Development of a Compact Self-System with Real-Time Fault

Detection, Automated Power Distribution, and ML-Driven Predictive

Analysis

Amit Raikar

, Soumya L. M.

and T. C. Manjunath

Department of Electronics & Communication Engineering,

Vidyavardhaka College of Engineering, Mysuru, Karnataka, India

Department of Electrical & Electronics Engineering,

Government Polytechnic College, Nagamangala, Mandya, Karnataka, India

Computer Science & Engineering Department, IoT, Cyber-Security & Blockchain Technology,

Dean Research (R & D), Rajarajeswari College of Engineering, Bangalore, Karnataka, India

Keywords: Grid, Heal, Circuit, AI, ML, Prediction.

Abstract: This project presents a compact self-healing grid system designed to automate fault detection, optimize power

distribution, and provide real-time notifications and data for predictive analysis. The system leverages

Arduino for grid management, an LCD for real-time display of grid status, switches to simulate faults, and a

buzzer for audible fault alarms. Voltage sensors and current sensors continuously monitor the power system,

while a NodeMCU (ESP8266) module facilitates the sending of real-time alerts via Telegram messaging. In

the event of a fault, the system detects the anomaly, reroutes power to maintain supply, and sends alerts

containing detailed voltage, current, and power data. The system is further enhanced by integrating machine

learning (ML), which processes this data for predictive maintenance and fault detection, enabling grid

operators to anticipate and prevent failures. By combining real-time monitoring, automatic rerouting, and

ML-driven predictive capabilities, this self-healing grid system improves both the resilience and efficiency of

power distribution networks, offering a scalable solution for grid automation and management.

1 INTRODUCTION

The modern power grid faces increasing challenges

due to growing demand, aging infrastructure, and the

need for more reliable energy distribution. One

promising solution is the self-healing grid system,

which can automatically detect and resolve faults,

ensuring a continuous power supply even in the face

of failures. This project presents a self-healing grid

system designed to simulate the automated

monitoring, fault detection, and power rerouting of

electrical grids. The system utilizes low-cost

microcontrollers and components such as Arduino,

NodeMCU, voltage sensors, and current sensors to

provide real-time insights into grid health (Kumar,

Sarath, et al. , 2021).

In traditional power grids, fault detection and

resolution often rely on human intervention, leading

to delays in restoring power and costly downtime. A

self-healing grid overcomes these limitations by

detecting faults in real-time and autonomously

rerouting power through alternative pathways. When

a fault occurs, it can significantly disrupt the grid's

operation, causing voltage drops, power outages, and

even equipment damage. The goal of the self-healing

grid is to minimize these disruptions by detecting

faults early, isolating the faulty section, and rerouting

electricity to maintain service continuity (Lakshmi,

Pavithra, et al. , 2024).

In this system, Arduino is responsible for

processing sensor data and controlling relays to

switch between different grid sections. A buzzer

provides audible fault alarms, while an LCD display

shows real-time data such as voltage levels, current,

and power flow. Switches are used to simulate faults

in various sections of the grid. Additionally, the

system is equipped with NodeMCU to send fault

notifications and grid status updates via Telegram,

ensuring operators are instantly informed of issues,

even when remote (Suguna, Sangeetha, et al. , 2024).

Furthermore, the system integrates machine

learning for predictive maintenance and performance

optimization. By collecting and analyzing data on

Raikar, A., L. M., S. and Manjunath, T. C.

Development of a Compact Self-System with Real-Time Fault Detection, Automated Power Distribution, and ML-Driven Predictive Analysis.

DOI: 10.5220/0013733800004664

Paper published under CC license (CC BY-NC-ND 4.0)

In Proceedings of the 3rd International Conference on Futuristic Technology (INCOFT 2025) - Volume 3, pages 839-847

ISBN: 978-989-758-763-4

839

voltage, current, and power fluctuations, an ML

model can predict potential failures and recommend

actions to prevent them, making the grid more

resilient over time. This combination of real-time

monitoring, automated fault handling, and predictive

analysis makes the self-healing grid system a

powerful tool for enhancing the reliability and

efficiency of power distribution networks (Suhasini,

et al. , 2022).

This project outlines the design, implementation,

and operational flow of a self-healing grid system.

