Own Studt Project

A Self Study Project Report On
EMBEDDED BASED SMART ACCIDENT PRE-ALERT AND PREVENTION

SYSTEM WITH MACHINE LEARNING
Submitted to
University College Of Engineering& Technology
In partial fulfillment of the requirements for the award of

the degree of
BACHELOR OF TECHNOLOGY
IN
ELECTRONICS AND COMMUNICATIONS ENGINEERING
By
GANJI ANIL KUMAR
(HT.NO:4511-21-735-047)
Under the guidance of

Mr. K. Avinash Babu
Asst. professor
DEPARTMENT OF ELECTRONICS AND COMMUNICATIONS ENGINEERING
DEPARTMENT OF ELECTRONICS AND COMMUNICATION ENGINEERING
UNIVERSITY COLLEGE OF ENGINEERING & TECHNOLOGY
MAHATMA GANDHI UNIVERSITY, NALGONDA

NALGONDA – 508001
2024
UNIVERSITY COLLEGE OF ENGINEERING & TECHNOLOGY
Nalgonda – 508001, Telangana
DEPARTMENT OF ELECTRONICS AND COMMUNICATION ENGINEERING
CERTIFICATE
Certified that this is a Bonafied record of the self study project work entitled,
“Embedded based smart accident pre-alert and prevention system with machine learning”, done
by G. Anil Kumar (4511-21-735-047), submitted to the faculty of Electronics and
Communications Engineering, in partial fulfillment of the requirements for the Degree of
Bachelor of Technology in Electronics and Communications Engineering from the Mahatma
Gandhi University, Nalgonda during the year 2024-2025.
Project Guide: Head of the Department:

Mr. K. Avinash Babu Mr. K. Avinas Babu
Assistant Professor(c), Assistant Professor(c),
Dept. of Electronics and Dept. of Electronics and
Communication Engineering, Communication Engineering,
University College of Engineering and University College of Engineering and
Technology, Nalgonda. Technology, Nalgonda.
___________________________________________________________________________
Viva-voice held on…………………………………………………………………..
ACKNOWLEDGEMENT
The satisfaction that accompanies the successful completion of the task would be put incomplete
without the mention of the people who made it possible, whose constant guidance and
encouragement crown all the efforts with success.
We wish to express our deep sense of gratitude to Mr. K. Avinash Babu Assistant Professor(c)
& Project Guide, Department of Electronics & Communication Engineering, University College
of Engineering and Technology, for her able guidance and useful suggestions, which helped me
in completing the project work in time.
We would like to thank Mrs. Huma Afreen project coordinator, Department of Electronics
&Communication Engineering, University College of Engineering &Technology, for his expert
guidance and encouragement at various levels of my project.
We are particularly thankful to Mr. K. Avinash Babu, Head of Department of Electronics and
Communication Engineering for her guidance, intense support and encouragement, which helped
me to mould my project into a successful one.
We would like to thank Principal Dr. Sudha Rani for her expert guidance and encouragement at
various levels of my Project.
We show our gratitude to our honourable Registrar Prof. A. Ravi, for having provided all the
facilities and support.
We avail this opportunity to express our deep sense of gratitude to our honourable Vice.
Chancellor Prof. Qaza Altaf Husssain, congenial atmosphere to complete this project
successfully.
We also thank all the staff members of Electronics & Communication Engineering department
for their valuable support and generous advice. Finally, thanks to all our friends and family
members for their continuous support and enthusiastic help.
G. Anil Kumar (4511-21-735-047)
i
DECLARATION
We here by declare that this self study project report titled “Embedded based smart accident pre-
alert and prevention system with machine learning” submitted in Partial Fulfillment of the
requirements for the award of degree of Bachelor of Technology in Electronics &
Communications Engineering. This dissertation is our original work and the project has not
formed the basis for the award of any degree, associate ship, fellowship or any other similar
titles and no part of it has been published or sent for the publication at the time of submission.
G. Anil Kumar
(4511-21-735-047)
ii
ABSTRACT
The rapid advancement of smart technologies has opened new avenues for enhancing road safety
and reducing accidents. This project proposes an embedded-based smart accident pre-alert and
prevention system integrated with machine learning (ML) techniques. The system leverages
sensor data from various on-vehicle devices (e.g., accelerometers, gyroscopes, and cameras) and
environmental factors (e.g., weather conditions, traffic, and road quality) to predict potential
accidents in real-time. The embedded system uses ML models trained on historical accident data
to analyze driver behavior, vehicle dynamics, and surrounding conditions. By continuously
monitoring parameters like speed, braking patterns, and proximity to other vehicles, the system
provides pre-emptive warnings to drivers, helping them avoid critical situations. Additionally, the
system can communicate with nearby vehicles and infrastructure to share vital information,
improving overall traffic safety. The proposed system aims to reduce accident rates by providing
timely alerts and taking preventive measures such as automatic braking or lane correction when a
potential collision is detected. The integration of machine learning enhances the accuracy and
adaptability of the system by learning from new data, enabling it to continuously improve its
predictive capabilities. This research will highlight the system architecture, key algorithms, and
the effectiveness of the proposed solution through simulation and real-world testing.
iii
TABLE OF CONTENTS
Acknowledgement i
Declaration ii
Abstract iii
CHAPTER NAME PAGE NAME
1. Introduction to Embedded system

