“DESIGN OF VOICE CONTROLLED ROBOT

AND BREAKDOWN ANALYSIS”

A THESIS REPORT
SUBMITTED IN PARTIAL FULFILLMENT OF THE
REQUIREMENTS FOR THE AWARD

OF
BACHELOR OF ENGINEERING
IN
DIVISION OF MANUFACTURING PROCESSES AND
AUTOMATION ENGINEERING

Submitted by:
KARAN PANCHAL 2018UMP1049
AYUSH SHARMA 2018UMP3598
PRIYANSHU NANDAN 2018UMP3603
ADITYA SRIVASTAVA 2018UMP3615

Under the supervision

of Dr. SIMRANJEET SINGH AND Dr. VIVEK KUMAR

DIVISION OF MANUFACTURING PROCESSES

AND AUTOMATION ENGINEERING
NETAJI SUBHAS INSTITUTE OF TECHNOLOGY
UNIVERSITY OF DELHI
MAY, 2022
 CERTIFICATE OF ORIGINALITY

We, Karan Panchal 2018UMP1049 , Ayush Sharma 2018UMP3598, Priyanshu

Nandan 2018UMP3603, Aditya Srivastava 2018UMP3615 of B.E.

Manufacturing Processes And Automation Engineering, hereby declare that the

Project-Thesis titled “DESIGN OF VOICE CONTROLLED ROBOT AND

BREAKDOWN ANALYSIS” which is submitted by us to the Division of

Manufacturing Processes and Automation Engineering, Netaji Subhas Institute

of Technology, Delhi(University of Delhi) in partial fulfillment of the

requirement for the award of the degree of Bachelor of Technology, is original

and not copied from source without proper citation. The manuscript has been

subjected to plagiarism check by Turnitin software. This work has not

previously formed the basis for the award of any Degree.

Place: Delhi

KARAN PANCHAL

AYUSH SHARMA

PRIYANSHU NANDAN

ADITYA SRIVASTAVA
 DIVISION OF MANUFACTURING PROCESSES AND AUTOMATION ENGINEERING

Netaji Subhas University of Technology

Dwarka, Delhi-110078, India

CERTIFICATE OF DECLARATION
This is to certify that the Project-Thesis titled “DESIGN OF VOICE

CONTROLLED ROBOT AND BREAKDOWN ANALYSIS” which is being

submitted by Karan Panchal 2018UMP1049 , Ayush Sharma 2018UMP3598,

Priyanshu Nandan 2018UMP3603, Aditya Srivastava 2018UMP3615 to the

Department of Manufacturing Processes and Automation Engineering, Netaji

Subhas Institute of Technology (university of Delhi) in partial fulfillment of the

requirement for the award of the degree of Bachelor of Technology, is a record

of the thesis work carried out by the students under our supervision and

guidance. The content of this thesis, in full or in parts, have not been submitted

for any other degree or diploma.

Place: Delhi

NAME OF SUPERVISORS : Dr. Simranjeet Singh

Dr. Vivek Kumar

 ACKNOWLEDGEMENT

This report has been prepared for project titled “DESIGN OF VOICE

CONTROLLED ROBOT AND BREAKDOWN ANALYSIS” that has been done

with the purpose of fulfilling the requirements of the course of B.E.

MANUFACTURING PROCESS AND AUTOMATION ENGINEERING .

We would love to thank our Project Leader – Dr. SIMRANJEET SINGH and Dr.

VIVEK KUMAR for his benevolent guidance and providing us all the knowledge on

this project topic and for his constant motivation throughout the journey of this

project.

We would also like to pass on our true gratitude to the Department Of Manufacturing

Process and Automation Engineering, NETAJI SUBHAS UNIVERSITY OF

TECHNOLOGY, NEW DELHI.

We would also like to thank our classmates, friends and family who have always

been our support in these tough times and were always available for help.We hope

our report will reflect the knowledge we gained through this project and the hard

work and innovation we put in for completion of the same.

- KARAN PANCHAL (2018UMP1049)

- AYUSH SHARMA (2018UMP3598)

- PRIYANSHU NANDAN (2018UMP3603)

- ADITYA SRIVASTAVA (2018UMP3615)

ABSTRACT:

Unplanned outages have proven to be one of the main factors that reduce process
capacity and reduce product quality and customer satisfaction in the manufacturing
industry. These turmoil lasted an average of four hours over the last three years and
cost $ 2 million. Unplanned downtime leads to goodwill, customer trust, and lost
productivity. 46% were unable to serve their customers. Use CNN modeling to
create real-time feedback software for machines that show subtle changes in
behavior before failure, such as frequency, force, and noise signals. Signals are
used by software to predict failures in real time and alert operators. Using this
model, we were able to detect punctures with an accuracy of over 99% on
machines that have been studied for a long time, that is, machines with relatively
little data available for inspection.