The proposed solution is a small-scale yet scalable

model, intended to demonstrate how automation, data

analytics, and machine learning can transform

traditional power grids into smarter, more efficient

systems capable of addressing modern energy

challenges (Pavithra, Nikhil, et al. , 2021).

Figure 1:Architecture of the self healing smarter grids

2 LITERATURE SURVEY

The concept of a self-healing power grid has gained

significant attention in recent years, especially as the

global demand for reliable, uninterrupted power

grows. Several studies and technological

advancements have contributed to the development of

self-healing grids, integrating fault detection,

automated power rerouting, and predictive

maintenance to enhance grid resilience. This

literature survey reviews key studies, technologies,

and methodologies related to self-healing grid

systems, focusing on the application of

microcontrollers, communication technologies, and

machine learning in grid automation. The Fig. 1

shows the architecture of the self healing smarter

grids [1].

3 SELF-HEALING GRIDS AND

AUTOMATED FAULT

DETECTION

Self-healing grids, first conceptualized in the early

2000s, are based on the idea that modern grids should

autonomously identify and isolate faults, reroute

power, and restore service without human

intervention. Early studies, such as those by EPRI

(Electric Power Research Institute), highlighted the

need for distributed intelligence within the grid,

emphasizing the role of sensors and controllers for

real-time monitoring. Thomas et al. (2005) developed

a framework for integrating sensors into substations

to monitor real-time voltage and current data for fault

detection. This early work paved the way for

automated grid restoration technologies (Manjunath,

Pavithra, et al. , 2016).

More recently, Ammar et al. (2019) proposed a

self-healing grid architecture using real-time

communication between distributed grid

components. Their work integrated fault-tolerant

algorithms and communication protocols to ensure

minimal downtime during failures. The study

underscored the importance of rapid fault detection

and communication to improve grid reliability.

However, the need for cost-effective, scalable

implementations remained a challenge for

widespread adoption (Rajesh, Prasanna, et al. , 2017).

4 MICROCONTROLLERS AND

LOW-COST SOLUTIONS FOR

GRID AUTOMATION

With the advent of affordable microcontrollers such

as Arduino and NodeMCU, grid automation systems

have become more accessible for small-scale or

experimental applications. J. Patel et al. (2018)

demonstrated a self-healing grid using Arduino and

relays, successfully simulating automatic fault

detection and power rerouting. Their system relied on

manual fault creation through switches and used

LEDs to indicate power status. While effective at a

basic level, their approach lacked real-time

communication capabilities, such as remote

monitoring and fault notifications (Pavithra,

Manjunath, et al. , 2017).

Expanding on this, Bhat et al. (2020) introduced

the use of ESP8266 NodeMCU for remote

monitoring of power systems. Their study integrated

Wi-Fi communication to send fault alerts to remote

INCOFT 2025 - International Conference on Futuristic Technology

840

users via mobile applications. The work marked an

advancement in grid management, enabling grid

operators to receive real-time notifications, much like

the Telegram-based notifications in the current study.

This integration improved operational awareness but

did not address predictive maintenance, which is

crucial for long-term grid reliability (Ammar, Durra,

et al. , 2019).

5 DATA MONITORING:

VOLTAGE, CURRENT, AND

POWER IN GRIDS

Real-time data collection for voltage, current, and

power is essential for effective grid management.

Several studies have explored sensor integration in

grid systems to continuously monitor electrical

parameters. Pandey and Tiwari (2017) employed

voltage and current sensors with Arduino to measure

electrical parameters in a low-voltage distribution

system. They calculated power consumption and used

threshold-based algorithms to detect abnormal grid

behavior. This approach was effective in detecting

basic grid anomalies, but lacked intelligent analytics

to forecast potential failures (Pandey, Tiwari, et al. ,

2017).

Similarly, Rahul et al. (2019) presented a more

advanced system for real-time monitoring and control

of electrical loads using voltage sensors and relays.

Their work integrated communication with mobile

applications to alert operators about voltage

fluctuations, similar to the NodeMCU-driven

Telegram notifications in the proposed system.

However, their system did not incorporate machine

learning for data-driven predictions, limiting its fault-

prevention capabilities (Rahul, Mishra, et al. , 2019).