1.1 Embedded system 1
1.2 Characteristics of an Embedded system 1
1.3 Advantages 2
1.4 Disadvantages 3
1.5 Basic structure of an Embedded system 3
1.6 Difference between General purpose 4
computer and Embedded system
1.7 Architecture of Embedded system 5
1.7.1 Von-Neumann Architecture 5
1.7.2 Harward Architecture 5
1.7.3 Von-Neumann Architecture vs Harward 6
Architecture
1.8 Types of Embedded system 6
1.8.1 Stand alone Embedded system 7
1.8.2 Real time Embedded system 7
1.8.3 Networked Embedded system 7
1.8.4 Mobile Embedded system 8
1.8.5 Small scale Embedded system 8
1.8.6 Medium scale Embedded system 8
1.8.7 Sophisticated Embedded system 8
1.9 Applications of Embedded system 9
2. Machine Learning
iv
2.1 Machine Learning 10
2.2 Importance of Machine Learning 10
2.3 Machine Learning process 10
2.3.1 Data Collection 11
2.3.2 Data Preparation 11
2.3.3 Choosing Learning Algorithm 11
2.3.4 Training Model 11
2.4 Types of Machine Learning 12
2.4.1 Supervised Machine Learning 12
2.4.2 Unsupervised Machine Learning 13
2.4.3 Reinforcement Machine Learning 13
3. Project Description
3.1 Existing System 14
3.2 Proposed System 14
3.3 Working 14
3.4 Module Description 15
3.4.1 Data Collection and Data Processing 15
3.4.2 Data Cleaning 16
3.4.3 Data Visualization 16
4. Methodology Flow
4.1 System Architecture 17
4.2 Input Images 17
4.2.1 Accident Dataset Creation 18
4.3 Working 18
4.4 Disadvantages of Existing System 19
4.5 Advantages of the Proposed system 20
4.6 Applications 20
v
5. Tools and Techniques
5.1 Python 22
5.2 Pycharm 22
5.3 Software Code or Deployment 23
5.3.1 Embedded system Code (Arduino 23
Example)
5.3.2 Machine Learning Part (Python) 25
5.3.3 Machine Learning Model Training 27
6. Final Results
6.1 Final Results 29
7. Chapter
7.1 Conclusion 31
7.2 future Work 31
7.3 Scope of Future Study 32
8. References 33
vi
LIST OF THE TABLES
NAME OF THE TABLES PAGE NUMBERS
1.6 Difference between General Purpose 4

Computer and Embedded systems
1.7.3 Von-Neumann Architecture vs Harvard 6
Architecture
LIST OF FIGURES
NAME OF THE FIGURES PAGE NUMBERS
1.2 System-on-chip Architecture Components 2
1.5 Structure of an Embedded System 3
1.6.1 Embedded computer system block diagram 4
1.7 Von-Neumann and Harvard Architectures 6
1.8 Types of an Embedded system 7
1.9 Applications of an Embedded system 9
2.3 Machine learning process 10
2.4 Types of machine learning 12
4.1 System Architecture 17
6.1 Accident occurred 29
6.2 Accident will happen 29
6.3 No. of Accidents occurred 30
vii
Accident Prevention System
1. INTRODUCTION TO EMBEDDED SYSTEMS
1.1 Embedded System
An embedded system can be thought of as a computer hardware system having

software embedded in it. It can be an independent system or it can be a part of a large
system. An embedded system is a microcontroller or microprocessor based system
which is designed to perform a specific task. For example, a fire alarm is an embedded
system; it will sense only smoke.
An embedded system has

components −  It has hardware.
• It has application software.
• It has Real Time Operating System(RTOS) though a small scale embedded
system may not have RTOS.
• So we can define an embedded system as a Microcontroller based, software
driven, reliable, real-time control system.
1.2 Characteristics of an Embedded System:
• Single-functioned − An embedded system usually repeats a specialized

operation. For example: A pager always functions as a pager.
• Tightly constrained − All computing systems have constraints on design

metrics like cost, size, power, and performance but those on an embedded
system can be especially tight. Design metrics is a measure of an
implementation's features. It must be of a size to fit on a single chip, must
perform fast enough to process data in real time and consume minimum power
to extend battery life.
• Reactive and Real time − Many embedded systems must continually react to
changes in the system's environment and must compute certain results in real
time without any delay. For example, a car cruise controller that continually
monitors and reacts to speed and brake sensors. It must compute acceleration or
Dept of ECE, UCET, MGU, Nalgonda 1

decelerations repeatedly within a limited time; a delayed computation can result

in failure to control of the car.
• Microprocessors based − It must be microprocessor or microcontroller based.
• Memory − It must have a memory, as its software usually embeds in ROM.
• Connected − It must have connected peripherals to connect input and output

devices.
• HW-SW systems − Software is used for more features and flexibility. Hardware
is used for performance and security.
1.2 Figure: System-on-chip Architecture Components

1.3 Advantages
• Small size
• Low power consumption
• Low cost
• Enhanced performance
• Easily customizable

1.4 Disadvantages
• High development effort
• Larger time to market
1.5 Basic Structure of an Embedded System
The following illustration shows the basic structure of an embedded system −
1.5 Figure: Structure of an Embedded System

• Sensor − It measures the physical quantity and converts it to an electrical signal
which can be read by an observer or by any electronic instrument like an A-D
converter. A sensor stores the measured quantity to the memory.
• A-D Converter − An analog-to-digital converter converts the analog signal sent

by the sensor into a digital signal.
• Processor & ASICs − Processors process the data to measure the output and
store it to the memory.
• D-A Converter − A digital-to-analog converter converts the digital data fed by

the processor to analog data
• Actuator − An actuator compares the output given by the D-A Converter to the
actual (expected) output stored in it and stores the approved output.

1.6 Difference between General purpose computer and Embedded systems:
General Purpose Computer Embedded Systems
It is designed using a microprocessor as the It is mostly designed using a microcontroller as

main processing unit. the main processing unit.
It contains a large memory semiconductor It uses semiconductor memories but does not
memories like cache and RAM. It also require secondary memories like hard disk, CD.
contains secondary storage like hard disks It sometime has special memory called flash
etc. memory.
It is designed such that it can do multiple It is designed such that it can do a particular
tasks as per requirement. predefined task.
It is mostly costlier compared to the

It is cheaper compared to a computer.
embedded systems
It is cheaper as it requires less no of peripheral

It requires huge number of peripheral devices
devices and their controllers are microcontroller
and their controllers
chip itself.
The Operating system and other software for The operating system (mostly RTOS i.e Real
the general purpose computers, are normally Time Operating System) and other software
complicated and occupy more memory space. occupy less memory space.