KEYWORDS: Robotics, speech recognition, breakdown detection, Convolutional

Neural Networking
 TABLE OF CONTENTS

S.NO CONTENT PAGE NO.

Certificate Of Originality

Certificate Of Declaration

Acknowledgement

Abstract

1.0 Introduction 1

2.0 Problem Statement 2

3.0 Methodology And Work Elements 3

3.1 Speech Recognition 3

3.2 Working Algorithm 3

3.3 Pre Programmed Commands 4

3.4 Design And Implementation 5

3.5 Hardware Used 5

3.5.1 Arduino Uno 6

3.5.2 Bluetooth Module 6

3.5.3 H Bridge 7

3.5.4 Hc Sr04 Ultrasonic Sensor 8

3.5.5 DC Motors 9

3.6 Software Used 10

3.7 Circuit Diagram 11

3.7.1 Arduino Connections 13

3.7.2 Bluetooth Module Connections 15

3.8 Working Code 16

3.9 Working Of Bluetooth Application 21

3.10 Flow Chart For Robot Working 25

3.11 Fmea 26

3.11.1 Types Of Fmea 27

3.11.2 Relationship Between Different 28

FMEA’s

3.11.3 Procedure To Conduct Fmea 29

3.11.4 Severity Ranking 30

3.11.5 Quantitative Techniques 32

3.11.6 Tools And Methodologies Used In Fmea 32

3.11.7 FMEA Chart For Project 34

3.12 Gantt Chart For Project 37

4.0 Conclusion And Future Scope 38

References 40
LIST OF FIGURES
CHAPTER 3 page no.

Figure 3.1: HC SR04 Ultrasonic Sensor 5

Figure 3.2: Arduino Uno 6

Figure 3.3: Bluetooth Module 7

Figure 3.4: L 293 Bridge 8

Figure 3.5: Connections Of Ultrasonic Sensors 9

Figure 3.6: DC Motors 10

Figure 3.7: Circuit Diagram 11

Figure 3.8: Jumper Wires 12

Figure 3.9: Actual Connections Of Model With IR 13

Figure 3.10: Dc Motor Connections With Driver 14

Figure 3.11: Arduino Connections With IR Sensor 15

Figure 3.12: Working Program 21

Figure 3.13: Bluetooth Mobile Application 22

Figure 3.14: Bluetooth 23

Figure 3.15: Speech Conversion Using Bluetooth 23

Figure 3.16: Bluetooth Module Connection To Application 24

Figure 3.17: Flowchart Of Robot Working 25

Figure 3.18: Relationship Between Different Types Of FMEA’s 28

Figure 3.19: Working Of FMEA 29

LIST OF TABLES

chapter 3: page no.

TABLE 3.1 : Severity ranking table 30

TABLE 3.2 : FMEA chart for project 34

CHAPTER 1

INTRODUCTION
Robots and technology are evolving at a brisk pace and the world has transformed into
a competitive work environment with a very fine margin of error. With the advent of
robots in every field it has become a hefty task to manage and automate these robots to
work on optimum levels and reduce the breakdowns and to reduce the pause.

In this project we aim to provide a seamless way to operate the robots used in
manufacturing and various other fields using speech recognition i.e.voice commands
from the user controls the movements of the vehicle.

These commands or different functions are read using an Android App on the user’s
phone which is connected to the robot using a (wireless) Bluetooth Module. The
commands are first encoded and then transmitted over an RF (Radio Frequency)
channel and before receiving they are decoded by the Bluetooth Module.

The sole aim of Human Voice Controlled Robot (HVCR) is to immediately act on the
inputs received from the user in the form of commands, which microcontroller already
contains beforeto. To detect the obstacle, the most convenient way is to employ a
ultrasonic sensor on top of the chassis as as to track the breakdown much before it
happens and thus save us from downtime. CNN Algorithm is incorporated to predict
the breakdown of components using component sound.

Convolutional neural networks (CNNs) are mainly used to open the doors for learning
between images or videos and a desired label or output. The CNN in the model is
incorporated using Python and Keras and increasing the accuracy of CNN Model
while reducing the Dataset size.The application of this CNN will be applied to
improve real time breakdown detection using robots in manufacturing to increase
profits of the manufacturing sector and ensure lesser breakdowns and significant
improvement in process performance due to lower disruptions while manufacturing. In
this competitive world even the slightest increase in the profit percentage gives
significant benefit when compared to competition

page 1
CHAPTER 2 :

PROBLEM STATEMENT
Robots are used in manufacturing all over the world and are an essential
component of Computer Integrated Manufacturing Systems. It approximates about
one third at the global level that the industrial robot fulfills the aim of industries to
ship the products across the different countries of the world.