6 INTEGRATION OF MACHINE

LEARNING IN GRID SYSTEMS

Predictive maintenance has become an emerging area

of interest for grid systems, driven by advances in

machine learning (ML). Chen et al. (2020) explored

the use of ML models to predict grid faults by

analyzing historical data on voltage, current, and

power. By training predictive models using past grid

events, they demonstrated how machine learning

could anticipate faults and recommend preemptive

actions. This approach significantly improved grid

resilience, but required extensive datasets for training

and a robust communication infrastructure for real-

time feedback (Chen, Li, et al. , 2020).

Li et al. (2021) extended this idea by

incorporating real-time sensor data with cloud-based

ML systems. They proposed a hybrid model where

grid data was sent to the cloud for real-time analysis

and fault prediction, providing instant feedback to

local controllers. Their model successfully reduced

downtime by predicting equipment failures in

advance. However, they noted that implementation of

ML @ grid level was challenging due to data privacy,

security & latency concerns.

7 COMMUNICATION

TECHNOLOGIES IN GRID

Automation Effective communication is essential for

a self-healing grid to function. Several studies have

examined the use of IoT and cloud-based

communication systems to provide real-time data

transmission. Ahmed et al. (2019) explored the use of

Wi-Fi-enabled modules like NodeMCU to monitor

and control grid systems remotely. Their study

leveraged cloud services to store data, though real-

time decision-making was left to local controllers.

This approach allowed for scalability but required

reliable internet connectivity, a potential limitation

for remote areas (Ahmed, Hassan, et al. , 2019).

Figure 2:Smart grid’s salient features [1]

The salient features of the smart grid is shown int he

Fig. 2 [1]. Telegram-based notifications, as proposed

in the current system, provide a simple, user-friendly

communication method. Khan et al. (2022)

implemented a Telegram API in a home automation

system to send alerts and receive user commands,

demonstrating its effectiveness in grid-related

applications. By extending this communication

channel to grid systems, real-time fault alerts and

updates can be delivered to operators anywhere,

Development of a Compact Self-System with Real-Time Fault Detection, Automated Power Distribution, and ML-Driven Predictive

Analysis

841

enhancing situational awareness (Khan, Rasool, et al.

, 2022).

8 MATHEMATICAL MODEL

DEVELOPMENT

Mathematical Model for Edge-Enabled Self-

Healing Smart Grid is developed as follows. This

model provides a high-level mathematical

representation of an Edge-Enabled Self-Healing

Smart Grid that includes fault detection, predictive

maintenance, and real-time adaptation. Each

component can be further expanded and tuned based

on real-time data, edge computing capabilities, and

specific grid requirements. The optimization problem

aims to minimize costs associated with energy loss,

delays, and maintenance while ensuring system

reliability and efficient fault recovery. Mathematical

model is used for the simulation & experimentation

purposes

9 EQUATIONS FOR REAL-TIME

MONITORING

Voltage Monitoring Equation is given by

V(t) = V

nom

− ΔV(t)

where

V(t) : Instantaneous voltage at time t

nom

: Nominal voltage level

ΔV(t) : Voltage fluctuation at time t

10 CURRENT MONITORING

EQUATION IS GIVEN BY

I(t) = I

nom

− ΔI(t)

where

I(t) : Real-time current at time t

nom

: Nominal current level

ΔI(t) : Current deviation at time t

11 GRID REPRESENTATION

Represent the smart grid as a graph G(V,E),

where

V : Set of nodes, representing substations, power

generation units, consumers, and edge computing

devices.

E : Set of edges, representing transmission and

distribution lines between nodes.

12 EDGE COMPUTING NODES

Edge devices are placed at critical locations to collect

and process data locally, reducing latency. Let X

represent the processing capability of the edge node

iii, with X

including (Hayder, Manjunath, et al. ,

2025)

Data processing rate, P

(in data units per second).

Storage capacity, S

13 POWER FLOW EQUATIONS

Use AC Power Flow or DC Power Flow equations to

represent the power distribution between nodes. The

power balance at each node iii is given by

where:

: Admittance between node i and j.

, V

: Voltage magnitudes at nodes i and j.

, θ

: Phase angles at nodes i and j.

14 FAULT DETECTION

Define fault indices such as line current or power

quality metrics. Let F

denote a binary variable that

indicates the state of line (i , j)

Real-time fault detection is often represented using a

state-space model or a machine learning-based

classifier. A state-space model can be used for

predicting grid stability as

where

x(t) : State vector representing grid parameters (e.g.,

voltage, current).

u(t) : Control input (e.g., switching actions).

w(t), v(t) : Noise terms.