The Embedded system hardware includes elements like user interface, input/output interfaces,
display and memory, etc. Generally. An embedded system comprises power supply, processor,
memory, timers, serial communications ports and system applications specific circuits.
1.7 Architecture of Embedded Systems: The 8051 microcontrollers work with 8-bit data bus.
So they can support external data memory up to 64K and external program memory of 64k. They
can address 128k of external memory. When data and code lie in different memory blocks, then
the architecture is referred as Harvard architecture. In case data and code lie in the same
memory block, then the architecture is referred as Von Neumann architecture.
1.7.1 Von Neumann Architecture
The Von Neumann architecture was first proposed by a computer scientist John von Neumann.
In this architecture, one data path or bus exists for both instruction and data. As a result, the
CPU does one operation at a time. It either fetches an instruction from memory, or performs
read/write operation on data. So an instruction fetch and a data operation cannot occur
simultaneously, sharing a common bus. Von-Neumann architecture supports simple hardware. It
allows the use of a single, sequential memory. There are drawbacks to the Von Neumann design
especially when it comes to security, which was not even conceived as a problem until the
1980s. Program modifications can be quite harmful, either by accident or design. Since the
processor just executes the word the PC points to, there is effectively no distinction between
instructions and data. This is precisely the design flaw that attackers use to perform code
injection attacks and it leads to the theme of the inherently secure processor: the processor
cooperates in security.
1.7.2 Harvard Architecture

The Harvard architecture offers separate storage and signal buses for instructions and data. This
architecture has data storage entirely contained within the CPU, and there is no access to the
instruction storage as data. Computers have separate memory areas for program instructions and
data using internal data buses, allowing simultaneous access to both instructions and data.
Programs needed to be loaded by an operator; the processor could not boot itself. In a Harvard
architecture, there is no need to make the two memories share properties.

1.7 figure: Von-Neumann and Harvard Architectures

1.7.3 Von-Neumann Architecture vs Harvard Architecture
The following points distinguish the Von Neumann Architecture from the Harvard
Architecture.
Von-Neumann Architecture Harvard Architecture
Single memory to be shared by both code and Separate memories for code and data.
data.
Processor needs to fetch code in a separate Single clock cycle is sufficient, as separate
clock cycle and data in another clock cycle. So buses are used to access code and data.
it requires two clock cycles.
Higher speed, thus less time consuming. Slower in speed, thus more time-consuming.
Simple in design. Complex in design.
1.8 Types of Embedded systems
Embedded systems are specialized computing systems designed to perform dedicated tasks
within larger systems. Embedded systems can be classified into two different types based on
performance, functional requirements and performance of the microcontroller.

1.8 Figure: Types of an Embedded Systems

Embedded systems are classified into four categories based on their
performance and functional requirements: Stand alone embedded systems, Real time embedded
systems, Networked embedded systems and Mobile embedded systems.
Embedded Systems are classified into three types based on the performance of the
microcontroller such as: Small scale embedded systems, Medium scale embedded systems,
Sophisticated embedded systems
1.8.1 Stand Alone Embedded Systems

Stand alone embedded systems do not require a host system like a computer, it works by itself. It
takes the input from the input ports either analog or digital and processes, calculates and converts
the data and gives the resulting data through the connected device-Which either controls, drives
and displays the connected devices. Examples for the stand alone embedded systems are mp3
players, digital cameras, video game consoles, microwave ovens and temperature measurement
systems.
1.8.2 Real Time Embedded Systems

A real time embedded system is defined as, a system which gives a required o/p in a particular
time. These types of embedded systems follow the time deadlines for completion of a task. Real
time embedded systems are classified into two types such as soft and hard real time systems.
1.8.3 Networked Embedded Systems

These types of embedded systems are related to a network to access the resources. The connected
network can be LAN, WAN or the internet. The connection can be any wired or wireless. This
type of embedded system is the fastest growing area in embedded system applications. The

embedded web server is a type of system wherein all embedded devices are connected to a web
server and accessed and controlled by a web browser. Example for the LAN networked
embedded system is a home security system wherein all sensors are connected and run on the
protocol TCP/IP
1.8.4 Mobile Embedded Systems

Mobile embedded systems are used in portable embedded devices like cell phones, mobiles,
digital cameras, mp3 players and personal digital assistants, etc. The basic limitation of these
devices is the other resources and limitation of memory.
1.8.5 Small Scale Embedded Systems

These types of embedded systems are designed with a single 8 or 16-bit microcontroller, that
may even be activated by a battery. For developing embedded software for small scale embedded
systems, the main programming tools are an editor, assembler, cross assembler and integrated
development environment (IDE).
1.8.6 Medium Scale Embedded Systems

These types of embedded systems design with a single or “16 or 32” bit microcontroller, RISCs
or DSPs. These types of embedded systems have both hardware and software complexities. For
developing embedded software for medium scale embedded systems, the main programming
tools are C, C++, JAVA, Visual C++, RTOS, debugger, source code engineering tool, simulator
and IDE.
1.8.7 Sophisticated Embedded Systems

These types of embedded system have enormous hardware and software complexities, that may
need ASIPs, Ips, PLAs, scalable or configurable processors. They are used for cutting-edge
applications that need hardware and software Co-design and components which have to assemble
in the final system.
1.9 Applications of Embedded Systems:
Embedded systems have widespread applications across various industries due to their ability to
perform specific tasks with high efficiency and reliability. Embedded systems are used in

different applications like automobiles, telecommunications, smart cards, missiles, satellites,

computer networking and digital consumer electronics.
1.9 Figure: Applications of Embedded System:

2.Machine Learning
2.1 Machine Learning
Machine Learning (ML) is coming into its own, with a growing recognition that ML can
play a key role in a wide range of critical applications, such as data mining, natural
language processing, image recognition, and expert systems.
" A computer program is said to learn from experience E with respect to some task T and some
performance measure P, if its performance on T, as measured by P, improves with experience E.”
It is widely used in applications such as recommendation systems, speech recognition, image
processing, and autonomous driving.
2.2 Importance of Machine Learning
Machine learning is important because it gives enterprises a view of trend in customer

behavioral and business operational patterns, as well as supports the development of new
products. Many of today's leading companies, such as Facebook, Google and Uber, make
machine learning a central part of their operations.
2.3 Machine Learning Process

Machine learning workflow refers to the series of stages or steps involved in the process of
building a successful machine learning system. The various stages involved in the machine
learning workflow are-

Figure 2.3 Machine Learning Process
Figure 2.3 Machine Learning Process

2.3.1 Data Collection- In this stage,
Data is collected from different sources. The type of data collected depends upon the type of
desired project. Data may be collected from various sources such as files, databases etc. The
quality and quantity of gathered data directly affects the accuracy of the desired system. The
quality and quantity of data are crucial. Data is gathered from various sources, which can
includes data bases, sensors, or web scraping, depending on the problem.
2.3.2 Data Preparation- In this stage,

Data preparation is a crucial step in the machine learning process, as the quality and structure of
data directly impact the performance of the model. Data preparation is done to clean the raw
data. Data collected from the real world is transformed to a clean dataset. Raw data may contain
missing values, inconsistent values, duplicate instances etc. So, raw data cannot be directly used
for building a model.
2.3.3 Choosing Learning Algorithm- In this stage,
Choosing a learning algorithm in the machine learning process involves several key
considerations that can significantly impact the performance and outcome of the model. The best
performing learning algorithm is researched. It depends upon the type of problem that needs to
solved and the type of data we have. If the problem is to classify and the data is labeled,
classification algorithms are used.

