Engineering, Karpagam College of Engineering, Coimbatore, Tamil Nadu 641032, India.
5
Department of Computer Science and Engineering, KoneruLakshmaiah Education Foundation, Vaddeswaram, Guntur, Andhra
Pradesh 522502, India.
6
School of CSA, REVA University, Rukmini Knowledge Park, Bengaluru, Karnataka 560064, India.

E-mail : [email protected], [email protected], [email protected], [email protected],

[email protected], [email protected]

Abstract- Technology has come a long way, and now many data from sensors, networks, actuators, databases, and
problems can be solved with the help of new developments in cutting-edge service delivery mechanisms. AI and IoT
the field. With this technology, we can finally put an end to will highlight the gap between the virtual and physical
the suffocating effects of air pollution. Air pollution is on the worlds, as seen in autonomous vehicles (AVs). This
rise due to the widespread use of automobiles powered by
internal combustion engines. By replacing internal
article examines how AI and IoT could benefit self-
combustion engines with electric motors, we can end our driving cars [2]. VPPs have power generators, energy
reliance on polluting fossil fuels and ensure future storage, and flexible consumers. When correctly
generations have access to clean air. Electric vehicles are implemented, VPP will make EV energy distribution
similar to machines in that they are powered by a more stable and equitable. This study starts an AI-based
rechargeable battery. To what extent an EV functions EV integration system. This system predicts electric
depends on its battery. Voltage, current, and temperature vehicle battery life using federated learning and ANN.
are the indicators of battery health. These factors are used to The methods in this article reduce EV power fluctuations.
calculate the state of charge (SOC). These results are [3].
tracked by a BMS (Battery Management System). This
article details how to create an EV monitoring system that
makes use of AI and IoT to collect data from the vehicle's The primary responsibility of the battery management
battery and send it to the cloud. Using a cloud-based system is to monitor and maintain the battery in optimal
smartphone app, we can track how well the battery is doing. condition in order to fulfill its primary function. It is
possible to extend the lifespan of a car battery by keeping
Keywords—AI, Internet of Things, battery monitoring, voltage track of the number of times it is charged and discharged.
sensor, current sensor, temperature sensor, NodeMCU, and The battery management system (BMS) will monitor the
the cloud are all components of this project. current, temperature, and voltage of the battery. These
variables are used in the calculation of the power output,
I. INTRODUCTION as well as the State of Charge (SoC) and State of Health
(SoH) of the battery. In this study, an Internet of Things is
Electric automobiles (EVs) help reorganize transportation used to monitor the state of the electric vehicle's battery
by removing gasoline-powered autos. Daily EV use [4]. In order to ensure that the battery is balanced in the
promotes a cleaner, healthier environment. There will be appropriate manner, the BMS will do an analysis on both
less air pollution, making it easier to live a carbon-free the battery as a whole and on each individual cell. If the
lifestyle. Because they emit no pollution, EVs will help BMS detects a weak cell within a collection of cells, it
global warming. The UK and most industrialized nations will send an alarm message to the user. At that time, the
like the US use EVs to produce emission-free user should promptly replace the weak cell. The battery
surroundings. Even in poor countries like India, EVs management system (BMS) provides feedback regarding
eliminate emissions. Artificial intelligence (AI) will the temperature, any fluctuations that may have occurred,
program computers to think and act like humans, the voltage and current levels of the system, and the
eliminating the need for humans. Industry relies on AI charging status of the battery [5]. It is vital for both
substantially. Industry 4.0 is driven by the Internet of autonomous and adaptive systems to have an accurate
Things. The Internet of Things [1] collects and processes estimate of the SoC in order to ensure the continuing

Authorized licensed use limited to: VIT University. Downloaded on January 30,2025 at 03:40:41 UTC from IEEE Xplore. Restrictions apply.
health and efficiency of the battery, which will ultimately artificial neural networks (ANN) in some circles. Both of
lead to an extension of its usable life. In this article, we these terms have a significant amount of usage. In order
propose a monitoring system for batteries that makes use to produce predictions based on the battery SOC data, the
of both artificial intelligence (AI) and algorithms in order NN model is utilized. A solid understanding of voltage,
to provide accurate readings of the batteries' levels of current, and temperature is necessary for producing
charge. This method has been demonstrated to be highly reliable estimates of SOC.
efficient and precise on multiple occasions. [6].