If we see globally, approximate cost of downtime in automobile industry only , just

for a single hour may cost us upto 3 million USD. But even after so much efforts, the
true cost to us is unknown globally for the downtime. It is widely believed that more
than half of the industrial machineries fails in estimating the downtime without any
error. An approximate idea is of upto 5 - 20%, depending upon the type of industries,
loss in productivity just because of the downtime. Machines are not solely
responsible, humans are also at the fault and accounts to approximate 23% of the
unplanned downtime.

This project will precisely address the use of speech control in Industrial robotics in
the field of manufacturing and to find a way using Convolutional Neural Networks for
early detection of machine faults to predict machine breakdown and thus reduce
downtime cost as this cost is significant.

The Objective of this project will be to: Use and integrate various sensors and modules
to integrate speech in manufacturing industrial robotics, and to develop a working code
to detect early mechanical breakdown by detecting noise anomalies of the machine.

NORMALIZATION:

All the features and parameters used in the program were normalized by dividing the
dataset with the maximum value to get parameters 0 and 1 for better speed, reliability
and accuracy of the neural network. Normalization was used with the help of unit
standard deviation and zero mean. Training time increased for number of features but
not proportionally.

page 2
CHAPTER 3:

METHODOLOGY AND WORK ELEMENTS

3.1 Speech Recognition

When a machine has the ability to sense the word spoken and convert them into the text,
then this transition of the speech to text is known as the speech recognition. Each and every
sound is analyzed before converting it into the text employing the algorithm based on the
machine learning. It also employs use of artificial intelligence upto small extent. It works
on the premise of algorithms that matches the sound of the set speech with word sequences
and interpret it as a command.
However various factors will have an effect on the performance of speech recognition like
numerous accents, the distinction in pitch and volume, manner of pronunciation, and
conjointly background signal.

3.2 Working Algorithm of The Robot

step 1: Start.
step 2: Provide connections between BT module and the bluetooth application and when
connected we have to provide predefined commands with the help of mobiles.
step 3: Characters that are scanned from a single word on the robot application.
step 4: This is going to be sent wirelessly to Bluetooth module HC-05.
step 5: HC-05 can automatically transmit it serially to the Arduino UNO. Arduino UNO
can perform action on the DC motors as per the command read. Obstacles will be
automatically discovered by the ultrasonic sensor when they come in a certain range. as
soon as they come in range, speed will be decreased and robot comes to a halt.
step 6: By the time we are able to give the succeeding command
step 7: Stop

page 3
3.3 Preprogrammed Commands for Robot:
Forward: start each motor at low speed and move it forward.

Backward: starts each motor at low speed and moves it backward.

Right: starts left motor and take a left spin • Left: starts right motor and take aright spin

Stop: stops all motors.

3.4 Design and Implementation

Power Supply Converts nine volts to five volts and five volts to 3.3 volts.

H-bridge - The engine control and regulator depends on L298D. The left engine is related
to input 1 and 2 while the right engine is associated with input 3 and information data 4.
Alter or the enable pin of L298D is related to the advanced pin number 6 of Arduino Uno
for speed management.

Ultrasonic sensing element - It can detect the obstacles with in the range of 400cm and
shunning is HC SR04. The results are accurate maximum time. It consists of ultrasonic
transmitter, receiver and the negative control feedback circuit and has worked by causation
pulse and check echo to check the gap between the obstacle Emitter and detector of
Ultrasonic sensor can produce 40 kHz sound wave and detect same frequency additionally
and send the received electrical signal back to the microcontroller

page 4
 FIGURE 3.1

3.5 Hardware Used

3.5.1 Arduino UNO-

The Arduino Uno is an open-source microcontroller. A microcontroller may be a compact

IC designed to try to do a unique jak na system The role here is to retuit to basse
commands accordingly, that is pre-programmed within the microcontroller and is joined
with the motor driver. ATmega328P chip is used in supporting the Uno as it has 14 number
of pins for providng digital outputs and inputs and six pins for analogue inputs. It also
contains a reset button, a power jack, sixteen MHz quartz crystal, an ICSP header, a USB
connection.

page 5
 Figure 3.2

3.5.2 Bluetooth Module HC-05

HC 65 is a Bluetooth module which is used to schipa wireless communication with the

device. It is either in the slave configuration or a master configuration depending upon the
red light. when the connection establishment is in process, the red light is continuously on.
Even before setting up a connection the RED light periodically blinks. After setting up the
connection, blinking speed reduces to once in every 2 seconds. The module usually
operated on 3.3 V. A 5V supply voltage can also work as we have on board 5 to 3.3 V
regulators. All the RX/TX pins usually operate on 3.3 V level and the microcontroller also
can detect 3.3 V level.