If the problem is to perform a regression task and the data is labeled, regression algorithms are
used. If the problem is to create clusters and the data is unlabeled, clustering algorithms are used.
2.3.4 Training Model- In this stage,
Training a model in the machine learning process involves several key steps designed to enable
the model to learn patterns from data and make predictions.
The model is trained to improve its ability. The dataset is divided into training dataset and testing
dataset. The training and testing split is order of 80/20 or 70/30. It also depends upon the size of
the dataset. Training dataset is used for training purpose. Testing dataset is used for the testing
purpose.
2.4 Types of Machine Learning
There are 3 types of Machine Learning

1. Supervised Machine Learning
2. Unsupervised Machine Learning
3. Reinforcement Machine Learning
Figure 2.4 Types of Machine Learning
2.4.1 Supervised Machine Learning:
Supervised machine learning is based on supervision. It means in the supervised learning

technique, we train the machines using the "labelled" dataset, and based on the training, the
machine predicts the output. Here, the labelled data specifies that some of the inputs are
already mapped to the output. More preciously, we can say; first, we train the machine with

the input and corresponding output, and then we ask the machine to predict the output using
the test dataset.
Supervised machine learning can be classified into two types of problems, which are given
below:
a) Classification: Classification algorithms are used to solve the classification problems in
which the output variable is categorical, such as "Yes" or “No”,
Male or Female, Red or Blue, etc.
b) Regression: Regression algorithms are used to solve regression problems in which there
is a linear relationship between input and output variables. These are used to predict
continuous output variables, such as market trends, weather prediction, etc.
2.4.2 Unsupervised Machine Learning:
In unsupervised learning, the models are trained with the data that is neither classified nor
labelled, and the model acts on that data without any supervision. The main aim of the
unsupervised learning algorithm is to group or categories the unsorted dataset according to
the similarities, patterns, and differences.
Unsupervised Learning can be further classified into two types, which are given below:
a) Clustering: The clustering technique is used when we want to find the inherent
groups from the data. It is a way to group the objects into a cluster such that the
objects with the most similarities remain in one group and have fewer or no
similarities with the objects of other groups.
b) Association: Association rule learning is an unsupervised learning technique, which
finds interesting relations among variables within a large dataset. The main aim of
this learning algorithm is to find the dependency of one data item on another data
item and map those variables accordingly so that it can generate maximum profit.
2.4.3 Reinforcement Machine Learning:

Reinforcement learning works on a feedback-based process, in which an AI agent (A software

component) automatically explore its surrounding by hitting & trail, taking action, learning
from experiences, and improving its performance. Reinforcement learning is categorized
mainly into two types of methods/algorithms:
a) Positive Reinforcement Learning: Positive reinforcement learning specifies

increasing the tendency that the required behavioral would occur again by adding
something. It enhances the strength of the behavioral of the agent and positively
impacts it.
b) Negative Reinforcement Learning: Negative reinforcement learning works exactly
opposite to the positive RL. It increases the tendency that the specific behavioral
would occur again by avoiding the negative condition.

3.PROJECT DESCRIPTION
3.1 EXISTING SYSTEM:
The existing system is designed to enhance road safety by leveraging advanced technologies.
This system integrates sensors, such as accelerometers and cameras, with machine learning
algorithms to monitor real-time vehicle dynamics and environmental conditions. By analyzing
data patterns, it can detect potential accident scenarios, such as sudden braking or erratic driving
behaviors. When a risk is identified, the system sends alerts to the driver through auditory or
visual signals, providing timely warnings to prevent collisions. Additionally, the system can
communicate with other vehicles and traffic infrastructure to improve situational awareness,
ultimately aiming to reduce the incidence of road accidents and enhance overall safety on the
roads.
3.2 PROPOSED SYSTEM:
The proposed system, aims to enhance road safety by leveraging embedded systems and
machine learning algorithms. The system is designed to detect potential accident scenarios in
real-time by analyzing sensor data from vehicles and the environment, such as speed, distance,
and weather conditions. Embedded systems collect and process this data locally, while machine
learning models predict accident risks based on historical and real-time data patterns. If a high-
risk situation is identified, the system can trigger pre-alerts for drivers to take preventive actions
or automatically adjust vehicle behavior, such as reducing speed or applying brakes. This
intelligent system aims to reduce accidents, enhance driver awareness, and improve overall
traffic safety by predicting and preventing hazardous situations before they occur.
3.3 WORKING
An Embedded-based Smart Accident Pre-Alert and Prevention System with Machine Learning
integrates embedded systems with sensors, communication modules, and machine learning
algorithms to improve road safety. The system collects real-time data from vehicle sensors, such
as speed, acceleration, proximity, and driver behavior (e.g., fatigue detection through cameras or
wearables). Machine learning models process this data to detect abnormal patterns or risky
driving behaviors. If the system predicts a potential accident, it sends pre-alerts to the driver and,
in critical situations, can take preventive actions such as automatically slowing down the vehicle
or activating emergency brakes. Additionally, the system can notify emergency services in case

of an accident, providing vital data like location and accident details. The embedded system
ensures real-time processing, making it an effective tool for both accident prevention and
immediate response.
3.4 MODULE DESCRIPTION
1. Data collection and Data Preprocessing
2. Data cleaning
3. Data Visualization
3.4.1 Data collection and Data Preprocessing:
Collecting data for training the ML model is the basic step in the machine learning pipeline. The
predictions made by ML systems can only be as good as the data on which they have been
trained. Following are some of the problems that can arise in data collection:
Inaccurate data. The collected data could be unrelated to the problem statement. Missing data.
Sub-data could be missing. That could take the form of empty values in columns or missing
images for some class of prediction.
Data imbalance. Some classes or categories in the data may have a disproportionately high or
low number of corresponding samples. As a result, they risk being under-represented in the
model.
Data bias. Depending on how the data, subjects and labels themselves are chosen, the model
could propagate inherent biases on gender, politics, age or region, for example. Data bias is
difficult to detect and remove.
Data Preprocessing: Real-world raw data and images are often incomplete, inconsistent and
lacking in certain behaviors or trends. They are also likely to contain many errors. So, once
collected, they are pre-processed into a format the machine learning algorithm can use for the
model.
Pre-processing includes a number of techniques and actions:
Data cleaning. These techniques, manual and automated, remove data incorrectly added or
classified.
Data imputations. Most ML frameworks include methods and APIs for balancing or filling in
missing data. Techniques generally include imputing missing values with standard deviation,
mean, median and k-nearest neighbors (k-NN) of the data in the given field.