The remainder of this article is organized into sections

that explain various aspects of the research process. These
sections include explanations of the methodology that was
proposed in Section II, the hardware devices that were
used in Section III, the software that was utilized in
Section IV, the findings that were presented in Section V,
and the conclusion that was presented in Section VI.

II.PROPOSED WORK
EV battery SOC monitoring is the focus of this article. Fig 1:Model based on neural networks
The combination of AI and IoT makes this possible. After
AI training and evaluation, the SOC prediction model is The data from Mendeley [9], which can be located at this
released to cloud production. Through IoT, physical link, serve as an input for the NN model. The NN model
devices can be connected to the cloud, where consumers is utilized in order to do the task of processing the data
can check battery status via a smartphone app. The battery that was obtained from the Mendeley database. Mendeley
contains temperature, current, and voltage sensors is the source of the data on current, voltage, and
coupled to NodeMCU, a microprocessor. Wireless temperature, which are then processed by an artificial
technology links the microcontroller to the cloud. The neural network (ANN) model. The output of the model is
combination of AI and IoT makes this possible. After AI the estimated state of charge, abbreviated as SOC, of the
training and evaluation, the SOC prediction model is battery.
released to cloud production. Through IoT, physical
devices can be connected to the cloud, where consumers A system that monitors and keeps tabs on electric vehicles
can check battery status via a smartphone app. The battery makes use of both the Internet of Things and artificial
contains temperature, current, and voltage sensors intelligence. AI is able to create predictions regarding the
coupled to Node MCU, a microcontroller. The SOC of an electric vehicle's battery by using the data
microcontroller wirelessly links to the cloud. We use obtained from Mendeley and the model. The data that is
Firebase for research. After the AI-powered SOC is gathered by the Internet of Things is then analyzedusing
finished, IoT deploys it to the cloud, and a mobile app lets certain thresholds that have been established in order to
users check battery health. The first half covers AI, while decide the most effective next step. The numerous
the second covers IoT. sensors, which detect things such as voltage, current, and
temperature, are connected to the battery as well as the
Current, voltage, or temperature can all be used in the CPU. After that, the NodeMCU will connect to the
calculation of EV SOC [7]. With the use of AI, a Firebase cloud via a wireless connection that it will
determination is made on the battery's final SOC. A create. The user is able to check the battery's state of
Neural Network (NN) is utilized in this piece in order to charge (SOC) using a mobile app thanks to the cloud. In
provide an estimate of the remaining SOC in the battery. this scenario, the Internet of Things is utilized to take
AI algorithms will make it straightforward [8] for users to readings of the temperature, current, and voltage of the
manage the batteries of electric vehicles and communicate battery. The information required for this collection
with nearby charging stations to facilitate the generation comes from the sensors that are linked to the battery.
of electricity. The Neural Network for this piece, which Next, the information is uploaded to the cloud storage
can be found in Figure 1. service Firebase by the NodeMCU microcontroller. Each
and every piece of SOC data is uploaded to the cloud
The neural network model that was utilized for this article storage provided by Firebase. Any time the user wants to
can be seen depicted in the graphic that can be found on check the state of charge of their battery, they can do so
the right. Deep learning is a subfield of machine learning, by accessing the cloud through a mobile application. The
and the foundation for deep learning, also known as suggested method for using this system is presented in
neural networks, is the subfield of computer learning Figure 2 in the form of blocks.
known as deep learning (DL). These neural networks are
also known as stimulated neural networks (SNN) and

Authorized licensed use limited to: VIT University. Downloaded on January 30,2025 at 03:40:41 UTC from IEEE Xplore. Restrictions apply.
 Fig 2: A block diagram of the work that is being proposed