page 6
 FIGURE 3.3

3.5.3 L293D Bridge-

The L298 Driver is also known as an H-bridge. we can drive stepper moros and DC
motors with this l293 bridge as it is a motor driver. It has a dual bridge driver having high
current as well as high voltage. Transistor logic (TTL) levels are in the plan to acknwle
dge standard Transistor . It has two enable out of which one pin is used to enable the
device while the other is used to disable. This module has the capacity to drive any DC
motors with voltage ranging from 5 to 35 V and can tolerate peaking current up to 2 A.
This driver is used to control the direction and working speed simultaneously.

page 7
 Figure 3.4

3.5.4 HC-SR04 Ultrasonic Sensor

HC-SR04 is a distance sensor, which can be used with both microcontroller platforms like
Arduino and microprocessors like Raspberry Pie. With the help of sonar, it is used to -
find the distance. Using non-contact ultrasonic sonar, it measures the distance between an
object and itself. It has two ultrasonic transmitters (basically speakens), one is a receiver,
and another one is a control circuit. The transmitter transmits a high-frequency ultrasonic
sound which after hitting returned back to the receiver, and the receiver receives the
returned echo. To compute the time difference between the signal first transmitted and the
obtained echo is calculated by the control circuit, to figure the distance between the sensor
and the reflecting body.

page 8
 Figure 3.5

3.5.5 DC Motors

A DC Motor is a motor which works on direct current I is a rector which operates at direct
current and converts electrical energy into mechanical energy The master consists of a
home shoc permanent magnet
and a tuning coil which is known as armature Loop of wire with a current going through it
produces an electromagnetic field lined up with the focal point of the curl. The heading and
greatness of the attractive field delivered by the curl can be altered with the course and size
of the current moving through it.

page 9
 Figure 3.6

3.6 Software Used

Arduino IDE The Arduino Integrated Development Environment Arduino Software (IDE)
has a text editor for writing and editing code, a console for text, a toolbar with catches
containing basic capacities and a progression of menus. The projects are then transferred to
the Arduino equipment. Projects composed are called portraits which are written in the
content manager and are saved with the record extension (ino). The proof reader gives
highlights, for example, cutting/gluing and looking/supplanting text. The message region
gives input while saving and furthermore shows mistakes. The reassignment shows text
yield by the IDE. The arranged board and sequential port are shown at the base right-hand
corner of the window. The check and transferring of the projects, manifestations, opening,
and saving representations, and opening the chronic screen should be possible utilizing the
catches of the toolbar.

• Android Application - The Android Application lets you connect to HC-05 module via
Bluetooth and allows you to send signals to the Arduino so that it can perform desired
operations successfully.

page 10
3.7 CIRCUIT DIAGRAM FOR CONNECTION OF
HARDWARE COMPONENTS

FIGURE 3.7

page 11
Before making the connection, you should be sure about the right type of jumper wires.
Jumper wires have the connector pins at each end. As many electronic applications require
the soldering for a connection, but sometimes you need to disassemble those connections
due to any update in the application or project. So, jumper wires are the best choices for
such applications.
These jumper wires are available in different colors, this color property doesn’t mean
anything but this property of the jumper wires can be used to clear the doubt about the
various connections.There are three types of jumper wires female-to-female in this type the
connector pins will not be available at both the ends, there will be empty connector pins,
male-to-female, in this type, one end has a connector pin and the other end has an empty
pin (a connector pin of male type can be inserted into this empty pin) & at last
male-to-male.

FIGURE 3.8

page 12
REAL MODEL & THE CIRCUIT CONNECTIONS:

FIGURE 3.9

page 13
3.7.1 CONNECTION OF MOTORS WITH MOTOR DRIVER &
ARDUINO UNO:
The battery is connected to the motor driver module directly, the red wire of the battery is
connected to the +ve pole of the motor driver module & the black wire of the battery is
connected to the -ve pole of the motor driver module. Now the two wires of the left motor
and the other two wires of the right motor are connected to the motor driver’s MOTOUT2
and MOTOUT1 ports respectively.Now for connecting the arduino uno, as we had
assigned the pins 8,9 ,10 & 11of the arduino board in our program code, so the pins 8,9 are
connected to MC6 and the pins 10,11 are connected to MC5 respectively. It is not
mandatory that the pin numbers are fixed, later we can change the pins according to our
program & the requirements.

page 14
 FIGURE 3.10

3.7.2 CONNECTION OF BLUETOOTH MODULE TO ARDUINO UNO:

In our program we have defined pins 2,3 for the communication of the bluetooth
module with the arduino uno. The pins 2, 3 of arduino uno are connected to the TX
& RX of the bluetooth module which are known as the receiver & transmitter pins of
the bluetooth module. There are 4 pins in the bluetooth module, & remaining two
pins are GND & VCC

GND is the ground pin which is connected to the GND(anyone GND pin of the

page 15
arduino, since there are two GND pins) of the arduino uno & the VCC pin is
connected to the 5V pin on the arduino UNO, if we switch on the battery, then there
will be a red light on the bluetooth module which confirms the correct connections
has made.The connection of the IR sensor with the Arduino UNO board.