3.4.2 Data cleaning:

Data cleaning is one of the important parts of machine learning. It plays a significant part in
building a model. It surely isn’t the fanciest part of machine learning and at the same time, there
aren’t any hidden tricks or secrets to uncover. However, the success or failure of a project relies
on proper data cleaning. Professional data scientists usually invest a very large portion of their
time in this step because of the belief that “Better data beats fancier algorithms”.
3.4.3 Data visualization:
Data visualization is the graphical representation of information and data in a pictorial or

graphical format (Example: charts, graphs, and maps). Data visualization tools provide an
accessible way to see and understand trends, patterns in data, and outliers. Data visualization
tools and technologies are essential to analyzing massive amounts of information and making
data-driven decisions. The concept of using pictures is to understand data that has been used for
centuries. General types of data visualization are Charts, Tables, Graphs, Maps.

4.METHODOLOGY FLOW
This chapter describes the overall design. It also describes each module that
is to be implemented.
4.1 SYSTEM ARCHITECTURE
Figure: 4.1: System Architecture

4.2 Input Images
Input images form the cornerstone of any computer vision-based system, especially in
accident detection and prevention systems. In the proposed architecture, the input images are
primarily sourced from CCTV cameras installed at strategic locations such as highways,
intersections, and accident-prone areas. These cameras capture real-time footage of the traffic
and surrounding environment, which is then processed by the system to detect accidents or
predict potential collisions.
The quality and consistency of input images are crucial for accurate detection. Poor-quality
images due to environmental conditions like rain, fog, or low light can pose significant
challenges for image processing algorithms. Hence, preprocessing steps are applied to these
images to improve their quality and extract meaningful features. The images captured often
include various scenarios, such as moving vehicles, pedestrians, and road conditions, which
are critical in detecting accident-prone situations.
Additionally, the CCTV footage may contain other distractions such as irrelevant objects or
occlusions, which need to be filtered out during preprocessing. These images serve as the

foundational data that the machine learning model will analyze to predict accidents, making
them an indispensable element in the entire system architecture. The use of high-quality
images ensures that the system can detect even subtle indicators of potential accidents, such
as sudden stops, erratic vehicle movements, or traffic violations.
4.2.1 Accident Dataset Creation
The creation of an accident dataset is a vital step in training machine learning models to
detect accidents effectively. This dataset consists of a large collection of labeled images or
videos that depict various accident scenarios. The dataset should cover a wide range of
accident types, including rear-end collisions, side-impact crashes, pedestrian accidents, and
more. It should also incorporate different environmental conditions (e.g., daylight, nighttime,
rain, fog) to ensure the robustness of the model.
Accident datasets are typically created by gathering historical accident footage from CCTV
cameras or other sources. In cases where real accident data is scarce or unavailable, synthetic
datasets can be generated using simulation tools. These synthetic datasets simulate different
types of accidents in controlled environments, allowing for the generation of a diverse range
of accident scenarios. Once the raw data is collected, it must be manually or automatically
labeled with the relevant accident information. Each image or video frame is tagged with
labels that specify the type of accident, the objects involved (e.g., cars, pedestrians), and the
severity of the incident.
The process of creating an accident dataset also involves the removal of irrelevant data, such
as images without any accidents or frames that do not provide useful information. The final
dataset should be balanced, ensuring that it includes both accident and non-accident
scenarios in equal proportions. This balanced dataset is essential for training the machine
learning model to distinguish between normal and accident-prone situations.
4.3 Working
The working of the proposed accident pre-alert and prevention system follows a streamlined
flow, starting from data collection to the detection of accidents and real-time alerts. The
system's core components are embedded hardware and machine learning models, working in
tandem to process real-time data and predict accidents.

1. Data Collection: The system continuously gathers data from CCTV cameras installed
in various locations. These cameras capture real-time footage, which is then passed to
the preprocessing module.
2. Preprocessing: Before feeding the data into the machine learning model, it undergoes
preprocessing. This step involves cleaning the data, enhancing image quality, and
filtering out irrelevant information. Preprocessing ensures that only relevant and high-
quality data is used for accident detection.
3. Model Training: The system uses a machine learning model trained on a labeled
accident dataset. During the training phase, the model learns to recognize patterns
associated with accidents, such as sudden deceleration, erratic driving behavior, or
collisions. The model is fine-tuned to minimize false positives and ensure accurate
predictions.
4. Accident Detection: Once the model is trained, it is deployed on embedded hardware
that processes real-time footage from the CCTV cameras. The model continuously
monitors the incoming video data and uses its learned patterns to detect potential
accidents.
5. Alert Generation: Upon detecting an accident, the system immediately generates an
alert. This alert can be sent to the relevant authorities, such as traffic management or
emergency services, allowing for a swift response. The detected accidents are also
displayed on a monitoring screen for human operators to review.
The system's working is highly efficient due to the integration of machine learning models,
which improve over time as more data is fed into them. The system is also scalable, meaning
that more cameras and locations can be added to the network, and the system will adapt to
process the increased data flow.
4.4 Disadvantages of Existing Systems
Existing accident detection systems have several limitations, primarily due to their reliance
on traditional image processing techniques or manual observation. These systems often
struggle with the following disadvantages:
1. Limited Accuracy: Traditional systems may not be able to detect accidents with high
accuracy, especially under challenging conditions like low visibility, heavy traffic, or