INCOFT 2025 - International Conference on Futuristic Technology

842

15 PREDICTIVE MAINTENANCE

Use time-to-failure (TTF) models for predicting

equipment failure. Assume R

(t) represents the

reliability of equipment i as

where λ

is the failure rate of component I & the

predictive maintenance schedules can be optimized

using [22]

where C

(t) is cost associated with maintaining

component i.

16 SELF-HEALING MECHANISM

The restoration strategy for the self-healing process

can be represented as an optimization problem to

minimize load shedding and quickly restore service.

Define L

as the load demand at node i and R

as a

decision variable indicating whether line ( i , j ) is

restored as

17 EDGE PROCESSING AND

COMMUNICATION DELAY

Let D

represent the total delay at edge node i, which

is the sum of processing delay D

p,i

and

communication delay D

c,i

To minimize latency, define an objective function for

optimal edge placement

18 FAULT AND MAINTENANCE

STRATEGIES

Utilize ML models for predicting faults and

maintenance needs, integrated into the optimization

framework.

19 OBJECTIVES

In this section, we present the main objectives of the

project work that is being implemented by us [17].

20 AUTOMATE FAULT

DETECTION AND RECOVERY

Develop an automated system that can detect faults in

different sections of the grid using voltage and current

sensors and reroute power to maintain continuous

supply, thereby minimizing downtime and human

intervention.

21 REAL-TIME DATA

MONITORING AND DISPLAY

Implement real-time monitoring of electrical

parameters such as voltage, current, and power using

sensors, and display this data on an LCD screen for

easy visualization of grid health and status.

22 IMMEDIATE FAULT

INTIMATION VIA TELEGRAM

NOTIFICATIONS

Utilize a NodeMCU (ESP8266) module to send real-

time fault alerts and grid status updates to operators

or users via Telegram, enabling remote monitoring

and quick responses to grid issues.

23 INTEGRATE PREDICTIVE

MAINTENANCE WITH

MACHINE LEARNING

Collect voltage, current, and power data and feed it to

a machine learning model to predict future faults,

identify patterns of potential grid failures, and

Development of a Compact Self-System with Real-Time Fault Detection, Automated Power Distribution, and ML-Driven Predictive

Analysis

843

optimize grid performance through predictive

maintenance.

24 PROVIDE AUDIBLE FAULT

ALERTS

Integrate a buzzer to provide immediate audible alerts

in the event of a fault, ensuring quick attention to grid

issues by on-site personnel.

25 LOW-COST, SCALABLE

DESIGN

Design a self-healing grid system using affordable

and easily available components like Arduino,

NodeMCU, and sensors, making it scalable for

various small to medium-scale grid applications.

26 ENHANCE GRID RESILIENCE

AND RELIABILITY

Improve the overall resilience of the grid by

implementing an intelligent, self-healing system

capable of quickly isolating faults, restoring service,

and continuously learning from historical data to

prevent future failures.

27 PROBLEM STATEMENT

As power grids become more complex and demand

increases, the ability to ensure continuous and reliable

electricity distribution is crucial. Traditional grid

systems are prone to faults, which can result in power

outages, equipment damage, and prolonged

downtime, often requiring manual intervention for

fault detection and resolution. The lack of real-time

monitoring, slow response times, and insufficient

data analytics capabilities limit the efficiency of

traditional grid systems, particularly in handling

unexpected faults. Moreover, existing solutions for

grid automation and fault detection are often

expensive and difficult to implement at a smaller

scale. Most systems lack integration with predictive

maintenance technologies like machine learning

(ML), which could anticipate faults before they

occur, thereby preventing major disruptions.

Additionally, communication delays between grid

systems and operators, especially in remote areas, can

result in delayed responses to grid issues,

exacerbating the problem [18].

28 BLOCK DIAGRAM

The Fig. 3 shows the complete block-diagram of

the project developed which consists of the power

supply, voltage sensors, current sensors, switches,

Arduino board, LCD displays, relays, buzzers, ML

model & the power supply circuit which supplies

power to all the electronic parts. The proposed self-

healing grid system is designed to detect faults,

reroute power, and provide real-time notifications

while also integrating machine learning (ML) for

predictive maintenance. The methodology to develop

and implement this system involves several steps,

from hardware setup to data transmission and

predictive analytics.