The IR sensor is directly connected to the battery, & the VCC pin of the IR sensor
is connected to the 5V of the Arduino UNO Board, GND of the IR sensor is
connected to the GND of the Arduino UNO.

FIGURE 3.1

page 16
3.8 WORKING CODE

1. WORKING CODE FOR ROBOT

Before using the Arduino IDE we will need some libraries, these libraries can be included
in our project by using the Arduino tool tab which is given below:
Open your project and go to sketch:

then select include libraries and then select manage libraries

page 17
After clicking on the Manage libraries there will be another window pop up on the screen
which is Library Manager. Now in the Library Manager, we can search and install any
available library.

The softwareSerial library will help the programme to communicate with the digital pins of
the arduino uno board, meaning of (softwareSerial BT(2, 3)) this statement is that we have
defined the pin no 2, 3 for the bluetooth module. In the same way 8,9,10 & 11 are the pins
on the arduino uno which are assigned to the left & the right motor respectively. The
variable “readvoice” is a string or group of characters or integers, which will hold the
modified voice command using the speech recognition technique.

page 18
page 19
page 20
 FIGURE 3.12

After writing the code and making the connection of the working model, before uploading
the program into the arduino uno board, make sure that all the connection are tight and for
uploading the program first check whether the USB port is selected or not, this can be done
by going to arduino ide’s tool tab.
Also before uploading the code, first compile the program to overcome unwanted errors.
Finally, upload the program after connecting the arduino uno board to the computer, as
program uploaded, there is a LED bulb on the arduino uno board which will blink on the
successful transmission of the code, & the IDE’S UI will also show that the program has
uploaded successfully.

3.9 WORKING OF BLUETOOTH

APPLICATION
Arduino Bluetooth Voice Control Application
This is an android application provided by Unicia Robotics and it is available on play store

page 21
for free.
Using this application one can send voice commands to HC-05 Bluetooth module.

FIGURE 3.13

With the help of this app we can transmit integers, characters or strings, using voice.
After installing the app click on the bluetooth icon to see the available devices & tap on the
required device.
Before using this app, make sure that the connections of your model should be proper &
there should be sufficient supply of power for proper working of the components, else this
app will throw some errors.This app works on the principle of Speech Recognition
Technology, which is the process of transmitting sound into strings.

There are three branches of SR:

Automatic Speech Recognition:
Natural Language Processing:

Text-to-Speech: following image shows the UI, that the application takes the command for
processing.

page 22
 FIGURE 3.14

bluetooth pairing: following image shows us the UI of the available devices.

FIGURE 3.15

page 23
After selecting the device, it may ask for the password to pair up, and this can be done by
entering 0000 or 1234. When the pairing has finished it will show the UI as shown in
screenshot given below.

FIGURE 3.16

page 24
3.10

FIGURE 3.17

page 25
3.11 FMEA (Failure mode and effect analysis):

Failure modes and effects analysis (FMEA) is a task which aims to identify each possible

failure mode right from the design stage to the end stage after manufacturing. It was

introduced by the U.S. military during WWII. Concept of FMEA is purely based on

reliability engineering.

What we mean by failure mode is the way or ways by which a device or a system fails to
do its intended function or may fail to do so in the coming future.
Effect analysis basically deals with the detailed elaboration and judgment of all the
possible changes that may come into effect when the system fails by a certain mode of
failure.
In this whole process of Failure modes and effects analysis (FMEA), a list is usually
prepared with all the possible reasons of failure and they are ranked based on their
severity. Severity may differ based on the mode of failure. Severity might range from
“very minor” to“catastrophic” with very minor representing the almost no effect of the
failure to the almost complete failure of system in multiple ways being represented by the
catastrophic level of severity. These severity levels could also be represented using
numerals representing the severity in ranking on the basis of level of seriousness of
severity.

Reason for conducting FMEA : FMEA always aims at minimizing the risk
and losses when a machine, product or a system is used just by assessing and analyzing all
the possible modes of failures. Failures sometime depending upon their severity may
become risky for the operator hence there is a need to assess the level of risk associated
with the operation of a particular machine or system. Apart from risk, failures also results
in the downtime ranging from very short to very long downtime of several hours resulting
in high losses. In order to minimize the losses, FMEA is performed.
.

page 26
 3.11.1 Types of FMEA - we can categorize FMEA into three types as follows :
• Design FMEA

• Process FMEA (PFMEA)

• Functional FMEA (FFMEA) or System FMEA (SFMEA)

Sometimes apart from all these categories, service FMEA is also categorized as one of the
type of FMEA. But service FMEA is majorly suitable for the product based companies in
the IT sector. With the growing pace of time, service FMEA will be definitely and
generally accepted by all the companies worldwide.