unusual accident scenarios. This leads to missed detections or a high rate of false
positives.
2. High Latency: Many existing systems have significant delays between the occurrence
of an accident and the generation of an alert. This delay can result in slower response
times from emergency services, which can be critical in accident scenarios.
3. High False Positive Rate: Without the use of advanced machine learning algorithms,
many existing systems produce a high number of false positives, where normal traffic
incidents are incorrectly flagged as accidents. This leads to unnecessary interventions
and alerts.
4. Inflexibility: Traditional systems are not easily scalable or adaptable. Adding more
cameras or new locations often requires significant manual adjustments, which can be
time-consuming and expensive.
5. Manual Monitoring: Many existing systems rely on human operators to manually
monitor video feeds for accidents. This is not only inefficient but also prone to human
error and fatigue.
4.5 Advantages of the Proposed System
The proposed embedded-based accident pre-alert and prevention system offers several
significant advantages over traditional systems:
1. Real-Time Detection: The system operates in real-time, allowing for immediate
detection of accidents and rapid alert generation. This reduces the time between the
occurrence of an accident and the response by emergency services.
2. Higher Accuracy: By leveraging machine learning models trained on extensive
datasets, the system can detect accidents with much higher accuracy, even in complex
or low-visibility conditions.
3. Reduced False Positives: The use of machine learning minimizes the occurrence of
false positives, ensuring that only legitimate accidents trigger alerts.
4. Scalability: The system is highly scalable, allowing it to be deployed across multiple
locations with minimal additional configuration. New cameras and sensors can be
easily integrated into the system.

5. Automated Alerts: Once an accident is detected, the system can automatically alert the
relevant authorities, reducing human intervention and speeding up the emergency
response.
6. Data-Driven Improvements: The system can continually improve over time by learning
from new data, leading to better accident prediction and detection.
4.6 Applications
The proposed system has a wide range of applications in various domains, including:
1. Traffic Monitoring and Management: The system can be used by traffic management
authorities to monitor road conditions in real-time and detect accidents, allowing them
to manage traffic flow more efficiently and dispatch emergency services when
necessary.
2. Smart Cities: As part of smart city initiatives, this system can integrate with other
technologies, such as IoT devices and smart infrastructure, to improve overall road
safety and urban planning.
3. Insurance: Insurance companies can use the system to analyze real-time accident data
for quicker claims processing and to detect fraudulent claims.
4. Public Safety: The system can enhance public safety by reducing accident response
times, preventing secondary accidents, and ensuring that emergency services are
promptly dispatched.
5. Logistics and Fleet Management: Companies in the logistics sector can use the system
to monitor their fleets, ensuring driver safety and responding quickly in case of
accidents.

5.TOOLS AND TECHNIQUES

SOFTWARE REQUIREMENTS
1. Python
2. Pycharm
5.1. Python
In this project, python is used as the programming language. In technical terms,
Python is an object-oriented, high-level programming language with integrated dynamic
semantics primarily for web and app development. It is extremely attractive in the field of
Rapid Application Development because it offers the dynamic typing and binding options.
Python is relatively simple, so it's easy to learn since it requires a unique syntax that
focuses on readability. Developers can read and translate Python code much easier than
other languages. In turn, this reduces the cost of program maintenance and development
because it allows teams to work collaboratively without significant language and
experience barriers.
Python supports the use of modules and packages, which means that programs can
be designed in a modular style and code can be reused across a variety of projects. Once
you've developed a module or package you need, it can be scaled for use in other projects,
and it's easy to import or export these modules. One of the most promising benefits of
Python is that both the standard library and the interpreter are available free of charge, in
both binary and source form. There is no exclusivity either, as Python and all the
necessary tools are available on all major platforms. Therefore, it is an enticing option for
developers who don't want to worry about paying high development costs. If this
description of Python over your head, don't worry. You will understand it soon enough.

What you need to take away from this section is that Python is a programming language
used to develop software on the web and in app form, including mobile. It's relatively
easy to learn, and the necessary tools are available to all free of cost.
5.2 PYCHARM
PyCharm is an integrated development environment (IDE) specifically designed for

programming in Python. Developed by JetBrains, it provides a robust platform for Python
development, offering a wide range of features to simplify coding tasks. These features
include intelligent code completion, error highlighting, refactoring tools, debugging, and
testing support. PyCharm also integrates version control systems like Git, making
collaboration easier.
PyCharm is the Python IDE by JetBrains, designed for professional Python developers. Industry-
leading code completion, code navigation, safe refactoring, and smart debugging are just a few
important features that contribute to make professional software development a more productive
and enjoyable experience. PyCharm Professional Edition comes with wide support for Python
web frameworks, modern JavaScript development, as well as with advanced database tools and
scientific tools integrations.
Pros:
1."I have been using PyCharm since over 4 years now and It is one of the best IDEs out there for
python developers. Great product with auto-complete features."
2. "Also PyCharm have some tools to find errors and helps you to correct them. I also like the
style and colors very nice for me”.
3."The best all in one IDE out there, the python supporting features are great and it has as many
templates for different projects for ease of architecture."
4."PyCharm is probably the best IDE for Python projects as it has so many Python orientated
features. Personally, I only use PyCharm for the occasional Python project and I am very
satisfied with that."
5.3 Software Code or Deployment

5.3.1 Embedded System Code:

(Arduino Example)
// Libraries
#include <Wire.h> // I2C for accelerometer
#include <Adafruit_Sensor.h>
#include <Adafruit_ADXL345_U.h> // Accelerometer library
#include <NewPing.h> // Ultrasonic sensor
// Pin configurations
#define TRIGGER_PIN 7 // Ultrasonic trigger pin
#define ECHO_PIN 6 // Ultrasonic echo pin
#define MAX_DISTANCE 200 // Maximum distance to measure in cm
// Initialize sensors
Adafruit_ADXL345_Unified accel = Adafruit_ADXL345_Unified (12345);
NewPing sonar (TRIGGER_PIN, ECHO_PIN, MAX_DISTANCE);
// Variables
int distance = 0;
float accelX, accelY, accelZ;
// Setup
void setup () {
Serial.begin(9600);
// Initialize accelerometer
If (!accel.begin()) {
Serial.println("Accelerometer not found");
While (1);
}
accel.setRange(ADXL345_RANGE_16_G); // Set accelerometer range
Serial.println("System Initialized");
}
// Main loop
void loop() {
// Ultrasonic sensor for distance measurement

distance = sonar.ping_cm();
Serial.print("Distance: ");
Serial.print(distance);
Serial.println("cm");
// Accelerometer for sudden braking or impact
sensors_event_t event;
accel.getEvent(&event);
accelX = event.acceleration.x;
accelY = event.acceleration.y;
accelZ = event.acceleration.z;
Serial.print("Acceleration X: "); Serial.println(accelX);
Serial.print("Acceleration Y: "); Serial.println(accelY);
Serial.print("Acceleration Z: "); Serial.println(accelZ);
// Send data over serial to be processed by machine learning model
Serial.print(distance);
Serial.print(",");
Serial.print(accelX);
Serial.print(",");
Serial.print(accelY);
Serial.print(",");
Serial.println(accelZ);
Delay (100); // Small delay
}
5.3.2 Machine Learning Part (Python Example)