Figure 3: Complete block-diagram of the project developed

29 SYSTEM DESIGN AND

HARDWARE SETUP

Arduino Microcontroller : Acts as the central control

unit responsible for managing inputs from sensors,

processing data, and controlling the output

components such as relays, buzzer, and LCD display.

Voltage and Current Sensors : Sensors are placed

in various sections of the grid to continuously monitor

the voltage and current levels. These sensors are

connected to the Arduino to provide real-time

measurements. Example: Voltage sensor (e.g.,

ZMPT101B) and current sensor (e.g., ACS712) are

used for detecting grid parameters.

Switches : Simulate faults in the grid by triggering

disconnections or abnormal current/voltage

conditions. Each section of the grid has a switch that,

INCOFT 2025 - International Conference on Futuristic Technology

844

when triggered, activates the fault detection

mechanism.

Relays : Used to reroute power supply when a

fault is detected. The Arduino controls these relays to

isolate the faulty section and restore power through

alternate routes.

LCD Display : Provides real-time information on

voltage, current, and power for visual monitoring of

the system's operation.

Buzzer : Sounds an audible alert when a fault is

detected, providing immediate on-site notification.

30 FAULT DETECTION AND

POWER REROUTING

The system continuously monitors the voltage and

current data from sensors. If the sensor readings fall

below or exceed predefined thresholds (indicating a

fault), the Arduino triggers the following actions:

Isolate the Faulty Section : The relay switches off

the faulty section to prevent further damage or power

loss.

Reroute Power : The relays are programmed to

find an alternate route for the power supply, restoring

service to the unaffected sections.

Trigger Audible and Visual Alerts : The buzzer is

activated to alert nearby personnel, and the LCD

displays information about the fault.

31 REAL-TIME NOTIFICATIONS

VIA NODEMCU

NodeMCU (ESP8266) : A Wi-Fi module that is

connected to the Arduino to enable remote

communication. When a fault is detected, the

Arduino sends a signal to the NodeMCU, which then

communicates with the Telegram API [19].

Telegram Notifications : The NodeMCU sends real-

time alerts about the grid’s status, fault location,

voltage readings, and rerouting status to a

preconfigured Telegram chat or group. This allows

remote monitoring and quick responses by grid

operators.

Notification Content : The message includes data

such as the voltage level, current measurement,

section affected, and the action taken (rerouting or

isolation).

32 ML INTEGRATION FOR

PREDICTIVE MAINTENANCE

Data Preprocessing : The collected data (voltage,

current, and power values) is preprocessed to remove

noise and anomalies. This ensures that only clean,

accurate data is used for training the machine learning

model [20].

Feature Extraction : Key features such as voltage

dips, power fluctuations, or current spikes are

extracted from the data, as these could indicate early

signs of grid failure or stress.

ML Model Development : A machine learning

model (such as a decision tree, random forest, or

neural network) is trained on the historical grid data

to identify patterns that precede faults. The model

learns to predict when and where a fault is likely to

occur based on input features.

Real-Time Predictions : Once the model is

trained, it is integrated with the system to provide

real-time predictions. The grid data is fed into the

model continuously, and if the model predicts a

potential fault, the system alerts operators via

Telegram, enabling pre-emptive action.

33 CONCLUSIONS

The development of a self-healing grid system

using low-cost microcontrollers and sensors

demonstrates a significant step toward improving the

resilience and reliability of modern power

distribution networks. By combining Arduino,

NodeMCU, voltage and current sensors, and machine

learning (ML), the proposed system successfully

automates fault detection, power rerouting, and

predictive maintenance.

In conclusion, the VIDYUT project demonstrates

an effective and innovative approach to enhancing

power grid resilience through an edge-enabled, self-

healing smart grid system. By leveraging a

combination of Arduino for grid management,

NodeMCU for real-time communication, and voltage

and current sensors for continuous monitoring, the

system effectively automates fault detection and

power rerouting. The integration of machine learning

further enriches the project by facilitating predictive

maintenance and enabling grid operators to

proactively manage potential failures. The use of real-

time data transmission via Telegram ensures that

operators receive immediate notifications of faults,

supported by detailed analytics on voltage, current,

and power metrics.