Different FMEAs in detail:

1. System FMEA : Main focus in this FMEA are systems and subsystems. Interaction
between the system and different elements of the system is studied in this type of FMEA
System FMEA can be beneficial in the following ways:
a) it can reduce redundancy.
b) it can help in chalking out the best system level design.
c) it can help in determining the possible system failures and hence can mitigate the risk

2. Design FMEA : It is used at the initial stage before the manufacturing of a certain
product is started. All the aspects of failures are considered and are analyzed along with the
proper solution for each mode of failure. Design FMEA can be beneficial in the following
ways:
a) It prioritizes the actions that are to be taken necessarily for design improvement.
b) it helps in determination of primary characteristics.
c) it helps in eradicating the design features from a safety point of view.
d) it helps in determination of alternative design solutions.
e) it eliminates the risk of product failure at the very early stage of product failure.

3. Process FMEA : It aims majorly at eradicating the possible flaws leading to failures at
the manufacturing and assembling stage. Each process involved in the manufacturing is
analyzed with great depth.

Process FMEA can be beneficial in the following ways:

page 27
a) different possible flaws in the manufacturing process are identified.

b) It prioritizes the actions that are to be taken necessarily for improvement.

c) It helps in determination of alternative manufacturing techniques or solutions.

3.11.2 Relationship between system, design and

process FMEA

FIGURE 3.18

page 28
3.11.3 Procedure to conduct FMEA: usually it involves 7 steps to conduct FMEA.

FIGURE 3.19: FMEA WORKING

Following steps are included in FMEA process:

System Analysis

● Step 1: Planning & Preparation : At this stage, scope of study is defined. This
step is completed by the management team.

● Step 2: Structure Analysis : This step aims at understanding the complete

process and further breaking down the process into the sequential steps and some
logical elements.

● Step 3: Function Analysis : this step aims at analyzing each element

separately based on their requirements and the function which they perform.

Failure Analysis and Risk Mitigation

● Step 4: Failure Analysis : This step involves analyzing the focus element which
results in the failure of the system to achieve the intended function. At this stage, focus is
mainly on three things – failure mode, failure effect and failure cause. This is the most
important stage directly aiming to identify the potential causes of failure.
● Step 5: Risk Analysis: This stage aims at assessing the risk associated with each
of the modes of failure identified at the previous stage. These failures are then arranged

page 29
 in a table known as severity table, ranking each mode of failure based on the assessed
harm that it could do to the system.
● Step 6: Optimization : At this stage, we need to find steps that need to be taken
to avoid all the modes of failure and thus bringing the system at minimal risk when
operated.

Risk Communication

● Step 7: Result Documentation : Final results are to be documented at this stage

and this stage holds equal importance to all other steps.

3.11.4 SEVERITY RANKING : It is a process of listing all the failure

modes in a way such that each failure mode is assigned a certain ranking which denotes
the level of severity of each mode of failure.

RANKING EFFECT DESIGN FMEA PROCESS FMEA

SEVERITY SEVERITY

10 Hazardous –without affects safe may endanger

warning operation without machine or
warning operator without
warning

9 Hazardous – with affects safe may endanger

warning operation with machine
or operator with
warning warning

8 Very high makes product major disruption in

inoperable operations

7 High makes product minor disruption in

operable at operations (may
reduced require sorting and

page 30
 performance some scrap)
(customer
dissatisfaction)

6 moderate results in customer minor disruption

discomfort in operations (no
sorting but some
scrap)

5 Low results in comfort minor disruption in

and convenience at operations (portion
a reduced level may require
rework)

4 Very low results in minor disruption in

dissatisfaction by operations (some
most customers. sorting and portion
may require
rework)

3 minor results in minor disruption

dissatisfaction by (some rework but
the average little effect on
customer. production rate)

2 Very minor results in minor disruption

dissatisfaction by (minimal effect on
few customers. production rate)

1 None No effect No effect

TABLE 3.1

page 31
3.11.5 QUANTITATIVE TECHNIQUES

There are three major classifications:

1. Time series analysis and projections : This method is completely dependent

upon the accuracy of the past data. Data of some of the previous years is rigorously
analyzed in this technique and the situation for the current scenario is estimated by
extrapolating the past data. This technique has some of the limitations. This technique
shows accurate results if the past relationship along with all the parameters remains the
same. Once it fulfills this criteria, then this technique is very fast in obtaining the
accurate results.

2. Regression analysis : By determining the relationships between the

activities and the factors which create demand, one will be able to forecast different
modes of failures and

hence this method is a very important quantitative technique for FMEA. Weightage to
each factor is assigned suitably depending upon the importance of each factor.