Figure 2 presents the block diagram representation of the the voltage of a target item. This can be done by using the
suggested operational approach. Mendeley is used to sensor. The battery's voltage can be determined by the use
gather data, which is then fed into the NN model that is of the voltage sensor, which is a component of the battery
utilized in artificial intelligence. The state of charge of the management system [10]. The battery voltage of an
battery will be utilized during the training and validation electric vehicle is one of the criteria for state of charge
processes of the model. After the model has been put (SOC), which is used to calculate state of charge. A
through its paces in terms of testing, the SOC might then voltage sensor will determine the voltage of a lithium
be considered complete and put into production on the battery that is installed in an electric car, and the
cloud. The Internet of Things has the potential to collect information will then be transmitted to a microcontroller.
data in real time if the EV's sensors are connected to the The sensor is always sending data to the microcontroller
internet. The readings from the sensors are uploaded to so it can keep track of how much voltage the battery is
the Firebase cloud in order to be analyzed. Following this, producing.
the information will be transferred from the cloud to the
SOC model, where it will be merged with the information 1. Second, in order to determine how much electricity is
that the AI model had expected. Continuously collected being supplied to an item, a voltaic device called a
and stored in the cloud is the data that is gathered from "current sensor" is used. EVs have a current sensor
the SOC. The user is presented with a graphical depiction built into the BMS that measures the current flowing
of the cloud data upon which the SOC model relies in through the battery. The battery's current is the most
order to generate predictions through the Firebase mobile crucial variable in determining its SOC. Calculates
app for mobile devices. the battery's SOC by summing the current being
drawn as it is being charged and discharged [11]. The
III.HARDWARE REQUIREMENTS OF THE NodeMCU microcontroller within the current sensor
PROPOSED WORK will receive the data that has been read from the
sensor.
The most important pieces of hardware that are addressed 2. The device's temperature may be determined with the
in this article are the voltage sensors, temperature sensors, use of a temperature sensor. In this work, we utilise
current sensors, and the microcontroller known as the an EV's temperature sensor to take readings from the
Node MCU. The microcontroller and the sensors are battery pack. It is inevitable that the battery's
given connection points so that they can communicate temperature would fluctuate due to the chemical
with one another. The sensor is responsible for collecting processes occurring inside its cells [12]. There is a
sensor data and sending it to the Node MCU in risk of serious harm to the battery if its temperature
accordance with the specifications of the data. The rises over the maximum threshold. The
information is uploaded to the cloud by the microprocessor and temperature sensor are used for
microcontroller, where it is readily available for the user temperature regulation. The data from the
to access at any time. A voltage sensor is a specialized temperature sensor is sent to the NodeMCU
piece of electrical equipment that may be used to measure microcontroller.
3. The Internet of Things platform has given rise to a