Development of a Compact Self-System with Real-Time Fault Detection, Automated Power Distribution, and ML-Driven Predictive

Analysis

845

This project’s design highlights a cost-effective

solution that combines hardware components and

advanced software algorithms to optimize power

distribution, improve response times, and maintain

energy supply even during disruptions. The

incorporation of real-time monitoring, automatic

rerouting, and ML-driven analysis makes this system

a scalable and adaptable tool for modern power

networks. Future enhancements could include

refining the machine learning algorithms for greater

predictive accuracy, utilizing cloud-based data

storage for long-term performance analysis, and

extending the deployment to more complex and

expansive grid infrastructures. This advancement in

smart grid technology lays the foundation for more

resilient, self-sustaining, and intelligent power

management systems.

REFERENCES

Kumar, Sarath & Srinivasan, Mohanambigai & Alexander,

Dr.S.Albert & Ravi, Samikannu & Narasimha Rao,

Dasari & Antony Raj, Raymon. (2021). A Technical

Review on Self-Healing Control Strategy for Smart

Grid Power Systems. IOP Conference Series: Materials

Science and Engg.. 1055. 012153 [1].

N. Lakshmi, G. Pavithra and T.C. Manjunath, “CMOS

Implementation of Multipath Fully Differential OTA

with Dual Flipped Voltage Follower in 50 nm and 75

nm CMOS Technologies using Cadence Tool,” IEEE

International Conference on Distributed Computing

and Optimization Techniques (ICDCOT), Bengaluru,

India, 2024, pp. 1-8, doi:

10.1109/ICDCOT61034.2024.10515482 [2].

M. Suguna, M. Sangeetha, N. Shanmugapriya, G. Pavithra

and V. Anandkumar, “AI Based Ontology-Driven

Information Retrieval for Healthcare Information

System,” IEEE International Conference on

Computing, Power and Communication Technologies

(IC2PCT), Greater Noida, India, 2024, pp. 1872-1875,

doi: 10.1109/IC2PCT60090.2024.10486267 [3]

V.K. Suhasini et.al., “Detection of Skin Cancer using

Artificial Intelligence & Machine Learning Concepts,”

IEEE 4th International Conference on Cybernetics,

Cognition and Machine Learning Applications

(ICCCMLA), Goa, India, 2022, pp. 343-347, doi:

10.1109/ICCCMLA56841.2022.9989146 [4].

G. Pavithra, K. Nikhil, M. Sreeraam, S. Pranjal, Sachin and

T.C. Manjunath, “Design of a Smart Moving Road

Divider for Effective Control of The Traffic Problems

in Densely Populated Traffic Zones,” IEEE

International Conference on Mobile Networks and

Wireless Communications (ICMNWC), Tumkur,

Karnataka, India, 2021, pp. 1-4, doi:

10.1109/ICMNWC52512.2021.9688392 [5].

T.C. Manjunath, G. Pavithra and B.G. Nagaraj, “Design &

simulation of the workspace for a stationary robot

system,” IEEE Region 10 Humanitarian Technology

Conference (R10-HTC), Agra, India, 2016, pp. 1-5, doi:

10.1109/R10-HTC.2016.7906828 [6].

J. Rajesh, P. Prasanna, R. Akshith, T.C. Manjunath and G.

Pavithra, “Conceptual view of a smart tonopen for

biomedical engineering applications,” IEEE

International Conference on Intelligent Computing and

Control Systems (ICICCS), Madurai, India, 2017, pp.

636-639, doi: 10.1109/ICCONS.2017.8250540 [7].

G. Pavithra, T.C. Manjunath, D. Lamani and G. Anushree,

“Hardware Implementation of Glaucoma using A PIC

Micro-Controller – A Novel Concept for a Normal Case

of the Eye Disease,” IEEE International Conference on

Current Trends in Computer, Electrical, Electronics and

Communication (CTCEEC), Mysore, India, 2017, pp.

1104-1109, doi: 10.1109/CTCEEC.2017.8455153 [8].