3. Econometrics

3.11.6 TOOLS or METHODOLOGIES used frequently

in FMEA

1. Affinity Diagram : this consists of small cards, each containing some idea or
solution to the problem.
2. Box and Whisker Plot : its appearance is similar to the rectangles, containing two
lines passing through its center. This provides us an alternative to a histogram. We
can compare several datasets using this technique at the same time.
3. Cause-and-Effect Diagram. This tool is known as the fishbone diagram because
of its shape. It aims at finding the cause of each problem (effect) and thus leading

page 32
 finally to the solution of all problems.
4. Computer Simulation. Computer-based technique probably requires the
assistance of operations research to prepare the programs.

Some other important tools widely used are:

● Control Chart
● Design of Experiments (DOE)
● Dot Plots
● Fault-tree Analysis
● Gauge Repeatability and Reproducibility

page 33
 3.11.7 FMEA CHART FOR THE
PROJECT

DESCRIPTI FAILURE MODE ANALYSIS ACTION PLAN ACTION

ON RESULTS
Subsystem Potential Effects Causes Probability severity RECOMM RESP. AND ACTION
name: Failure of failure ENDED COMPLETI TAKEN
function and Mode ACTION ON DATE
performance
requirements
COMMAND No Downtime Command Often minor To give - Changed
CAPTURE Motion varying capturing command the device
in
FAILURE from 1 to 5 error either disturbance for giving
minutes due to free command
microphone environmen
t
or too
much
disturbance

page 34
 COMMAND No Downtime Command occasional No To list as - Included all
RECOGNITIO Motion varying different relevant many the possible
N FAILURE from 5 to from set of effect command command
15
minutes command possible
provided that could

bring robot
in motion
PROCESSING No Downtime Processo Extremely severe Check - Reloaded
FAILURE motio varying r failure unlikely arduino the
n from 20 to properly program on
to compile and arduino
120 comman replace it
minutes
d in case
found faulty
CIRCUIT No Downtime Wrong occasional severe Check - Loose
FAILURE motio varying circuit connections connections
n from 5 to connection properly tightened
15
minutes

page 35
ULTRASONIC No Downtime faulty remote catastro Replace - Ultrasonic
FAILURE motio varying senso phic the sensor
n from 5 to r ultrasonic replaced
15 sensor
minutes
POWER No Downtime Either frequent minor Replace the - Replace
FAILURE motio varying discharged batteries or the old
n from 10 battery or charge discharged
to partially them in cells with
120 drained out case if new one
minutes discharged

TABLE 3.2

page 36
3.12 GANTT CHART

page 37
CHAPTER 4

CONCLUSION AND FUTURE SCOPE

In the project, we have successfully developed a speech recognition robot and built a
code for breakdown detection in machines in the whole manufacturing industry or
firms . Robotics will be gaining a fair increase of role in the automotive sector during
the next few decades. Using our code we have successfully been able to detect
mechanical failures using custom sensors depending on the machine , such as
accelerometers for bearings, frequency sensors for transducers, sound and noise
sensors in heavy machinery used in different firms etc. All these sensors will give
input data set for the code and this was normalized to give inputs in the range from -1
to 1 to improve training speeds and accuracy of the CNN model. The project uses AI
concepts like Pandas, Keras, ANN, CNN, Normalization, Stochastic Gradient etc to
create custom models for each machine.

It was also found that the model could predict with accuracy close to 99.8 % for
machines which could provide with large datasets and if companies invest good
amounts in R&D to find out their root causes of failure. Otherwise even with minimal
availability of dataset it gives accuracy close to 70%. This could prove to be a
breakthrough for small industries that could not invest big amounts of money but now
will get a boost to match the process quality and accuracy to those of big players in
manufacturing.

To conclude, this code could be used in various industries, but the major use will be in
the fields of manufacturing and bio-medics.

This project has tremendous future scope potentials. Major motivation for
researchers and future students will be the benefit in cost and improvement in
process quality that this work can bring to any firm. Future work can be on

● Custom codes for each machine

● Build codes and implement household machines to improve longevity.

page 38
● Improve the low dataset accuracy from 70%.

● Industry can improve their quality concepts and remove unplanned

breakdown as one of the headaches. Incorporate the project with the
concepts of TQM, TPM, and 6 Sigma principles.
● Use it to substantially improve worker productivity and boost customer
satisfaction, and reduce the evil of bullwhip.
● Can be used in the manufacturing industry after automation.

● Can be used for military purposes after incorporation of voice recognition.

● Can be used in critical fire fighting scenes employed with voice reduction
algorithms.
● Can be used in situations that are harmful for human purposes.

● Voice driven small cars can be made after increasing the speed and flexibility
of the commands.

● For the sake of security and monitoring.

page 39
REFERENCES:

1. A Review Paper on Infrared sensor Pooja Ajmera Department of Electronics &

Communication Vivekananda Institute of Technology, Jaipur Jaipur, India

2. Allocca, A.J., and A. Stuart. 1984. Transducers: Theory and Application.

Englewood Cliffs, New Jersey: Reston.