Authorized licensed use limited to: VIT University. Downloaded on January 30,2025 at 03:40:41 UTC from IEEE Xplore. Restrictions apply.
 novel type of microcontroller that is known as a Node V.RESULTS AND DISCUSSION
MCU. As open-source software, it can be
downloaded without cost. An ESP8266 Wi-Fi Internal combustion engines (ICEs), which are fueled by
module is incorporated into the System-on-Chip gasoline, diesel, and other fossil fuels, are losing ground
(SoC) Node MCU. Because it contains a Wi-Fi to electric vehicles (EVs), which are progressively
module that is built right in, this microcontroller has gaining in popularity. Electric vehicles (EVs) are
the capability of establishing wireless connections gradually replacing ICEs. The batteries in electric
with any other device. Within the software vehicles may be recharged thanks to the connectivity
environment of the Arduino IDE, instructions are provided by electrical connections. When it comes to
created for the hardware that is attached to the modes of transportation, it is more cost-effective to
microcontroller. The microcontroller and the central simply top off the battery with a charge rather than use
processing unit (CPU) are connected to one another gasoline or diesel. Electric cars greatly cut down on the
by means of a Universal Serial Bus (USB). This USB quantity of pollution that is caused by conventional motor
serves the purpose of moving the program code from vehicles. The operation of electric vehicles (EVs) does
the microcontroller to the CPU of the Node MCU. A not in any way add to the problem of pollution in the
microcontroller functions as a go-between, surrounding environment. Incorporating electric vehicles
connecting the sensors to the Firebase cloud and (EVs) into your day-to-day routine is a step you can take
transmitting data from the sensors to the cloud, where toward making the planet a better place to live in terms of
it can be stored and monitored. In this way, the air quality and overall health. As a direct result of this,
microcontroller acts as a link between the sensors and there will be a sizeable drop in the total amount of
the cloud. pollution in the air, and the natural world will become
more conducive to living a life free of carbon emissions.
IV.SOFTWARE REQUIREMENTS Because they do not emit any pollution, electric vehicles
will have a positive impact on the phenomenon known as
A mobile application known as Firebase Cloud was global warming. In order for the battery to be able to
developed by members of the Google community. With perform its primary purpose, the battery management
the assistance of this cloud, the application may be system's major duty is to oversee the monitoring and
designed, constructed, and improved. Some of the cloud- maintenance of the battery in order to keep it in the best
based services that have been developed include the possible condition. You can increase the life of the battery
storage of files, authentication, push notifications, in your vehicle by keeping track of the number of times it
analytics, database management, and system settings. is charged and discharged. This will allow you to get
Databases that are kept in the cloud and updated in real more miles out of the battery. The current, temperature,
time so that users can access them at any time are called and voltage of the battery will be monitored by the battery
"cloud-based." Because of this characteristic, the cloud management system (BMS), which is also abbreviated as
application is frequently referred to as a "real-time" cloud "BMS." These are the inputs that can be utilized in the
application. Using Firebase, the data may be effortlessly process of calculating the SoC of a battery. Figure 3
saved and accessed from the cloud at any time. Google is depicts the temperature, voltage, and current of the battery
responsible for all aspects of operations and maintenance as they were measured in real time. Additionally, an
related to the Firebase cloud. As a consequence of this, estimate of the amount of charge that is still left in the
the data that is stored in the cloud is safeguarded, which battery is shown in the figure.
makes this mobile app a safer and more secure option
overall. The manner that Firebase is utilized is entirely Figure 3 provides an illustration of the measured
unique in comparison to any other cloud application. It is properties of the electric vehicle's battery. The electric
essential to have both front-end and back-end developers vehicle monitoring system incorporates both artificial
working on cloud-based applications. The disclosed intelligence and the internet of things into its operations.
endpoints of the back end are utilized by the front end of The data that were presented earlier are derived from real
the structure. The backend of the system is where the measurements of the battery. The sensor reports that the
majority of the development work is being done. Google voltage in the environment at this very moment is 3.741
has also developed Firebase Cloud Messaging (FCM) volts. A measurement of the amount of current, in
with the intention of facilitating communication between amperes, that is moving across a certain circuit can be
itself and users. In this investigation, the microcontroller obtained via the current sensor. At this very moment, the
establishes a wireless connection to the Firebase cloud, actual current value is 0.8913 A. The temperature of the
which enables the cloud to later upload the data. Through battery, as measured by a sensor, has been determined to
the usage of the Firebase mobile app, a user is able to be 304.55 degrees Fahrenheit when it is functioning
view a representation of the data that they have stored in properly. These values can be used to obtain an
the Firebase cloud. approximation of SOC. It is anticipated that the current
SOC will be 86%. This information is monitored by the
Firebase cloud mobile app at all times. A button labeled

Authorized licensed use limited to: VIT University. Downloaded on January 30,2025 at 03:40:41 UTC from IEEE Xplore. Restrictions apply.
"History" will serve as the page's final element. A user and actual worth comparison to be displayed as a graph
has the ability to view measured and forecasted SOC on the Firebase cloud service. On the SOC graph, the
statistics from previous days' worth of data. The user is percentage is drawn along the y-axis, while the number of
given the opportunity to examine the AI's forecasts days is plotted along the x-axis. A line graph in green
regarding the state of charge of the battery. color depicts the actual SOC, whereas a graphic in black
color represents the anticipated SOC. Simply entering the
numeric identification for the day in question enables the
user to view the results of that particular day's activities.
After that, the cloud service displays both the projected
and the actual percentage results for that day. The reader
needs to be aware of the results on Day 11 in order to
have a complete understanding of the diagram. Therefore,
both the expected and the actual SOC percentages equal
82% as of day 11 of the study. The user is able to monitor
the state of charge of the electric vehicle's battery in real
time once they have downloaded a cloud-based mobile
app onto their phone.

VI.CONCLUSION
The main goal of this research is to use a metric similar to
the state-of-charge indicator to keep track of the status of
EV batteries. Artificial intelligence (AI) and the Internet
of Things (IoT) are used to bring this to fruition. The
SOC-predicting model is then trained and evaluated with
the help of AI before being put into production in the
cloud. As a result of IoT, physical devices may be
Fig 3: The conclusion reached by the EV monitoring system connected to the cloud, and the user's state of charge can
be tracked in real time through the cloud's mobile app.
The main purpose of a battery management system is to
regulate and track the battery's overall health. To prevent
damage to the battery and ensure the EV can make it to its
destination safely, its charging and discharging processes
are constantly monitored. The BMS will keep tabs on
critical battery information, including current,
temperature, and voltage. With these values, the battery's
SoC may be determined. The SOC was predicted by
utilising the Mendeley information to train and test a NN
model of AI. AI employing a neural network to
implement this system has a 96.2% success rate. Better
battery health monitoring is possible with this setup.