This part uses a pre-trained model to predict accident risk based on sensor data received from
the embedded system. You can use libraries like TensorFlow, Scikit-learn, or PyTorch.
This example assumes the data is coming through serial communication (e.g., from Arduino
or any other microcontroller) and the machine learning model is pre-trained using labeled
sensor data.
CODE:

import serial
import numpy as np
import joblib # For loading a pre-trained model
import time
# Load pre-trained model (assume it's in joblib format)
model = joblib.load('accident_prevention_model.joblib')
# Open serial connection (adjust to your port)

ser = serial.Serial('/dev/ttyUSB0', 9600)
def preprocess_data(data):
"""
Function to preprocess the data received from sensors
"""
try:
# Convert string to float array
values = np.array([float(i) for i in data.split(",")])
return values.reshape(1, -1) # Reshape for ML model input
except:
return None
def predict_accident(data):
"""
Use ML model to predict accident
"""
# Preprocess and predict
processed_data = preprocess_data(data)
if processed_data is not None:
prediction = model.predict(processed_data)
return prediction [0]
else:
return None
while True:

try:
if ser.in_waiting > 0:
# Read data from serial
sensor_data = ser.readline().decode('utf-8').strip()
print(f"Sensor Data: {sensor_data}")
# Predict accident
prediction = predict_accident(sensor_data)
if prediction == 1:
print ("Accident risk detected! Alert!")
# Code to trigger alerts (e.g., LED, buzzer)
else:
print ("No accident risk.")
time.sleep(1) # Adjust frequency as needed
except Exception as e:
print (f"Error: {e}")
5.3.3 Machine Learning Model Training

You would train a model using data from the sensors, where the features include distance,
accelerometer values, and labels (e.g., accident/no accident).
Here’s a simple example in Python (using Scikit-learn):
CODE:
from sklearn.ensemble import RandomForestClassifier
from sklearn.model_selection import train_test_split
from sklearn.metrics import accuracy_score
import joblib
# Sample data (distance, accelX, accelY, accelZ, label)
data = [
[10, 1.2, 0.5, 0.6, 1], # Accident
[100, 0.1, 0.2, 0.3, 0], # No accident
[5, 1.5, 1.8, 2.0, 1], # Accident
[150, 0.0, 0.1, 0.2, 0], # No accident

# Add more data samples here

]
X = [row [-1] for row in data] # Features
y = [row [-1] for row in data] # Labels
# Split data into training and test sets
X_train, X_test, y_train, y_test = train_test_split (X, y, test_size=0.2)
# Train model
model = RandomForestClassifier ()
model.fit (X_train, y_train)
# Test model
predictions = model.predict (X_test)
accuracy = accuracy_score (y_test, predictions)
print (f"Model Accuracy: {accuracy * 100:.2f} %")
# Save the model
joblib.dump(model, 'accident_prevention_model.joblib')

6.FINAL RESULTS
We take the input video from a Source (CC TV camera) and divide the video into several frames.
On each frame a Vehicle is detected whether Accident has occurred or not. If accident occurred it
will print the image number with “accident” message.
6.1 FINAL RESULTS
Fig 6.1: Accident occurred

This figure shows a clear scenario where an accident has been detected. Two vehicles are shown
in a collision, confirming that the system accurately identifies situations where an accident has
taken place. The CCTV camera captures the incident in real time, and the system, after dividing
the video into frames, successfully detects the accident. The detection is confirmed when the
system prints the relevant frame along with the "accident" message.

Fig 6.2 Accident will happen

In this figure, the system predicts that an accident is imminent. A vehicle is shown traveling in
the wrong direction, and the machine learning model detects this deviation from the norm. As a
result, the system flags the situation as an impending accident.
This prediction capability is critical for accident prevention. By identifying situations where an
accident is likely to occur, authorities can take preventive measures, such as issuing warnings to
drivers or adjusting traffic signals to reduce the risk of a collision. This feature could prove
especially valuable in high-risk areas like highways or intersections where accidents are more
likely to happen due to speed or traffic violations.
Fig 6.3: No of Accident occurred
In this picture we got 4 images displayed with name accident, because in that video 4 accidents
has happened so we got displayed that. In this figure, the system displays multiple frames, none
of which indicate an accident. The output demonstrates that, for the majority of the video, the
system did not detect any abnormal events or collisions. The accuracy of the system in these non-
accident scenarios is crucial. A high rate of false positives—where the system incorrectly

identifies an accident—could lead to unnecessary interventions, resource allocation, and

inefficiencies.
The system's success in avoiding false positives ensures that it can be relied upon for real-world
applications. Emergency response teams, for example, can trust that an accident alert represents a
genuine incident, allowing for quicker and more effective responses.
7.CHAPTER
7.1 CONCLUSION
The embedded-based smart accident pre-alert and prevention system with machine learning
offers a promising solution for enhancing road safety by integrating advanced technology into
real-time accident prediction and prevention mechanisms. This system uses sensors to monitor
vehicle parameters, driver behavior, and environmental conditions, feeding this data into
machine learning algorithms to identify potential accident risks. By providing timely alerts to
drivers and potentially activating preventive measures, it can reduce the likelihood of accidents,
saving lives and minimizing injuries. The continuous learning capabilities of machine learning
models enable the system to evolve, becoming more accurate and effective over time, thus
offering a proactive approach to road safety.
7.2 FUTURE WORK
The future work of an embedded-based smart accident pre-alert and prevention system with
machine learning focuses on enhancing real-time data processing, predictive accuracy, and
system adaptability. Future developments will likely integrate advanced sensors such as LiDAR,
RADAR, and IoT-enabled devices to gather comprehensive environmental and vehicular data.
Machine learning algorithms, especially deep learning models, could improve the system's
ability to predict accidents by identifying complex patterns in driver behavior, road conditions,
and traffic dynamics. The use of edge computing will ensure faster response times by processing
data locally, while cloud connectivity will facilitate continuous updates and large-scale data

analysis. Moreover, incorporating adaptive learning models will allow the system to improve
over time as it encounters new scenarios, leading to more personalized and precise pre-alerts and
prevention measures. Enhanced vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I)
communication will also play a crucial role in accident prevention by enabling coordinated
responses across multiple vehicles.
7.3 SCOPE OF FURTHER STUDY