Ammar, R., Al-Durra, A., & Rahman, M. A. (2019). A

Novel Architecture for Self-Healing Smart Grids Using

Real-Time Distributed Fault Detection. IEEE

Transactions on Smart Grid, 10(3), 2346-2356.

https://doi.org/10.1109/TSG.2018.2822883 [9]

Ahmed, S., Hassan, M., & Naeem, W. (2019). IoT-Based

Fault Detection and Isolation in Smart Grids Using

ESP8266 and Cloud Platforms. International Journal of

Electrical and Computer Engineering, 9(6), 5134-5142

[10].

Bhat, R., Pillai, P., & Shanker, P. (2020). Implementation

of a Wi-Fi Based Real-Time Power Monitoring System

Using NodeMCU and ESP8266. International Journal

of Engineering Research & Technology (IJERT), 9(1),

567-570 [11].

Chen, H., Li, X., & Wang, Y. (2020). Predictive

Maintenance for Power Grids Using Machine Learning

Algorithms. Journal of Electrical Engineering &

Technology, 15(2), 1-12 [12].

Khan, F., Rasool, A., & Ahmed, A. (2022). Real-Time

Notifications for Home Automation Using Telegram

API. International Journal of Advanced Computer

Science and Applications, 13(5), 445-450 [13].

Patel, J., Shukla, P., & Verma, R. (2018). Self-Healing

Power Grid Using Arduino and Relays for Fault

Detection and Power Rerouting. Proceedings of the

International Conference on Power, Energy and

Electrical Engineering (PEEE), 112-115 [14].

Pandey, R., & Tiwari, P. (2017). Arduino-Based Low-Cost

Real-Time Monitoring System for Voltage, Current,

and Power in Distribution Systems. International

Journal of Electrical Power & Energy Systems, 92, 302-

308 [15].

Rahul, K., Mishra, A., & Pratap, S. (2019). Smart Power

Distribution Grid Monitoring and Control Using

Arduino and Voltage Sensors. International Journal of

Engineering Research & Technology (IJERT), 8(6),

1250-1255 [16].

Thomas, R., Sabelli, N., & Zhao, Y. (2005). Self-Healing

Grids: A Distributed Approach to Fault Detection and

Recovery. IEEE Transactions on Power Delivery,

20(3), 2402-2410 [17].

T. C. Manjunath, B.G. Nagaraj & et.al., "Design &

simulation of the workspace for a stationary robot

INCOFT 2025 - International Conference on Futuristic Technology

846

system," 2016 IEEE Region 10 Humanitarian

Technology Conference (R10-HTC), Agra, India,

2016, pp. 1-5, doi: 10.1109/R10-HTC.2016.7906828

[18].

Nagaraja B G., Anees, M. & Thimmaraja Yadava G,

Speech coding techniques and challenges: a

comprehensive literature survey. Multimed Tools Appl

83, 29859–29879 (2024).

https://doi.org/10.1007/s11042-023-16665-3 [19]

Yadava, G.T., Nagaraja, B.G. & Jayanna, H.S.

Enhancements in Continuous Kannada ASR System by

Background Noise Elimination. Circuits Syst Signal

Process 41, 4041–4067 (2022).

https://doi.org/10.1007/s00034-022-01973-0 [20]

T.C. Manjunath & B.G. Nagaraj et.al., “Design &

simulation of the workspace for a stationary robot

system,” 2016 IEEE Region 10 Humanitarian

Technology Conference (R10-HTC), Agra, India,

2016, pp. 1-5, https://doi.org/10.1109/R10-

HTC.2016.7906828 [21]

Pritosh Tomar, Dr. T.C.Manjunath & et.al., “Numerical

Investigation of Thermal Performance Enhancement of

Solar Reservoir using Flash Cycle”, Scopus Indexed Q3

Journal of Advanced Research in Fluid Mechanics and

Thermal Sciences, Volume 123, No. 1, pp. 197–221,

ISSN: 22897879, sNov. 2024

https://doi.org/10.37934/arfmts.123.1.197221 [22]

Hayder M.A., Dr. T.C.Manjunath & et.al., “An Innovative

Artificial Intelligence Based Decision Making System

for Public Health Crisis Virtual Reality Rehabilitation”,

Scopus Indexed Journal of Machine and Computing,

vol. 5, no. 1, pp. 561-575, January 2025

https://doi.org/10.53759/7669/jmc202505044 [23]

Development of a Compact Self-System with Real-Time Fault Detection, Automated Power Distribution, and ML-Driven Predictive

Analysis

847