REFERENCES
[1] Costa, Carlos & Barbosa, João& Castro, Hélder&Gonçalves,
Renato &Lanceros-Méndez, Senentxu, “Electric vehicles: To
what extent are environmentally friendly and cost effective? –
Comparative study by european countries”, Renewable and
Sustainable Energy Reviews, vol. 151, pp. 111548, 2021, Doi:
10.1016/j.rser.2021.111548.
[2] Zubi, G.; Dufo-López, R.; Carvalho, M.; Pasaoglu, G. “The
lithium-ion battery: State of the art and future perspectives”.
Renew. Sustain. Energy Rev, 2018, vol. 89, pp. 292–308, Doi:
10.1016/J.RSER.2018.03.002
[3] R, G. et al. (2021). “A Short-Term Solar Photovoltaic Power
Optimized Prediction Interval Model Based on FOS-ELM
Fig 4:Comparison of the SOC Prediction to the Actual Percentage Algorithm” International Journal of Photoenergy, Volume 2021,
Article ID 3981456, 12 pages,
https://doi.org/10.1155/2021/3981456
The expected and the actual values of the SOC are [4] Saravanan, A., Chitra, L., Chandran, S. S., Aravind, B. S.,
presented in Figure 4. It is possible for the SOC forecast Kumar, J. N., Jayaprakash, S., & Ramkumar, M. S. (2022,

Authorized licensed use limited to: VIT University. Downloaded on January 30,2025 at 03:40:41 UTC from IEEE Xplore. Restrictions apply.
 March). Distinguished DC-DC Converter for an Electric [9] Amudha, D. A., Ramkumar, M. S., & Krishnan, M. S. DESIGN
Vehicle. In 2022 6th International Conference on Computing AND SIMULATION OF ZETA CONVERTER WITH ZVZCS
Methodologies and Communication (ICCMC) (pp. 578-582). SWITCHING TECHNIQUE. Journal of Engineering and
IEEE. Applied Sciences, 14(9), 2764-2774.
[5] Ahmed, Z., Zeeshan, S., Mendhe, D., & Dong, X. "Human [10] M. T et al, "A novel based mix-column architecture for AES-
gene and disease associations for clinical-genomics and 128bit algorithm" International Journal of Applied Engineering
precision medicine research," Clinical and Translational Research, ISSN 0973-4562 Vol. 10 No.20 (2015).
Medicine, Vol 10 no 1, pp. 297-318, 2020. [11] V. Ganesan, K. Vijayakumar, V. A. Devi and P. Ramadoss,
https://doi.org/10.1002/ctm2.28 "Hybrid Intelligence for Multimedia Data in Intra IoT (IIoT)
[6] A. Santhi Mary Antony, D. Godwin Immanuel, “An Cloud by Persistent homology," 2022 International Conference
enhancement approach of Reboost converter for unification of on Advances in Computing, Communication and Applied
disparate WECS with the utility grid system”, Elsevier, Informatics (ACCAI), 2022, pp. 1-5, doi:
Environmental Technology & Innovation, Sep 2021, Impact 10.1109/ACCAI53970.2022.9752558.
Factor 7.758, 101953, ISSN 2352-1864, [12] Nagarajan, N Ashok Kumar, JoshuvaArockiaDhanraj, T Senthil
https://doi.org/10.1016/j.eti.2021.101953. Kumar,” Delay Flip Flop based Phase Frequency Detector for
[7] W. Xiao, S. Miao, J. Jia, Q. Zhu and Y. Huang, "Lithium-ion Power Efficient Phase Locked Loop Architecture” International
batteries fault diagnosis based on multi-dimensional Conference on Electronics and Renewable Systems (ICEARS),
indicator," 2021 Annual Meeting of CSEE Study Committee of 2022, pp. 410-414,
HVDC and Power Electronics (HVDC 2021), 2021, pp. 96-101, doi.org/10.1109/ICEARS53579.2022.9752183.
Doi: 10.1049/icp.2021.2544.
[8] Transportation Electrification, vol. 7, no. 1, pp. 26-36, March
2021, Doi: 10.1109/TTE.2020.3006064.

Authorized licensed use limited to: VIT University. Downloaded on January 30,2025 at 03:40:41 UTC from IEEE Xplore. Restrictions apply.