3. Arthur, K. 1970. Transducer Measurements. Beaverton, Oregon: Tektronix.

National Academies of Sciences, Engineering, and Medicine. 1995. Expanding the

Vision of Sensor Materials. Washington, DC: The National Academies Press.

https://doi.org/10.17226/4782.

4. http://Infrared sensors - Sensor Technology - Metropolia Confluence

5. http://IR Sensor: Circuit Diagram, Types Working with Applications

(elprocus.com)What is an IR Sensor : Circuit Diagram & Its Working

6. Classroom.google.com/u/0/h

7. Sams Teach Yourself Arduino™ Programming in 24 Hours by Richard Blum -

PDF Drive

8. Figure 2: Bluetooth Module with Arduino, www. Circuit digest.com

9. Figure 3: L298D Pin Diagram google images ,

10. Figure 4: Ultrasonic sensor, www.the engineeringproject.com

11. Figure 5:DC Motor, www.indiamart.com

page 40
12. Figure 1:Arduino IDE, www.learn-sparksfun.com

13. Figure 6: Block Diagram, www.micro controllerslab.com

14. https://youtu.be/Jy9-aGMB_TE

15. https://youtu.be/umGJ30-15_A

16. https://youtu.be/PcNqAV
17. ,https://www.ijert.org/human-voice-controlled-robot-embedded-with-real-t
ime- obstacle-detection-and-
avoidance#:~:text=Human%20Voice%20Controlled%20Robot%20Embed
ded%
20with%20Real%20Time%20Obstacle%20Detection%20and%20Avoidan
ce,- Sourabh%20Marne&text=A%20Bluetooth%20module%20
18. https://how2electronics.com/wireless-voice-controlled-robot-car-using-
arduino/#:~:text=The%20circuit%20consists%20of%20Arduino,Module%20i
s
%20supplied%20with%205V

19. https://create.arduino.cc/projecthub/aqarrout/voice-controlled-car-ee1464

20. https://create.arduino.cc/projecthub/Yug_Ajmera/voice-controlled-car-983ef1

21. https://www.elementzonline.com/download/elementz_0797018001469441810V
OICE%20CONTROLLED%20ROBOT.pdf

22. https://www.electronicsforu.com/electronics-projects/voice-controlled-robotic-
car

23. Artificial Intelligence - PDF Drive

24. https://veltech.edu.in/international/visai19/uploads/VISAI19-
Voice%20Controlled%20Robot%20using%20Bluetooth%20Module.pdf
25. https://youtu.be/Jy9-aGMB_TE

26. https://youtu.be/umGJ30-15_A

27. https://youtu.be/PcNqAVNCZrE

28. https://youtu.be/ewtIxc-MHXE
29. https://www.hindawi.com/journals/mpe/2018/7354081/

page 41
30. https://www.biorxiv.org/content/10.1101/766022v1.full
31. https://github.com/PacktPublishing/Python-Robotics-Projects

32. Figure 2: Bluetooth Module with Arduino, www. Circuit digest.com

33. Figure 3: L298D Pin Diagram google images ,

34. Figure 4: Ultrasonic sensor, www.the engineeringproject.co

35. Figure 5:DC Motor, www.indiamart.com

36. https://www.circuitstoday.com/robot-using-arduino-bluetooth-module

37. Figure 1:Arduino IDE, www.learn-sparksfun.com

38. Figure 6: Block Diagram, www.micro controllerslab.co

page 42
s
ORIGINALITY REPORT

9 %
SIMILARITY INDEX
4%
INTERNET SOURCES
0%
PUBLICATIONS
6%
STUDENT PAPERS

PRIMARY SOURCES

1
Submitted to RMIT University
Student Paper 3%
2
www.ijert.org
Internet Source 2%
3
Submitted to The London College UCK
Student Paper 1%
4
discovery.ucl.ac.uk
Internet Source 1%
5
Submitted to Assumption University
Student Paper <1 %
6
Submitted to Middlesex University
Student Paper <1 %
7
Submitted to University of Huddersfield
Student Paper <1 %
8
Submitted to Middle East College of
Information Technology
<1 %
Student Paper

9
Submitted to Universiti Teknologi Malaysia
Student Paper <1 %
10
www.termpaperwarehouse.com
Internet Source <1 %
11
Submitted to University of Malaya
Student Paper <1 %
12
www.coursehero.com
Internet Source <1 %
13
www.cio.com
Internet Source <1 %
14
www.natalimak1.narod.ru
Internet Source <1 %
15
Submitted to Softwarica College Of IT & E-
Commerce
<1 %
Student Paper

16
Submitted to Universiti Teknikal Malaysia
Melaka
<1 %
Student Paper

Exclude quotes On Exclude matches Off

Exclude bibliography On