The future scope of an Embedded-based Smart Accident Pre-Alert and Prevention System
with Machine Learning is vast and promising. By integrating machine learning algorithms,
the system can continuously learn from real-time data, improving its accuracy in predicting
potential accidents. This can be expanded to cover a broader range of vehicles, road
conditions, and driver behaviors, making the system adaptable to diverse environments.
Further research can also focus on integrating advanced sensors, such as LiDAR and radar, to
enhance detection capabilities. Additionally, with the rise of autonomous vehicles, this
technology could be pivotal in ensuring safer navigation. Cloud-based solutions can enable
data sharing between vehicles, creating a network of interconnected systems that can predict
and prevent accidents at a larger scale. Lastly, combining it with IoT and 5G technologies
could allow for faster communication and more responsive safety mechanisms, further
reducing accident rates.

8.REFERENCES
1. Wang S, Liu Y, Zhu M, Zhang J. A machine learning approach for smart vehicle accident
prevention using sensor data. IEEE Trans Veh Technol. 2020; 69(2): 1053-1065.
2. Suryawanshi M, Kulkarni V. Embedded accident prevention system for vehicles with pre-
alert functionality. Int J Embedded Syst. 2019; 7(4): 315-322.
3. Ponnusamy K, Varghese B. Intelligent accident prevention system using IoT and machine
learning techniques. IEEE Access. 2021; 9: 57345-57354.
4. Mehta A, Chatterjee R. Real-time traffic accident detection and alert system using machine
learning algorithms. Int J Adv Comput Sci Appl. 2020; 11(3): 123-130.
5. World Health Organization. Global status report on road safety 2018. World Health
Organization, 2018. [Online]. Available: https://www.who.int/roadsafety. Accessed: 15-Jul-
2020.
6. Ramesh S, Bhaskar A, Kumar S. Intelligent accident pre-detection and alert system using
sensor fusion. Sensors. 2021; 21(5): 1987.
7. Gupta P, Mishra R, Verma K. Machine learning based accident prevention in connected
vehicles using edge computing. IEEE Trans Intell Transp Syst. 2019; 20(5): 1410-1420.
8. Sharma K, Tiwari R. Embedded GSM-based accident alert system with machine learning for
real-time vehicle monitoring. J Intell Veh Technol. 2020; 3(2): 76-85.
9. Kaur G, Singh P. A novel accident prevention and detection system using deep learning
techniques for autonomous vehicles. J Adv Res Comput Sci. 2021; 12(1): 45-54.

10. Zhang H, Xu Y, Li M. A comprehensive survey of machine learning techniques for accident

detection and prevention in intelligent transportation systems. IEEE Trans Emerg Top
Comput Intell. 2021; 5(4): 652-666.

Own Studt Project

Uploaded by

Copyright:

Available Formats

Own Studt Project

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Own Studt Project

Uploaded by

Copyright:

Available Formats

A Self Study Project Report On

EMBEDDED BASED SMART ACCIDENT PRE-ALERT AND PREVENTION

University College Of Engineering& Technology

In partial fulfillment of the requirements for the award of

Under the guidance of

DEPARTMENT OF ELECTRONICS AND COMMUNICATIONS ENGINEERING

DEPARTMENT OF ELECTRONICS AND COMMUNICATION ENGINEERING

UNIVERSITY COLLEGE OF ENGINEERING & TECHNOLOGY

MAHATMA GANDHI UNIVERSITY, NALGONDA

UNIVERSITY COLLEGE OF ENGINEERING & TECHNOLOGY

Nalgonda – 508001, Telangana

DEPARTMENT OF ELECTRONICS AND COMMUNICATION ENGINEERING

Project Guide: Head of the Department:

G. Anil Kumar (4511-21-735-047)

1. Introduction to Embedded system

1.6 Difference between General Purpose 4

1.2 System-on-chip Architecture Components 2

1.5 Structure of an Embedded System 3

1.6.1 Embedded computer system block diagram 4

1.7 Von-Neumann and Harvard Architectures 6

1.8 Types of an Embedded system 7

1.9 Applications of an Embedded system 9

2.3 Machine learning process 10

2.4 Types of machine learning 12

4.1 System Architecture 17

6.1 Accident occurred 29

6.2 Accident will happen 29

6.3 No. of Accidents occurred 30

1. INTRODUCTION TO EMBEDDED SYSTEMS

1.1 Embedded System

An embedded system can be thought of as a computer hardware system having

An embedded system has

1.2 Characteristics of an Embedded System:

• Single-functioned − An embedded system usually repeats a specialized

• Tightly constrained − All computing systems have constraints on design

Dept of ECE, UCET, MGU, Nalgonda 1

decelerations repeatedly within a limited time; a delayed computation can result

• Microprocessors based − It must be microprocessor or microcontroller based.

• Memory − It must have a memory, as its software usually embeds in ROM.

• Connected − It must have connected peripherals to connect input and output

1.2 Figure: System-on-chip Architecture Components

Dept of ECE, UCET, MGU, Nalgonda 2

1.5 Basic Structure of an Embedded System

The following illustration shows the basic structure of an embedded system −

1.5 Figure: Structure of an Embedded System

• A-D Converter − An analog-to-digital converter converts the analog signal sent

• D-A Converter − A digital-to-analog converter converts the digital data fed by

Dept of ECE, UCET, MGU, Nalgonda 3

1.6 Difference between General purpose computer and Embedded systems:

General Purpose Computer Embedded Systems

It is designed using a microprocessor as the It is mostly designed using a microcontroller as

It is mostly costlier compared to the

It is cheaper as it requires less no of peripheral

Dept of ECE, UCET, MGU, Nalgonda 4

1.7.1 Von Neumann Architecture

1.7.2 Harvard Architecture

Dept of ECE, UCET, MGU, Nalgonda 5

1.7 figure: Von-Neumann and Harvard Architectures

Simple in design. Complex in design.

1.8 Types of Embedded systems